The Ethics of Social Roles 0192843567, 9780192843562

Citation preview

The Ethics of Social Roles

ENGAGING PHILOSOPHY This series is a new forum for collective philosophical engagement with controversial issues in contemporary society. Disability in Practice Attitudes, Policies, and Relationships Edited by Adam Cureton and Thomas E. Hill, Jr. Taxation Philosophical Perspectives Edited by Martin O’Neill and Shepley Orr Bad Words Philosophical Perspectives on Slurs Edited by David Sosa Academic Freedom Edited by Jennifer Lackey Lying Language, Knowledge, Ethics, and Politics Edited by Eliot Michaelson and Andreas Stokke Treatment for Crime Philosophical Essays on Neurointerventions in Criminal Justice Edited by David Birks and Thomas Douglas Games, Sports, and Play Philosophical Essays Edited by Thomas Hurka Effective Altruism Philosophical Issues Edited by Hilary Greaves and Theron Pummer Philosophy and Climate Change Edited by Mark Budolfson, Tristram McPherson, and David Plunkett Applied Epistemology Edited by Jennifer Lackey The Epistemology of Fake News Edited by Sven Bernecker, Amy K. Flowerree, and Thomas Grundmann

The Ethics of Social Roles Edited by

ALEX BARBER AND SEAN CORDELL

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © the several contributors 2023 The moral rights of the authors have been asserted First Edition published in 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2022947950 ISBN 978–0–19–284356–2 DOI: 10.1093/oso/9780192843562.001.0001 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Contents Preface Acknowledgements List of Contributors Abstracts of Chapters

1. An Overview of Social Roles and Their Ethics Alex Barber and Sean Cordell

vii ix xi xiii

1

PART ONE: HISTORICAL FOUNDATIONS 2. Tempted Like Achilles: Reflections on Roles and Role-Recalcitrance in Ancient and Modern Ethics Sophie Grace Chappell

25

3. Roles and Virtues: Which is More Important for Confucian Women? Jing Iris Hu

49

4. John Dewey’s Analysis of Moral Agency as Pragmatist Role Ethics John Simons 5. A Good Doctor but a Bad Person? A Puzzle for Role Ethics from Løgstrup Robert Stern

66

87

PART TWO: THE NATURE AND NORMATIVITY OF ROLES 6. All Together Now: When Is a Role Obligation Morally Binding? Erin Taylor

107

7. The Part We Play: Social Group Membership as a Role Tracy Isaacs

133

8. Explaining Role-Based Reasons Reid Blackman

156

vi



PART THREE: ROLES, INSTITUTIONS, AND ORGANIZATIONS 9. Role Ethics and Institutional Functions Sean Cordell

177

10. Role Obligations to Alter Role Obligations Stephanie Collins

200

11. Are Obligations of Friendship Role Obligations? Diane Jeske

217

12. My Job and Its Requirements Thomas H. Smith

236

PART FOUR: WELLBEING, SELFHOOD, AND ROLES 13. Three Relations Between Roles and the Good Samuel Clark

263

14. Participatory Wellbeing and Roles Alex Barber

278

15. Virtuous Chameleons: Social Roles, Integrity, and the Value of Compartmentalization Luke Brunning

298

Index

321

Preface Many of the most significant practical deliberations we make turn on some social role or other we happen to occupy at the time (where ‘social role’ is read widely to include professional roles such as teacher or manager as well as parent, carer, friend, citizen, etc.). Members of the emergency services often put themselves in harm’s way, running towards a danger zone instead of away from it like everyone else. They are admired for doing so, but at the same time this is something they, as firefighters or police officers, are meant to be doing. Morally loaded decision making often requires us to reason and act with a particular hat on, so to speak. Sometimes we must first decide which hat to wear (of colleague, of friend, of manager) or whether to wear a hat at all. Despite these being features of daily experience, the ethical significance of role occupancy has long gone under-acknowledged as a topic within normative ethics. To be more accurate, while certain social roles (including legal, medical, business, military, gender, family, and friendship roles) have been recognized as ethically significant, their significance has been addressed piecemeal. We currently lack a developed literature on the ethical significance of social roles as such—on what they are, on why they appear to have ethical force, on the structure of that force, and on the significance of social roles for identity and wellbeing. The contributors to this volume have set out to fix this, building on the small body of work that does already exist and extending it into new and unexpected areas, bringing out the diversity of ethical questions that arise for social roles. The topic, it turns out, is important for the ethics of various individual social roles, but also for an integrated understanding of a suite of other live topics such as collective agency, impartiality, special relationships, wellbeing and self, and social justice. Our hope for the volume is that it advances the ethics of social roles as a subfield of ethics, at the interface of these other literatures but also deserving of attention in its own right. This will help the English-speaking philosophical world out of a temporary parochialism, since, as several of the contributors help to bring out, this relative neglect of the ethics of roles is something of a historical blip.

Acknowledgements Our work on this volume was supported by a grant from the Arts and Humanities Research Council, which funded the Role Ethics Network between 2015 and 2018 (Project ref: AH/N006321/1), allowing members to meet and discuss work in progress at four workshops and a conference. This same grant, plus The Open University Arts Faculty, supported Codes of Ethics workshops involving nonacademic professionals in 2018 and 2019, and through these we learned how some of the issues dealt with in this volume play out in different professional spheres. We are grateful to all the many participants at these events, as we are to those at a MANCEPT workshop at the University of Manchester in 2013. We are especially thankful, of course, to the authors of this collection for their contributions. They, and the OUP editor Peter Momtchiloff, were both patient and responsive during its long gestation. Three reviewers for OUP also provided helpful steers at a critical point.

List of Contributors Alex Barber is Senior Lecturer in Philosophy at The Open University. Reid Blackman is CEO of Virtue, and was a Fellow at the Parr Center for Ethics, University of North Carolina, and Assistant Professor of Philosophy at Colgate University. Luke Brunning is Lecturer in Applied and Inter-Disciplinary Ethics at the University of Leeds. Sophie Grace Chappell is Professor of Philosophy at The Open University. Samuel Clark is Senior Lecturer in Philosophy at Lancaster University. Stephanie Collins is Associate Professor of Philosophy at Monash University. Sean Cordell is Senior Lecturer in Philosophy at The Open University. Jing Iris Hu is Assistant Professor of Philosophy at Concordia University in Montreal. Tracy Isaacs is Professor of Philosophy at Western University, Ontario. Diane Jeske is Professor of Philosophy at University of Iowa. John Simons is an independent scholar who gained his PhD from the London School of Economics. Thomas H. Smith is Lecturer in Philosophy at the University of Manchester. Robert Stern is Professor of Philosophy at the University of Sheffield. Erin Taylor is Assistant Professor of Philosophy at Washington and Lee University.

Abstracts of Chapters 1. An Overview of Social Roles and Their Ethics This introductory chapter offers a basic framework for thinking about the topic of social roles and their ethics, and in doing so addresses latent resistance to the very idea that the norms associated with social roles have an ethical claim on us. It suggests a working understanding of what social roles are, a default terminology, a set of key questions, a description of links to neighbouring debates, an overview of the other chapters in the volume, and finally a list of key works on a wrongly neglected topic.

2. Tempted Like Achilles: Reflections on Roles and Role-Recalcitrance in Ancient and Modern Ethics This chapter considers the notions of role-compliance and role-recalcitrance by starting from a simple argument-schema for role-reasons: (1) that rolerecalcitrance is a human universal; (2) that at least some role-recalcitrance is ethically interesting; (3) that at least some ethically interesting role-recalcitrance is a very good thing. The argument for (1) and (2) examines some well-known claims that Alasdair MacIntyre offers about “heroic societies” in After Virtue: in particular, his connected claims (a) that people in those societies cannot “step back” from their roles, and (b) that there are arguments across the Is‒Ought Gap that are based on “functional concepts”. A re-examination of the Iliad’s central figure of Achilles refutes (a) and suggests an a fortiori argument for (1). Analysis of (b) leads to a distinction—which refutes (b)—between logical and psychological/sociological cogency; to some reflections on analyticity in general; and to the conclusion, which is a rephrasing of (3), that at least sometimes the ability to be role-recalcitrant is precisely “what makes us truly human”.

3. Roles and Virtues: Which is More Important for Confucian Women? Some influential scholars in Confucian philosophy argue that roles rather than virtues are at the centre of Confucian ethics. The advocates of this Confucian Role

xiv

  

Ethics maintain that roles are not only important for providing ethical guidance for those who fulfil them; roles also constitute the Confucian moral agent. This chapter defends a controversial claim—roles do not constitute moral agency for Confucian women, and nor can they offer sufficient moral guidance; instead, they are supplemented by virtues. The argument for this claim recognizes that, historically, Confucian women, unlike men, experienced—and were educated to expect—a drastic change of roles and social relationships when leaving their maternal family and moving in with their husbands and in-laws. These changes prepared Confucian women to embrace a view of moral agency through which they learn to focus on virtues as a more reliable source of moral guidance than roles.

4. John Dewey’s Analysis of Moral Agency as Pragmatist Role Ethics This chapter renders the analysis of moral agency by John Dewey, one of the founders of American pragmatism, as an account of pragmatist role ethics. For Dewey, moral obligations arise naturally in, and are constitutive of, the role relationships that are necessary to a society’s existence. The purpose of morality is regulation of the mutual reciprocal expectations that people have of one another’s conduct in their enactment of social roles. Dewey insists that acts with moral significance are enactments of the self. Thus, when making morally significant choices of role and performance, the agent is choosing the sort of person she is or wants to become. Deliberation over the choice is influenced by three independent sources of normativity, which often diverge in influence: the agent’s own judgement of the moral praiseworthiness of the choice; its socially recognized obligations; and whether the character trait it reveals will be approved by others.

5. A Good Doctor but a Bad Person? A Puzzle for Role Ethics from Løgstrup This chapter considers the following challenge to social role ethics: 1. Role ethics holds (a) that you can be a good occupant of a role by fulfilling the obligations associated with the role, and (b) that this is sufficient to make you a good person (at least to some degree) provided that the institution itself is a good one 2. But a person can count as having fulfilled their role obligations while being a moral monster 3. So role ethics is not an ethics.

  

xv

The chapter shows how this challenge is suggested by what K. E. Løgstrup says about social norms and roles, but also how he might provide resources with which to meet it, thereby offering a way for the role ethicist to escape the problem outlined above. Finally, the wider implications of Løgstrup’s view for some of the standard issues in role ethics are discussed.

6. All Together Now: When Is a Role Obligation Morally Binding? This chapter defends a novel account of the connection between social roles and their associated demands. Consider pairs of statements such as: (a) ‘Maura is Ethan’s mother’ and (b) ‘Maura has an obligation to provide for Ethan.’ It is natural to think that such pairs of sentences don’t merely state two unconnected truths about the agents involved. Rather, in each case, the truth of (b) seems to be in some sense explained by the truth of (a). Competing theories of the moral status of social roles, and their associated obligations, aim to explicate the nature and significance of this explanatory connection. The present chapter does this by arguing that many of the obligations or norms that attach to social roles morally bind us by default. These norms are morally binding since, first, they are ultimately explicable in terms of conventional moral duties and, secondly, those conventional duties enjoy a default but defeasible good standing.

7. The Part We Play: Social Group Membership as a Role This chapter explores the nature of social group membership understood as a role. We use social group categories such as gender, race, class, sexuality, etc., to identify social groups to which individuals belong. Roles are parts that people play within a given collective context. The context also defines the roles. Roles’ occupants have powers, obligations, and responsibilities qua role occupant. This chapter’s central question is: does considering social group membership as roles enable us to deepen our ethical understanding of oppressive social and political contexts and of what is morally required of us when we are living in such contexts? The chapter argues that a role analysis of social group membership in oppressive social contexts can help people living within them to gain a more determinate understanding of their obligations with respect to the collective goal of eradicating injustice.

8. Explaining Role-Based Reasons Occupants of social roles have normative reasons that non-occupants lack. How do we explain this? Answering the question hinges on issues both in social

xvi

  

ontology concerning the nature of social roles and in metaethics concerning the existence conditions for normative reasons. This chapter begins by laying out the conceptual terrain before sketching what is here called a ‘non-normative’ account of roles in social ontology: a view that rejects the normative account which takes roles to be identical to a set of normative relations. The chapter then articulates and defends a novel account of role-based reasons, Role Pluralism. According to Role Pluralism, role-based normative reasons are externalist (i.e. they do not depend on the motivational constitution of role occupants) and non-hierarchical (i.e. the reason-giving power of a role does not depend on any other role or anything external to a role for that power). Finally, the chapter explains why the most common objection to this view—that immoral roles cannot provide normative reasons—is irrelevant.

9. Role Ethics and Institutional Functions This chapter begins by setting out the problem of role-indeterminacy raised in Michael Hardimon’s ‘Role Obligations’. It then considers the prospects for a solution in terms of a functional account of institutions which determine those roles. After giving some instances of the role-indeterminacy problem and focusing on the case of role-requirements being mis-specified by institutions (‘ersatz obligations’), three ways in which a functional view might be formulated in response to this problem are discussed. First, a backward-looking etiological approach (roughly, what some institution is ‘there for’ in the first place); secondly a present-focused ‘practical’ view (roughly, what it is ‘used for’); and thirdly a forward-looking teleological account in terms of institutional ends (roughly, what it is ‘good for’). After running through the attractions and shortcomings of these accounts, an alternative view is defended as more coherent. Using language borrowed from Aristotle, this alternative treats institutions as having an ergon, a ‘characteristic activity’.

10. Role Obligations to Alter Role Obligations Many roles are situated within organizations. The occupants of these roles often confront a dilemma between (i) the occupancy conditions, performance conditions, and functions of the role, as bestowed upon the role by the organization’s decision-making procedures, and (ii) the occupancy conditions, performance conditions, and function that the role should ideally have. This chapter argues that this dilemma should be resolved in favour of (ii). Yet this does not require forgoing role-based considerations in favour of extra-role considerations. Instead, we should conceptualize organizationally specified roles as being one token of an underlying role-type. The underlying role-type provides the organizational

  

xvii

role-token with its fundamental occupancy conditions, performance conditions, and function. When the token is a poor instance of its type, organizational roleoccupants have obligations—grounded in the fundamental aspects of the role—to challenge this. These are role obligations that call upon one to alter one’s role obligations.

11. Are Obligations of Friendship Role Obligations? Are the obligations of friendship role obligations? This chapter argues that the ways in which ‘friend’ differs from paradigmatic social roles such as ‘spouse’ or ‘daughter’ are more significant than the ways in which it is similar to them, and, thus, we ought not to understand being a friend as a social role nor ought we to understand the obligations of friendship as role obligations. In fact, to conceive of ‘friend’ as a social role would be to diminish the ways in which friendship adds value to our lives and binds us to certain other people. The significance of friendship is in part a matter of how it is constructed by the parties involved and provides a space for understanding oneself in relation to another independent of everyone else and, very importantly, apart from societal expectations and norms.

12. My Job and Its Requirements This chapter concerns the ethics and metaphysics of occupations, such as teacher, waiter, and priest. It argues that teacher is a functional kind, but teachers are not functional objects. If you are a practising teacher, it is likely that you perform a function and serve a purpose, that of imparting knowledge and cultivating minds and skills. This is what teachers, generically, are for, and it is what your school is for. But it is not what you are for. Easily confused senses of ‘job’ are distinguished: occupation, position, requirement, and function. The chapter explains that the requirements of your position and occupation do not entail ability, reasons, or fault (if not complied with), and that if you are not as your position and occupation require, you are not as you ought to be, but it does not follow that there is anything wrong with you. The chapter includes some metaphysical speculation about positions, and examines some sources of, and remedies for, workplace alienation and anxiety.

13. Three Relations Between Roles and the Good What is the relation between roles and the human good? Between our construction, maintenance, and enaction of institutions, and the life which goes well for the

xviii

  

person whose life it is? This chapter reads selected martial autobiographies to explore three relations and what they mean for the nature of the good: 1. Tools for self-shaping: roles are social technology for shaping ourselves towards good understood as fulfilment of desires which are independent of those roles. 2. Good-making practices: roles are parts of good-making practices which transform individuals by creating goods and initiating individuals into them. 3. Self-discovery: roles are a method for gaining self-knowledge. They help each of us discover her unchosen, seedlike, initially opaque self, and thereby discover her particular good, which is that self ’s realization. The chapter concludes that some roles’ relation to the good is that they test and reveal the self and therefore its good.

14. Participatory Wellbeing and Roles The wellbeing that can accrue to individuals through their participation in collective endeavours, here called their participatory wellbeing, is a fundamental component of human wellbeing more broadly. It is also difficult to conceptualize, let alone quantify, and has been neglected in philosophy, apparently falling into a gap between the literature on collective agency and the literature on wellbeing. As a contribution towards filling in that gap, this chapter uses the notion of a role within a group—encompassing anything from familial and professional roles to being a friend or a citizen—to solve a puzzle about participatory wellbeing. The puzzle, crudely stated, is that while wellbeing is an essentially individualistic notion, participation is essentially social. By conceiving of participation as a matter of occupying and performing a role, we can recognize and model the complexity (i.e. the multifaceted nature) of participatory wellbeing.

15. Virtuous Chameleons: Social Roles, Integrity, and the Value of Compartmentalization Most people occupy several social roles. How are these roles to be managed? This chapter examines one strategy: the compartmentalization of roles, in which actions and mindsets change with the roles people occupy. People might worry, however, that compartmentalization is in tension with integrity or one’s commitment to the good. Might an integrating approach to roles be better? This chapter argues that compartmentalization is a fine way to manage multiple social roles. Central to the argument is a response to those who think compartmentalization

  

xix

must inhibit practical reflection, or is in tension with having integrity, or is alienating. In fact, the task of compartmentalizing roles is itself a valuable practical activity. Maintaining clear boundaries between roles and their attendant mindsets enables us to be guided by fully specified virtues, and flexible boundary work can help us avoid inappropriate forms of immersion in roles at the expense of other people.

1 An Overview of Social Roles and Their Ethics Alex Barber and Sean Cordell

1. Introduction A recent sociology article asks what that discipline could possibly have been like before the notion of a social role became as fundamental to it as it is today (Jacobs 2018). In philosophical ethics the situation is almost the reverse, at least in the English-language tradition. There was a time when the importance of social roles to ethical deliberation was widely acknowledged and discussed, most memorably in F. H. Bradley’s ‘My Station and Its Duties’, but that discourse fell dormant during the twentieth century (Bradley 1927 [1876]; Stern 2013). Exceptions to the silence (e.g. Cohen 1966; Emmet 1966; Downie 1971) were rare and largely overlooked, or else were confined to examining particular roles in isolation, with no accompanying generalization into an ethical theory of social roles as such (e.g. Firestone 1970; Walzer 2006 [1977]). This trajectory could be tied to the rise, in political philosophy and in society at large, of liberal individualism, Rawlsian or otherwise.¹ But whatever the cause, appeals today to the obligations of social roles, or of Bradley’s ‘stations’, make many uneasy. They bring to mind threats to the autonomy and equality of all those who would refuse to know their place. And yet social roles, as present-day sociologists (at least) are perfectly aware, are unavoidable in practice, and for good reason. They allow us to manage our coexistence and to act collectively. A moment’s reflection is enough to show that getting rid of them entirely or pretending they don’t exist is neither attractive nor feasible as a response to the restrictive or oppressive character of some of them. A person who insists they shouldn’t have to abide by role norms is, after a certain point, going to sound like a driver who insists they shouldn’t have to slow down for corners. And as with driving, roles have an ethical character, which is where philosophy comes in. Classic sociological discussions of roles (e.g. Linton 1936;

¹ Rawls himself typically subsumes discussion of social roles within his broader framework, as for example when considering the position of ‘certain representative individuals’ when applying the principles of justice to the basic structure of society, where this must be from a ‘suitably general point of view’ (Rawls 2000 [1971], 81–82). Alex Barber and Sean Cordell, An Overview of Social Roles and Their Ethics In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Alex Barber and Sean Cordell 2023. DOI: 10.1093/oso/9780192843562.003.0001

2

  

Goffman 1990; Dahrendorf 1968; Hilbert 1981) have been geared primarily towards a social-scientific understanding of roles, and so can inform but not settle normative questions about roles. And in case there is any doubt that role occupancy can be morally charged, here are just three commonplace ways in which social roles have ethical salience. Special obligations Firefighters and lifeguards discharge role obligations when they carry out lifeendangering acts of rescue, acts that for passers-by or sunbathers would be morally supererogatory or reckless. Less urgently but just as clearly, an admissions tutor has a duty to tell an ambitious undergraduate student that he is ill-equipped for doctoral study, assuming this is true. For someone outside that role to impart that same truth to the same student in the same words and in the same manner could be morally questionable, perhaps even cruel. Wearing many hats We all have to juggle the different roles we occupy, whether it be professional, familial, citizenship, or friendship roles. Sometimes these create conflict, as when a colleague who is also a friend starts underperforming in one or other of these roles. Roles provide us with normative levers and pulleys in our everyday practical reasoning, and when the mechanisms get jammed or clash, it is only by thinking about what we should do as friend, as colleague, as friend-andcolleague, that we can negotiate our way towards right action. Fulfilling roles Roles, and our performance of them, can influence the fuller story of an individual’s life as well as its episodes. Cases of life-changing career paths or vocations abound. Maybe there really are, for example, some social workers, schoolteachers, or carers whose ‘egoism has disappeared unobtrusively into the care and service of others’ (Murdoch 1992, 429). Or maybe not, but undoubtedly these and other roles (e.g. parent as a ‘transformative’ role, Paul 2015) can impact in differing degrees on their occupants’ development as persons.

On the face of it, then, social roles have an ineliminable impact on the moral status of certain actions and perhaps even on the assessment of moral character (Pettigrove 2020). In the sociology article mentioned earlier, Struan Jacobs identifies personality as the notion displaced by role within that discipline. His suggestion is noteworthy in the present context because a pervasive, latent, animosity to social roles within normative ethics often turns on the contrast between social roles and the persons who take them on. Most of us will at some point have felt a conflict between the ‘me’ who is required to face and interact with the world in some role-prescribed and perhaps stereotype-prone way, and the ‘real me’ who wants to shake off those

       

3

shackles. This urge is at its strongest when the role is a bad fit for ‘the person we truly are’, as we might put it. It is striking, then, that the modern word person derives from the Latin word persona, meaning mask. The masks in question were worn on stage, so a derivative usage emerged, naturally enough, and it came to mean persona on or off stage. Centuries later we have somehow ended up with the word ‘person’ coming to mean the individual behind the mask rather than the mask itself. The word has flipped, in other words, which is why we now use ‘person’ to mean the individual occupying the role, distinguishable from the role itself. The other word in this binary opposition, role, also has a noteworthy history. It appears to have the same root as roll in the sense of register or ‘roll call’, both deriving from the practice of keeping a list of those entitled to practise a profession on a roll of parchment (OED, roll, n.1, I.2.b). The idiom ‘struck off the rolls’ is a hangover from these origins. But the word soon spawned derivative senses. It began to be used to describe characters on stage, for example, and to be used to describe an individual’s contribution to a one-off act, as opposed to a role that is relatively stable through time. Thus, on a hike, we might now speak of its being soand-so’s role to shut the gate since they were the last person through it. A third expansion of meaning is that it can now be used of social roles that aren’t normally subject to control through a register, such as the role of friend or sister. So much for English etymology. In the end, how we use these and other words in philosophy is up for grabs. Nonetheless this short semantic history lesson does carry us quickly to the very heart of the philosophy of social roles. That is, every one of the ethical questions addressed by contributors to this volume turns in one way or another on the relation between these two abstract notions, that of an individual person and that of a social role—an occupier and a thing they can occupy, so to speak. This alone makes the topic a rich seam for philosophers to mine. In the rest of this chapter we offer an overview of the topic. This is to some extent a constructive or imaginative act since, as we have been emphasizing, there has been a latent disregard for the ethics of social roles as a unified topic in modern philosophy. That said, we do not wish to pretend there has been no prior work in this area to survey. We and other contributors are building on the steady trickle of publications that began to appear in the late twentieth century. This includes work on the moral dimensions of professional roles (Freedman 1978) and social roles more generally (Andre 1991), most notably Michael O. Hardimon’s influential paper ‘Role Obligations’ (1994). Samuel Applbaum (1999), for example, offers a systematic critique of the view that professional public roles can justify what would, outside those roles, be morally prohibited. There has also been growth in a virtue-theoretic perspective on the ethics of roles (Cordell 2011, Swanton 2016), and on particular professional roles (Solomon 1997; Van Hooft 1999; Oakley and Cocking 2001; Solum 2003; Hursthouse 2007; Swanton 2007), including more recently a special journal issue on the topic

4

  

(Hamilton 2016), and a collection of essays (Dare and Swanton 2020) which expands that topic into novel territory such as the relation between roles and reasons. Much of this work has been written in isolation from the rest or has focused on specific questions within the broader topic of roles and their ethics. With this in mind, our citations throughout this introductory chapter constitute a kind of proto-literature for a field we regard as still nascent. In Section 2 we address the obvious starting question of what social roles are, attempt to identify their most salient properties, and offer a neutral platform— including a default terminological toolkit—from which to think about their ethics. Section 3 surveys the range of key ethical questions one might ask about social roles, without taking a stand on any of them. Section 4 highlights important links to neighbouring literatures. We end, in Section 5, with an overview of the four sections into which contributions to the present volume have been organized.

2. What is a Social Role? Occupancy, Performance, Function Each contributor to this volume rightly presupposes some particular understanding or definition of social role, from which they go on to argue for a particular substantial claim. Here, we explore the question of what social roles are, less with an eye on advancing a particular thesis and more as a way of giving the volume’s readers a default way of talking and reasoning about social roles. These same readers are encouraged to diverge from this default as they see fit, just as other authors already do. The simplest way to capture the intended meaning of a phrase is often to point at paradigm cases. ‘Social role’ as it figures in the title of this collection is no exception. As we understand the term, spouse, doctor, friend, colleague, parent, and public intellectual are all social roles. But as with any definition-by-list, this leaves us wanting more. What makes these examples instances of the kind? That is what we really want to know as we attempt to get to grips with the nature of social roles and their ethics. It is also what we need if we want our definition to tell us how to categorize novel or controversial cases, such as ‘influencer’, or ‘being Black’ (Fanon 2008 [1952]; Mills 1997; Kisolo-Ssonko 2019), or to explain apparently similar but contrasting cases such as lorry driver (a social role in that it is a job) as opposed to car driver (arguably just something one does, not a social role). And finally, it is what we need if we are to establish whether there is even a useful, univocal notion underpinning our group of paradigm cases. Hardimon, for example, distinguishes non-contractual roles—such as citizen and familial roles—from contractual ones such as professional and occupational roles. He then includes both contractual and non-contractual roles within a stipulated category of institutional social roles from which he excludes ‘friend’, since it is non-institutional (1994, 335–336). Such distinctions are best thought of, we

       

5

suggest, as distinctions between different ethical sub-categories of social role rather than between things that are genuine social roles and things that are not.² In what follows, we offer what is intended to be a relatively neutral ontology of social roles. It is neutral (or so we suppose) in two senses. First, and as already stated, it is designed to supply a rough and ready framework for readers new to the topic, and not meant to settle substantial questions. But the account is neutral in a further sense: it spans across a very broad range of social roles, rather than homing in on one type in particular. And here it is worth pausing to notice just how many dimensions of variation there are here, all potentially ethically significant. We have just mentioned the contractual vs non-contractual contrast. To this we can add: Voluntary vs involuntary (e.g. adoptive parent vs conscript) Open to all vs by invitation only (e.g. consumer vs godparent) Institutionally embedded vs free-floating (e.g. Deputy Dean vs friend) Highly codified vs loosely specified (e.g. commanding officer vs parent) Socially endorsed vs contested or extinguished (e.g. teacher vs slave-owner) Biologically specified vs socially laden (e.g. sibling vs ‘Mum’) Short-term vs enduring (e.g. juror vs hereditary monarch) Broad in brief vs narrow in brief (e.g. public intellectual vs expert witness) Decision making vs decision facilitating (e.g. patient vs surgeon)

Despite this variety, we suggest, all social roles have three core features: they are occupied, they are performed, and they ostensibly have a social function of some kind. We can travel a long way towards a default understanding of social roles and their ethics by thinking about each feature in turn.

Occupancy Social roles have this first core feature in common with dramatic roles. Both can, in principle anyway, be occupied by different people. We can take on and let go of social roles, just as actors can take on and leave parts in a play. Shakespeare’s character Jaques in As You Like It spots the parallel: All the world’s a stage, And all the men and women merely players; They have their exits and their entrances; And one man in his time plays many parts.

² Hardimon himself would quite possibly accept as much, since his restriction of the term to institutional roles is motivated simply by a desire to say how he plans to use the term in a paper focused only on roles of that type.

6

  

Some social roles are more freely entered or left than others, it is true, while others do not survive the departure of a particular occupant, but all social roles seem in principle to have what we can call occupancy conditions, norms that specify when one is to count as a legitimate occupant of the role. Taking our cue from Shakespeare’s Jaques, we break occupancy conditions down into entrance conditions and exit conditions. The usual entrance condition for the role of monarch, for example, is that one be the eldest son, or else the eldest child, of the most recently deceased occupant of that same role. Its usual exit condition is death, a kind of limiting case in which one leaves the role by leaving the world though the role itself persists. Occupancy conditions are not always so readily stated. It is not easy to specify what counts as no longer being a person’s friend, for example, as we learn in the schoolyard. Such vagueness is nonetheless compatible with the role of friend being something one can enter and leave. The entrance and exit conditions for the monarch role, for that matter, are not as straightforward as was just implied. Usurpers have been bypassing primogeniture throughout history, and abdicators have sought to escape the role while still alive. It is unclear whether these scenarios constitute violations of the occupancy conditions or are instead an aspect of the occupancy conditions (on the grounds that showing willingness and having the power to hold onto the crown are a sign of being qualified). As soon as one scratches the surface one finds important distinctions: for example, between what the occupancy conditions are said to be and what they are in practice, or between what they are (officially or in practice) and what they should be. To specify a social role’s entrance and exit conditions—capturing or resolving such vagueness and ambiguity—is part of what it is to specify that social role. These occupancy conditions will differ vastly from role to role, of course, often in ethically salient ways. Voluntariness is a key variable here. In a liberal society the norm is to avoid coercive social roles wherever possible, since levering people unwillingly into social roles has been a way of denying their freedom throughout history. Even now, though, pure voluntariness, where it is entirely up to the individual whether and when they can take on or shed the role, is relatively rare. Contractual employment usually requires some notice, on pain of penalty; and to abandon a friendship simply because it suits you—because your friend has been struggling recently and it is getting bothersome—may well be to violate a norm of friendship. For other roles, such as the role of child or sibling, the nonvoluntary aspect is more obvious: you don’t get to choose your parents, and insofar as there are filial obligations they are handed to one at birth. A parent, likewise, may be thought of as locked into their role by default, long after they entered the role voluntarily or otherwise (though there is room for disagreement here; compare Jeske 1998 and Porter 2014). But what does being ‘locked in’ in this way entail? More generally, what follows from occupancy? For this we must turn to our second core feature.

       

7

Performance The analogy between social roles and acting is useful here, too. Just as social and dramatic roles are both things one can move into and out of, so too they are things the occupant is expected to perform. Getting the part is one thing, playing it something else. In the case of social roles, we must look to what we call its performance conditions, norms that attach to someone by virtue of their occupying that role. The most obvious examples of performance conditions are things a person has responsibility for doing by virtue of being the occupant. A parent, by virtue of being a parent, is obliged to safeguard the interests of the relevant child or children. To fail to do this is, in extremis, to have failed not simply as a parent but as a moral agent. But alongside these duties, these role obligations as they are generally called, performance conditions typically include a range of powers, what we will call role entitlements. The acting analogy will run out of steam eventually but it is still helpful to illustrate the point. An actor in the role of Richard II is obliged to deliver the lines, of course, but they are also entitled to have sufficient time in which to study and learn the part, a space in which to rehearse, a stage from which to deliver the lines, an audience to deliver them too, alongside other actors, with guidance from a director, and perhaps some financial recompense to make all of this feasible for them as human beings with other pressures on them outside the role. Likewise for an occupant of the role of police officer, who has both role obligations (primarily to maintain the peace and uphold the law, with other obligations flowing from that) and certain powers that come with the role. These latter include the power of arrest, for example, which is importantly different in scope from, say, citizen’s arrest. Because the power of arrest and the associated obligations come with the role, a police officer loses these as soon as she exits the role. ‘Come with’ is an understatement: that power and those obligations are partially constitutive of the role, and to specify a set of performance conditions is, as with specifying a set of occupancy conditions, part of what it is to specify the role. We find some of the same looseness, the same ambiguity and vagueness, with performance conditions as we did with occupancy conditions. Even with roles for which the duties are tightly specified, by a professional code of conduct for example, the adequacy of that specification is itself open to challenge. Still, all roles, excepting those with a purely honorary function, seem to come attached with some or other performance conditions. And sometimes the vagueness is only there if you set out to find it. Famously, there is ‘no single right way to parent’, but this is misleading. To parent requires one to act in accordance with the primary responsibility one has for the wellbeing of the child. The ‘no single right way’ mantra is correct insofar as there is more than one way to exercise that primary responsibility, but that is true of every role and does not necessarily amount to imprecision, any more than there being more than one way to commit murder stands in the way of that notion’s having a tidy definition.

8

  

Verbal slippage of a different kind can sometimes blur the distinction between performance and occupancy conditions. We can use the word ‘parent’ to refer both to what someone is and to what they do, for example, which is why an assertion of the sentence ‘My dad wasn’t a dad’ would make sense. Sometimes the blurring of the distinction goes beyond simple messiness of language, since failure to perform well can itself be the trigger for counting as having left the role—a tacit exit condition. Failure to fulfil the usual obligations of friendship, perhaps, is to have ceased to be a friend, just as a sentry who leaves their post is no longer a sentry except perhaps in name. (This is not always the case: a teacher between jobs is still in some sense a teacher.) Neither species of blurring—the slipperiness of ordinary language or the use of performance as an indicator of occupancy—shows that we should jettison the distinction between what counts as being in the role and what counts as performing the role.

Function Taken together, the occupancy (entrance and exit) conditions and performance (obligation and entitlement) conditions of a role specify what the role is. We might suppose that in a well-designed role, its entitlements will cohere with its obligations, and both should cohere with the occupancy conditions. The entitlements should enable a suitable occupant to carry out the associated duties effectively (otherwise we get ‘all responsibility and no power’) but should not be excessive (otherwise we get ‘all power and no responsibility’); there should be suitable occupants (the entrance conditions should not be too stringent and the exit conditions not be too liberal) but not too many occupants (with everyone in a business occupying a leadership role, for example). Society’s roles are defective to the extent that its various conditions fail to coalesce into realistic vehicles for individuals to contribute to and benefit from that society. The word ‘defective’ is warranted only because of the third core feature of roles: that they have some function, some purpose. Or rather, they ostensibly have a function, some purported benign or positive purpose in society or within an organization. One purported function of the role of entrepreneur, for example, is to generate wealth, and not only for the occupant of the role. This is what the role is ‘for’, or should be for. If a social role turns out to lack the purpose it was advertized as having—if entrepreneurs just use their vast wealth for whimsical projects and no one else benefits, for example—this doesn’t mean it is not a social role after all, it just means it is open to criticism from that angle. By contrast, it jars to describe being a thief as a social role, because it normally lacks even a purported social function, Hollywood and Robin Hood mythology aside. To talk of the criticizability of roles—because they lack an ethical or other function or have poorly balanced occupancy and/or performance conditions—is

       

9

to bring us close to their ethical character. Since we aim to remain neutral on substantial questions, we will not here attempt to be more specific on these matters. Still, if only to illustrate how the framework we have presented can be used to discuss ethical and socio-political topics in this area, consider the following ‘function first’ approach to the optimal design of roles. This approach says that, when thinking about what makes for a good role, one must always begin by clarifying what benign outcome its effective performance would bring about. This can be used to inform us what its role obligations should be—they will be the things that need to be done by the occupant if said benign outcome is to come about. The function of the role of fire-fighter, for example, is tied in a very obvious way to what someone in that role is obliged to do, with both obligation and function on display in the name. A set of obligations that does not guarantee this is hence criticizable. And from the role obligations we can infer a certain set of role entitlements, a set of conditions that make it de facto feasible for an occupant to perform the duties of the role and hence yield the assumed social good (safety equipment, right of access to mains water, and so on). This function-first approach appears more sophisticated than old-fashioned talk of ‘stations’, where static and unchallengeable authority structures were somehow given. But like that old-fashioned talk, and perhaps too like the functionalist perspective on roles associated with the sociologist Talcott Parsons (1937; see also Merton 1968), it can be criticized on the grounds that, in putting the function of roles first, it puts the humanity of their occupants last.

Other Properties of Roles ‘Core’ does not mean ‘necessary and sufficient’, and these three core traits of roles are not meant to be definitional, if only because some things have them but are not social roles. Promiser and promisee, for example, are roles but not social roles. Exactly what makes a role a social role is not something we try to settle here. Many authors, us included, often omit ‘social’ for brevity, but its use in the present book’s title is intended to signal that we mean roles that are in some degree shaping of or pervasive within a person’s life, in a way that being a promiser is not, save perhaps insofar as it enters into the specification of a social role (marriage vows, say). This will be a matter of degree, but it is enough to make sense of why lorry driver seems like a role while car driving is simply something one does. The three core features discussed above are not exhaustive of the common properties of roles, either. When discussing or reasoning about social roles, there are several other properties it helps to have in mind. The four we think are especially worth mentioning are: semantics, institutional context, geography, and realization. These are not features of all social roles, which are, as we saw, highly diverse in character; but they do help us to spot fault-lines running through debates in this domain.

10

  

The first property will seem banal: social roles often have names. Some of these, such as ‘friend’ or ‘teacher’, are well established, whereas some social roles have no name at all, as was presumably true of thought leaders before they began to call themselves that. Sometimes a name can be ad hoc or metaphorical, as with the description of someone as ‘the Art Garfunkel’ (their role is to be a good voice, so to speak, not to write the songs). Sometimes the persistence of a label is arbitrary. Surgery and even the use of leeches was once carried out by people called barbers, and it is a matter of choice whether to count today’s barbers as performing a different role with the same name, or an evolved version of the same role. Sensitivity to the practice of naming roles can be important to debates about their ethics. We have already seen one way in which this is true: role names, such as ‘dad’, oscillating between characterizing someone as an occupant and characterizing them as a performer. A second way is when trying to identify the performance conditions of a role. What should teachers be required to do? ‘Teach’, comes the obvious reply. But modern teachers are required to do a good deal besides teaching in the narrow sense, including administrative and social welfare work. Is that what ‘teaching’ means now? That seems a stretch. But nor does it seem reasonable for a teacher to reject such duties purely on semantic grounds. There is an interesting debate here about how to interpret assertions such as one that will have a familiar ring to many reading this article: ‘I’m an academic, and that means doing both teaching and research, not just teaching and administration, not obedience to a manager, and not being denied the time or resources to contribute to society’s knowledge base.’ Is this just a semantic claim, open to refutation by an academic manager who simply refuses to accept the stipulation? Or is it a semantic claim rooted in some real essence of the social role of academic? Another possibility is that it is a claim about what a well-designed institution of higher learning should have at its core, in the same way that a ferry company should have at its core things that move across water with the capacity to carry people and vehicles. We will not try to settle the issue here, and merely note that different roles may call for different answers. The role named by ‘teacher’ may correspond to a métier, but the one labelled ‘Systems quality assurance analyst: marketing’ seems to be more about slotting into the requirements of an institution. This brings us neatly to our next ‘other property’: many social roles have an institutional context. Professional roles such as academic and lawyer certainly do. It may be possible to have roles with no institution, as with a musical band or a casual friendship, unless one stretches the word ‘institution’ to mean any social group. Other roles seem to belong to more than one institution. If I am someone’s sibling, then I am so within the context of a family, which might be called the host institution. But alongside that particular family I am also participating in ‘the institution of the family’, which is society-wide, period-specific, and often harder to nail down. The role of spouse, too, implies a social and legal framework

       

11

alongside the particular family created or extended through the act of marrying. One can talk of the institution of friendship, too, though it is less clear this picks out anything definite since the ‘rules’ of friendship, unlike marriage, are generally left to the individuals in the friendship group. There is no legal framework for the ending of a friendship as there is for divorce, for example. This contrast brings out why institutionality matters in debates about the ethics of social roles: the institutions associated with social roles are a potential source of both its normative content—its occupancy and performance conditions as we have been putting it— and the apparent bindingness of those norms on the role’s occupants. A third property of social roles to be alert to is their ‘geography’. As we will put it, social roles sit within a rolescape. Both spatial metaphors seem fitting given how we can zoom in or zoom out when we are thinking about a social role. A single individual could be, say, a biologist at a university, but we could zoom in on that role to discover that, for this individual, it consists in part in their doing some teaching, or more specifically still of their doing some first-year undergraduate teaching, alongside much else. These are subsidiary roles, not extra and on top of the role of their being a professional biologist, and they do not represent the only possible way of having that role. Equally, we might zoom out and see the individual as occupying the role of academic (as opposed to biologist in particular), or further out still, to the roles of educated professional and citizen. Each of these may bring with it certain responsibilities or entitlements. The situation here is analogous to one familiar from discussions of action in philosophy of mind. Suppose someone fires a gun. If we zoom in, we see that they pulled a trigger. If we zoom out, we see that they have murdered someone. As with roles, the relatively fine-grained actions correspond to ways of implementing the relatively coursegrained ones, not numerically distinct actions. But notice that human beings do also sometimes occupy more than one social role, and this multiplicity is achieved without the different roles nesting within one another. Jaques again: ‘One man in his time plays many parts’. He meant in sequence, but one man can also end up playing many parts at the same time, which can give rise to friction, paradigmatically in the case of those juggling the role of care-giver at home with that of employee—though that particular friction has typically been experienced less by men than by women. This last example also illustrates an important feature of the rolescape: the social roles that populate it, and hence the repertoire of social roles available to take on, are not fixed. Gender roles, along with the language used to describe them, change over time. Like social roles more generally they are malleable, morphing under various pressures, including recognizably ethical pressures. The mutability of the rolescape is linked to the fourth and final property to beware of when thinking about the ethics of social roles, which is that in any particular society and period, some social roles are realized while others are not. We will not attempt to say here what it is for a role to be realized, save to say that it looks to be

12

  

an instance of a question discussed in social ontology as a whole (on which see Epstein 2018). Roles that have lost reality may have been forgotten or may be remembered, fondly or otherwise (e.g. lighthouse keeper, lord of the manor), and roles that have yet to exist (e.g. citizen of the world) may be championed or lampooned. Some roles are effectively inconceivable from within some social settings. An example, venture capitalist viewed from the perspective of feudal Europe, is a reminder that realizing or extinguishing a role is not, or is not always, simply a matter of having a good idea. Social roles are woven into the fabric of society and not just of organizations, informal groups, or families. Realizing or eliminating them can entail society-wide transformation. Any account of the ethics of social roles will for that reason often have a socio-political element.

3. Questions for an Ethics of Social Roles One ambition we have for this volume is that it should energize work on the ethics of social roles. To that end, in this section we offer Nine Outstanding Questions in the field as it now stands. Some of these arise from points made in Section 2, and they are all expressed in the language introduced there. The questions are interconnected, inevitably, and a plausible answer to any one of them is going to presuppose a credible answer to at least some of the others, and indeed a plausible perspective on neighbouring literatures (see Section 4 below). Q1

What is a social role?

We have already offered a tentative, broad, default answer to this, but that answer is open to refinement or indeed rejection. Definitions are always theoretically committed, in normative ethics as in any other sphere of enquiry. The group of things that we might initially think of as roles could turn out not to belong to a single kind, with a unified ethics, and this might be reflected in one’s operating definition of social role (Thomas 1987, 230; Jeske, this volume). It is easy, too, to envisage perspectival divisions like those found elsewhere in philosophy, such as between reductionism, the view that social roles are compounded out of (and inherit their properties from) other socio-ethical categories, and anti-reductionist alternatives that treat social roles as a sui generis category. Q2

When should an existing social role be defended, when should it be reluctantly tolerated, and when should it be actively eradicated; or, if it is not currently realized, when should it be introduced or resisted?

In the terminology we introduced earlier, this is a question about the kinds of roles we want realized in our rolescape. We take this question to include questions

       

13

about whether a role should be reformed, not just whether it should be displaced, since that distinction will often be arbitrary (see our point about semantics, above). Q3

Why do the norms associated with social roles have a claim on us (if indeed they do)?

This is a question about the source of the normative force attaching to or governing roles. For role obligations to bind us, must they be in some sense contractual in nature, whether through our consent or more weakly through our identification with the role (see Hardimon 1994; Simmons 1996; Sciaraffa 2011)? Or, might rolenormativity be conventional in nature, and so potentially not subject to consent yet binding all the same (Taylor, this volume)? Lastly, could some or even all roles have a distinct and possibly sui generis source—as with, perhaps, the special norms arising out of strong familial bonds (Jeske 1998; Kane 2019)? Q4

If the norms associated with a role (including both its occupancy and performance conditions) are not clearly specified, how can we determine which actions it prescribes? Even if the norms are clearly specified, e.g. in a robust code of conduct, how should we evaluate the codes themselves?

Unlike Q3, this is a question (or two closely related questions) about a social role’s normative content rather than about the source of its normative power. The next question is about neither the source nor the content of a social role’s normativity but about its reach. Q5

When may a social role’s occupants ignore or override its prescriptions, and to what extent are they entitled or obliged to shape the role themselves?

This question turns on a tension at the heart of the way social roles operate in practice. On the one hand, social roles are a tool, and like any tool they exist to serve humanity rather than the other way round. Non-optimal roles should therefore be open to challenge, we might think, not least by the occupant, who will often have special insights. On the other hand, and again as with any tool, social roles need to be respected and used properly if they are to do their job. You don’t get efficient collaboration if every occupant of a role is challenging its dictates at every turn. Q6

If it has one, how independent is a role from its host institution?

This question overlaps with several of the previous ones but is framed as a question about the extent to which the institutions to which social roles often belong are final arbiters. It also raises the question of what should steer the

14

  

decisions of the institution. This illustrates a point made at this section’s outset, that answers to any one of these questions must ultimately be part of a package that answers them all. The final three questions show that answers also need to be informed by perspectives in adjacent fields, several of which are described in Section 5. Q7

How does role occupancy inform the self-conception and wellbeing of the occupant, or of someone barred from occupancy?

This question in turn raises the question of how access to roles, and the social goods (or in some cases harms) that come with that access, should be governed in a just society: Q8

How should theories of social justice accommodate the twin phenomena of fulfilling-but-scarce roles and unfulfilling-but-necessary roles?

Our final question concerns the conflicts that can arise between the demands of two or more different roles, sometimes occupied by different individuals and sometimes occupied by a single individual. Q9

Is there an ethical perspective we should take that sits above the norms attached to particular roles, which we can and should look to if we need to settle questions about their ethics?

This last question has both practical and theoretical resonance. Practical: most of us have been in the situation of not knowing how to act because we are unsure which hat we should be wearing, or whether we should be taking all hats off and responding sub specie aeternitatis, assuming there is such a thing. The abolition of slavery was justified by appeal to a universal humanistic ethics, yet other cases are less straightforward. Being a spy unquestionably requires dishonesty, and much of what Machiavelli says about political leadership in The Prince (1999 [1532]) is still pertinent today (Tillyris 2015). So, must a spy or a political leader choose between their role and morality? Theoretical: this question is asking, in effect, how the ethics of roles should mesh with the rest of normative ethics. A good answer to it will need to be informed as much by one’s wider ethical perspective as by one’s views on social roles.

4. Related Ethical Debates Social roles figure in several neighbouring branches of ethical, social, and political philosophy, in ways often acknowledged only in passing in the relevant literatures.

       

15

Social roles arguably constitute a kind of missing link between these other debates, and recognizing these interconnections has the potential to bring clarity across the board. In this section we briefly highlight how social roles sit in relation to six more established ethical topics: collective agency, impartiality and special relationships, wellbeing, social justice, dirty hands, and discussions of specific social roles. Interest in the metaphysics and ethics of collective (or group) agency has surged over the past few decades (French 1984; Rovane 1998, 2004; Gilbert 1996; Schmitt 2004; Miller and Mäkelä 2005; Sheehy 2007; Isaacs 2011; List and Pettit 2011; Lawford-Smith 2012; Tollefsen 2015; Hess, Igneski, and Isaacs 2018; Thomasson 2019). Any social group or collective is made up of individuals, and the relationship between these individuals and the collective can usefully be thought of as mediated by their roles within it (Miller 2010, 52–54; Ritchie 2018, 22). The framework of occupancy and performance conditions outlined above helps us to see why. Thus, what we have been calling ‘occupancy’ looks to be what constitutes membership of a group, whether it be a highly structured organization (Ludwig 2017, 2) or more of an aggregate, such as people on a cruise ship or in a country. It is not easy, certainly, to see how one could be the member of a collective without having some role or other within it, however passive (, , ,  ). In addition, an individual’s contribution to the agency of a collective (as opposed to their bare membership of it) looks to be nothing other than their performance, effectively or negligently as it may be, of their role. If this is right, then many established debates over the metaphysics and ethics of collective agency will be intimately tied to debates over the nature and ethical properties of roles (Collins 2016, 2017). A second related literature is that of impartiality and special relationships. Certain relationships, such as friendship or familial bonds, have puzzled ethicists because they seem to justify attitudes and behaviours that pull against the spirit of impartiality (Scheffler 1997; Jeske 2001, 2019; Keller 2013), a spirit that to many has seemed central to the very notion of morality, which is why Lady Justice wears a blindfold. Viewing our positions in these special relationships as roles may help us to resolve this tension, though any such dividend will depend on us also having an improved understanding of the ethics of roles more generally. But for example, and again using the framework of the previous section, we might better understand debates about friendship if we distinguish questions over when and how to enter or exit a friendship from questions about how to behave when in one. Putting it this way makes friendships sound rather like quagmires of duties and demands, making a mystery of the fact that we prize our friendships as a primary source of human wellbeing. That holds true of our social-role relationships more generally, which is why they deserve a higher place in a third existing literature, or rather cluster of literatures, dealing with wellbeing, self-conception, and identity (Brunning 2018). Sociologists have been ahead of philosophers, here, at least since Irving Goffman’s The Presentation of Self in Everyday Life (1959), as have

16

  

philosophers outside the mainstream English-language tradition—think of Simone de Beauvoir’s account of womanhood or Jean-Paul Sartre’s waiter (Beauvoir 1973 [1949], 301; Sartre 2018 [1943], 102–103; and see Smith, this volume). In any case, what we are and do, how others regard us, how we regard ourselves, are all ingredients in how well a person’s life is going for that person and are all intimately tied to the social roles we occupy. A comprehensive theory of the nature and value of wellbeing must therefore recognize and accommodate rolegenerated personal fulfilment. Discussions of wellbeing (or welfare) are themselves closely related to discussions of the nature of social justice. Social roles are occupied by individual agents, but they are also parts of wider social structures and so it will be important to be able to recognize when certain roles are parts of unjust, oppressive structures (Kisolo-Ssonko 2019), and when roles could afford their occupants opportunities to ameliorate injustices (Zheng 2018; Isaacs, this volume). Another debate concerns the types and patterns of roles a just society could or should instantiate. To put it bluntly, some roles are more rewarding than others. The ones we get to occupy and perform shape our lives, even to the point of making or breaking them. How should theories of social justice incorporate this fact about roles? How, for example, can the egalitarian impulse be reconciled with the obvious need for a diversity of roles? Should what we have called the ‘rolescape’ be manipulated into something humane and fair, or would such manipulation inevitably be inefficient, a violation of personal liberty, or both? Our fifth example of an adjacent literature is the one on the problem of dirty hands (Walzer 1973; de Wijze 2007; Coady 2009). This is a peculiar kind of moral conflict in which a person must do evil to do right, or in Michael Stocker’s words, do what is ‘justified, even obligatory, but none the less wrong and shameful’ (1990, 9). The problem has been cast as particularly acute or pervasive for political actors and roles of high public office, especially those serving pluralistic democracies (Bellamy 2010; Archard 2013). But the dirty hands conflict can potentially arise for a wider range of social roles. A spy is obliged qua spy to engage in lying— an honest spy is a rubbish spy. On the assumption that spy is not a role we should banish to the museum of history to sit alongside slave owner and absolute monarch, we seem to have to acknowledge that lying is not always the universal moral transgression it is commonly assumed to be. The dirty-hands puzzle, viewed from this role perspective, is a question about whether, how, and why role requirements sometimes bypass, place limits on, or violate ordinary moral standards. It raises the possibility that the assumption of an ‘ordinary’ or role-neutral morality is a fantasy; but even if we decide it is not a fantasy, we are left facing the question of how role-situated and role-transcendent norms are supposed to interact. A better understanding of the ethics of roles, building on existing work on dirty hands, ought to help us address these questions. Finally, although the volume aims to further our understanding of the ethics of roles as such, its contributions are inevitably informed by—and will in turn

       

17

inform—literatures on specific social roles within applied ethics. We have already mentioned friendship, but others include gender roles (Moller Okin 1989; Haslanger 2012, chapter 7), military roles (Sandin 2007), business roles (Sternberg 2000; Swanton 2007), medical roles (Alexandra and Miller 2009; Cribb 2011), legal roles (Luban 1988; Dare 2009); the roles of engineer (Nichols 1997) and journalist (Harcup 2007), and familial roles (Jeske 1998).

5. The Contributions to This Volume If the topic of social roles and their ethics were better established, and if this were a handbook, we might have asked contributors to address some specified subliterature or sub-topic, perhaps by dishing out some variant of our Nine Outstanding Questions (Section 3). We were not in that situation, so we instead invited authors to explore the unexplored, to follow their instincts as to which questions most urgently needed examining and which points deserved greater recognition. What has resulted is thus, unsurprisingly but pleasingly, wide in scope. While it ranges over each of those nine questions there is no neat mapping. Indeed, the contributors have helped us as editors to see just how multifaceted this topic is, and this introductory chapter has been shaped by that discovery. Individual abstracts summarizing each contribution appear prior to this chapter, and we will not attempt replication here. Instead, we limit ourselves to giving a brief overview of how they fit together in the volume’s four parts.

Part One: Historical Foundations As the philosophical silos imposed by history and geography collapse, it has become increasingly difficult to make plausible generalizations. Accordingly, whilst our earlier claim that social roles are a neglected topic is true of recent and contemporary anglophone normative ethics, critical discussion of social roles and their ethics thrives, or has thrived, in other intellectual ecosystems. The contributions to Part One make this clear by advancing contemporary debates in role ethics through four historical lenses. These are: women in Confucian ethics (Jing Iris Hu), the Homeric tradition via Alasdair MacIntyre (Sophie Grace Chappell), John Dewey’s pragmatism (John Simons), and the ethics of the Danish philosopher K. E. Løgstrup (Robert Stern).

Part Two: The Nature and Normativity of Roles The second part of the volume comprises work on roles in metaethics, social ontology, and social and political philosophy. Its diverse contributions have in

18

  

common their asking questions about the power of social roles, by which we mean the way roles appear both to exert normative force on their occupants and to serve as a potential source of political power. In the former sense of normative force, Erin Taylor considers, and answers, the question of whether and how the conventional obligations of social roles are morally binding on occupants, whilst Reid Blackman develops an account of a class of normative reasons grounded distinctly in social roles. In the latter sense of potential political power, Tracy Isaacs argues that seeing membership of certain social groups as roles can help explain the obligations those role occupants have to resist and ameliorate oppressive social structures of which their groups are part.

Part Three: Roles, Institutions, and Organizations The topic of this section, broadly speaking, is the social constitution of roles. Three of the chapters focus on what we might call paradigm social roles—those circumscribed by social institutions (Sean Cordell), organizations (Stephanie Collins), and employers (Thomas H. Smith)—and on what requirements such roles generate for their occupants. Each of these three pieces tackles that question by starting with the thought that, in one way or another, roles or their host social structures have certain functions or purposes. Diane Jeske, on the other hand, uses such socially determined or ‘institutional’ social roles as a foil to argue—and not merely stipulate—that ‘friend’ is like a bat among birds, so to speak. Despite its similarity to certain social roles involving intimate relationships, friendship is in important respects so much unlike those paradigm cases that we ought not to think of ‘friend’ as a social role at all.

Part Four: Wellbeing, Selfhood, and Roles The final part of the volume moves into the realm of moral psychology, considering roles’ influence on or constituting of their human subjects. The mood of this section is upbeat, focusing as it does on benefits and goods which roles afford their occupants. By looking at autobiographical exposition, Samuel Clark gives an analysis of different ways in which some roles can be, and are, basic ingredients in good human lives. Alex Barber discusses the wellbeing of individuals which obtains only through their being part of a social collective. On his account, recognizing that we occupy certain roles within such collectives is key to understanding such ‘participatory’ wellbeing. Luke Brunning takes seriously, but rejects, the common suspicion that one person’s occupying multiple contrasting roles is inimical to their integrity or good. On the contrary, he argues, managing a diversity of roles in the right ways can, for several reasons, be a good thing for a person to do.

       

19

References Alexandra, Andrew, and Miller, Seumas. 2009. ‘Ethical Theory, “Common Morality,” and Professional Obligations’. Theoretical Medicine and Bioethics 30: 69–80. Andre, Judith. 1991. ‘Role Morality as a Complex Instance of Ordinary Morality’. American Philosophical Quarterly 28: 73–79. Applbaum, Arthur Isak. 1999. Ethics for Adversaries: The Morality of Roles in Public and Professional Life. Princeton: Princeton University Press. Archard, David. 2013. ‘Dirty Hands and the Complicity of the Democratic Public’. Ethical Theory and Moral Practice 16 (4): 777–790. Beauvoir, Simone de. 1973 [1949]. The Second Sex. New York: Basic Books. Bellamy, Richard. 2010. ‘Dirty Hands and Clean Gloves: Liberal Ideals and Real Politics’. European Journal of Political Theory 9 (4): 412–430. Bradley, F. H. 1927 [1876]. ‘My Station and its Duties’. In Ethical Studies: Selected Essays, pp. 160–213. Oxford: Oxford University Press. Brunning, Luke. 2018. ‘Cultivating an Integrated Self ’. In M. Dennis and S. Werkhoven (eds.), Ethics and Self Cultivation: Historical Perspectives, pp. 174–196. Abingdon: Routledge. Coady, C. A. J. 2009. Messy Morality: The Challenge of Politics. Oxford: Oxford University Press. Cohen, Gerald A. 1966. ‘Beliefs and Roles’. Proceedings of the Aristotelian Society 67: 17–34. Collins, Stephanie. 2016. ‘Distributing States’ Duties’. Journal of Political Philosophy 24: 344–366. Collins, Stephanie. 2017. ‘Duties of Group Agents and Group Members’. Journal of Social Philosophy 48: 38–57. Cordell, Sean. 2011. ‘Virtuous Persons and Social Roles’. Journal of Social Philosophy 42: 254–272. Cribb, Alan. 2011. ‘Integrity at Work: Managing Routine Moral Stress in Professional Roles’. Nursing Philosophy 12: 119–127. Dahrendorf, Ralf. 1968. ‘Homo Sociologicus: On the History, Significance and Limits of the Category of Social Role’. In Essays in the Theory of Society, pp. 19–87. Stanford, Calif.: Stanford University Press. Dare, Tim. 2009. The Counsel of Rogues: A Defence of the Standard Conception of the Lawyer’s Role. Farnham: Ashgate. Dare, Tim and Swanton, Christine (eds.). 2020. Perspectives in Role Ethics: Virtues, Reasons, and Obligation. Abingdon: Routledge. Downie, R. S. 1971. Roles and Values. London: Methuen. Emmet, Dorothy. 1966. Rules, Roles and Relations. London: MacMillan. Epstein, Brian. 2018. ‘Social Ontology’. The Stanford Encyclopedia of Philosophy. Edward N. Zalta (ed.) .

20

  

Fanon, Frantz. 2008 [1952]. Black Skin, White Masks. Richard Philcox (trans.). New York: Grove Press. Firestone, Shulamith. 1970. The Dialectic of Sex: The Case for Feminist Revolution. New York: Bantam Books. Freedman, Benjamin. 1978. ‘A Meta Ethics for Professional Morality’. Ethics 89: 1–19. French, Peter. 1984. Collective and Corporate Responsibility. New York: Columbia University Press. Gilbert, Margaret. 1996. Living Together: Rationality, Sociality, and Obligation. Lanham, Md.: Rowman and Littlefield. Goffman, E. 1990 [1959]. The Presentation of Self in Everyday Life. Harmondsworth: Penguin. Hamilton, Richard Paul. 2016. ‘My Role and Its Virtues’. Journal of Value Inquiry 50: 683–685. Harcup, Tony. 2007. The Ethical Journalist. London: Sage. Hardimon, Michael O. 1994. ‘Role Obligations’. Journal of Philosophy 9: 333–363. Haslanger, Sally. 2012. Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press. Hess, Kendy M., Igneski, Violetta, and Isaacs, Tracy (eds.). 2018. Collectivity: Ontology, Ethics, and Social Justice. Rowman and Littlefield. Hilbert, Richard A. 1981. ‘Toward An Improved Understanding of “Role” ’. Theory and Society 10 (2): 207–226. Hursthouse, Rosalind. 2007. ‘Two Ways of Doing the Right Thing’. In Colin Farrelly and Lawrence B. Solum (eds.), Virtue Jurisprudence, pp. 236–255. Houndmills: Palgrave MacMillan. Isaacs, Tracy. 2011. Moral Responsibility in Collective Contexts. Oxford: Oxford University Press. Isaacs, Tracy. This volume. ‘The Part We Play: Social Group Membership as a Role’. Jacobs, Struan. 2018. ‘How Role Replaced Personality as a Major Category of Sociology’. American Sociologist 49: 280–298. Jeske, Diane. 1998. ‘Families, Friends, and Special Obligations’. Canadian Journal of Philosophy 28 (4): 527–555. Jeske, Diane. 2001. ‘Special Relationships and the Problem of Political Obligations’. Social Theory and Practice 27: 19–40. Jeske, Diane. 2019. ‘Special Obligations’. The Stanford Encyclopedia of Philosophy. Edward N. Zalta (ed.) Jeske, Diane. This volume. ‘Are Obligations of Friendship Role Obligations?’ Kane, Laura Wildemann. 2019. ‘What is a Family? Considerations on Purpose, Biology, and Sociality’. Public Affairs Quarterly 1: 65–88. Keller, Simon. 2013. Partiality. Princeton: Princeton University Press. Kisolo-Ssonko, Joseph. 2019. ‘Race and the Responsibility to Abide by the Norms of Unchosen and Unjust Social Roles’. The Monist 102: 172–186.

       

21

Lawford-Smith, Holly. 2012. ‘The Feasibility of Collectives’ Actions’. Australasian Journal of Philosophy 90: 453–467. Linton, Ralph. 1936. The Study of Man. New York: Appleton-Central. List, Christian and Pettit, Philip. 2011. Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press. Luban, David. 1988. Lawyers and Justice. Princeton: Princeton University Press. Ludwig, Kirk. 2017. From Plural to Institutional Agency: Collective Action II. Oxford: Oxford University Press. Machiavelli, Niccolo. 1999 [1532]. The Prince. London: Penguin Books. Merton, Robert K. 1968. Social Theory and Social Structure. New York: Free Press. Miller, Seumas. 2010. The Moral Foundations of Social Institutions: A Philosophical Study. New York: Cambridge University Press. Miller, Seumas and Mäkelä, Pekka. 2005. ‘The Collectivist Approach to Moral Responsibility’. Metaphilosophy 36: 634–651. Mills, Charles W. 1997. The Racial Contract. Ithaca, NY: Cornell University Press. Moller Okin, Susan. 1989. Justice, Gender, and the Family. New York: Basic Books. Murdoch, Iris. 1992. Metaphysics as a Guide to Morals. London: Vintage Books. Nichols, Steven P. 1997. ‘Professional Responsibility: The Role of the Engineer in Society’. Science and Engineering Ethics 3: 327–337. Oakley, Justin, and Cocking, Dean. 2001. Virtue Ethics and Professional Roles. New York: Cambridge University Press. Parsons, Talcott. 1937. The Structure of Social Action. New York: McGraw-Hill. Paul, Laurie. 2015. ‘What You Can’t Expect When You’re Expecting’. Res Philosophica 92: 149–170. Pettigrove, Glen. 2020. ‘Characters and Roles’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons and Obligations, pp. 11–30. New York: Routledge. Porter, Lindsey. 2014. ‘Why and How to Prefer a Causal Account of Parenthood’. Journal of Social Philosophy 45 (2): 182–202. Rawls, John. 2000 [1971]. A Theory of Justice. Oxford: Oxford University Press. Ritchie, Katherine. 2018. ‘Social Creationism and Social Groups’. In Kendy Hess, Violet Igneski, and Tracy Isaacs (eds.), Collectivity: Ontology, Ethics and Social Justice, pp. 13–34. Lanham, Md.: Rowman and Littlefield. Rovane, Carol. 1998. Bounds of Agency. Princeton: Princeton University Press. Rovane, Carol. 2004. ‘What is an Agent?’ Synthese 140: 181–198. Sandin, Per. 2007. ‘Collective Military Virtues’. Journal of Military Ethics 6: 303–314. Sartre, Jean-Paul. 2018. Being and Nothingness. Trans. S. Richmond. London and New York: Routledge. Scheffler, Samuel. 1997. ‘Relationships and Responsibilities’. Philosophy and Public Affairs 26: 189–209.

22

  

Schmitt, Fredrick F. (ed.). 2004. Socializing Metaphysics. New York: Rowman and Littlefield. Sciaraffa, Stefan. 2011. ‘Identification, Meaning, and the Normativity of Social Roles’. European Journal of Philosophy 19: 107–128. Sheehy, Paul. 2007. ‘Blaming Them’. Journal of Social Philosophy 36: 428–441. Simmons, Andrew John. 1996. ‘External Justifications and Institutional Roles’. Journal of Philosophy 93 (1): 28–36. Smith, Thomas H. This volume. ‘My Job and Its Requirements’. Solomon, Robert C. 1997. ‘Corporate Roles, Personal Virtues: An Aristotelian Approach to Business Ethics’. In Daniel Statman (ed.), Virtue Ethics: A Critical Reader, pp. 205–226. Edinburgh: Edinburgh University Press. Solum, Lawrence B. 2003. ‘Virtue Jurisprudence’. Metaphilosophy 34: 178–213. Stern, Robert. 2013. ‘ “My Station and its Duties”: Social Role Accounts of Obligation in Green and Bradley’. In K. Ameriks (ed.), The Impact of Idealism, vol. 1: Philosophy and Natural Sciences, pp. 299–322. Cambridge: Cambridge University Press. Sternberg, Elaine. 2000. Just Business: Business Ethics in Action. Oxford: Oxford University Press. Stocker, Michael. 1990. Plural and Conflicting Values. Oxford: Oxford University Press. Swanton, Christine. 2007. ‘Virtue Ethics, Role Ethics, and Business Ethics’. In Rebecca L. Walker and Philip J. Ivanhoe (eds.), Working Virtue, pp. 207–224. Oxford: Oxford University Press. Swanton, Christine. 2016. ‘A Virtue Ethical Theory of Role Ethics’. Journal of Value Inquiry 50: 683–819. Taylor, Erin. This volume. ‘All Together Now: When Is a Role Obligation Morally Binding?’ Thomas, Laurence. 1987. ‘Friendship’. Synthese 72: 217–236. Thomasson, Amie. 2019. ‘The Ontology of Social Groups’. Synthese 196: 4829–4845. Tillyris, Dimitri. 2015.‘ “Learning How not to be Good”: Machiavelli and the Standard Dirty Hands Thesis’. Ethical Theory and Moral Practice 18 (1): 61–74. Tollefsen, Deborah. 2015. Groups as Agents. Cambridge: Polity. Van Hooft, Stan. 1999. ‘Acting from the Virtue of Caring in Nursing’. Nursing Ethics 6: 189–201. Wijze, Stephen de. 2007. ‘Dirty Hands: Doing Wrong to do Right’. In Igor Primoratz (ed.), Politics and Morality, pp. 3–19. New York, Palgrave Macmillan. Walzer, Michael. 1973. ‘Political Action: The Problem of Dirty Hands’. Philosophy and Public Affairs 2: 160–180. Walzer, Michael. 2006 [1977]. Just and Unjust Wars: A Moral Argument with Historical Illustrations. New York: Basic Books. Zheng, Robin. 2018. ‘What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice’. Ethical Theory and Moral Practice 21: 869–885.

PART ONE

HI STO R I CAL F OUNDA TIONS

2 Tempted Like Achilles Reflections on Roles and Role-Recalcitrance in Ancient and Modern Ethics Sophie Grace Chappell

1. The most obvious way for a role to prompt intentional action is by way of something like the following simple argument-schema: (P1) If you occupy role R, then you have reason to do action φ. (P2) You occupy R. _______________ (C)

You have reason to φ.

Anyone who rejects an argument of this form is what I shall call role-recalcitrant. Role-recalcitrance is what I want to talk about in this chapter. What I want to say about it is (1) that role-recalcitrance is a human universal; (2) that at least some role-recalcitrance is ethically interesting; and (3) that at least some ethically interesting role-recalcitrance is a very good thing. I won’t say much here, by the way, about what roles are. As we know from Nietzsche, Genealogy Treatise 2, most philosophically interesting concepts outside formal logic have no useful necessary-and-sufficient-conditions definition. That is not only true of roles, but particularly true of roles: roles are part of the ontology of our social world, and our social world is historically situated and always changing. What is useful is to point to paradigm cases of roles (as I shall start doing shortly), begin with those, and move outwards from them towards greyer areas and more disputable instances. But one general point that is certainly worth noting—if no more than noting— at the outset is the Gricean social reflexivity of typical roles. I mean by this that it is part of my having any typical role R that I know I have R, that I am known by my social peers to have R, that I know I am so known, and so on. This social reflexivity is a key factor in generating the nexuses of expectation that surround typical roles.

Sophie Grace Chappell, Tempted Like Achilles: Reflections on Roles and Role-Recalcitrance in Ancient and Modern Ethics In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Sophie Grace Chappell 2023. DOI: 10.1093/oso/9780192843562.003.0002

26



Indeed, this reflexivity at least partly constitutes most of our roles. It is as a rule impossible to occupy a social role unless you are recognized to occupy it. One quick way for others to destroy some individual’s standing as an occupant of some role, or indeed, in more drastic cases, the role itself, is for them simply to withdraw this recognition. It can come about that others no longer recognize me in the role of Wise Woman, or Priest, or Paterfamilias, or King (“Citoyen Capet”), or the like. Sometimes this comes about because my individual standing fails, perhaps because of blots on my reputation; sometimes it happens because they no longer recognize the role itself. Think of a shaman in 1840s Arizona, in what was till now a Navajo town. As the population of the pueblo changes with the influx of European settlers who do not share the shaman’s world view, there is a sense in which he increasingly becomes an ex-shaman, simply because the role of shaman no longer secures social uptake: he cannot sustain the role of shaman as a social reality all on his own and just by choosing to. Change in the converse direction is possible too, of course: together, the shipwrecked schoolboys in The Lord of the Flies quickly and spontaneously create the social roles of Conch-Holder and Tribe-Leader. But, together: both the creation and the decay or destruction of roles depend on collective action and understanding. That neither can happen solely by individual choice is not just a truth, but a conceptual truth. Another general point is that complex human societies typically use a wide variety of types of roles, and that this variety too is at least partly constituted by social reflexivity. For something to be, for example, a legal role that I occupy (“Solicitor”, “Defendant”, “Judge”, “Police Officer”), I must occupy it explicitly understanding it to be a legal role, and being explicitly understood so to occupy it, and so to understand it—and so on. Here too the behaviour of others is crucial for the possibility that such types of role can so much as exist. Some roles are legal, and others political, and others again academic or police or medical, because we as a society represent these distinctions of role-type as real. Here too no individual can make or unmake these social realities just by choosing to.¹ With these preliminary remarks in place, I turn to thesis (2): role-recalcitrance can sometimes be ethically interesting. And, of course, sometimes not; not at least in the sense of “ethically interesting” that I have in mind. For a start, (P1) and/or (P2) might just be obviously false. If someone tries to use an argument of that form to force or urge me to fulfil the duties of the city coroner, then I will just point out against their (P2) that I am not the city coroner. If someone tries to tell me that as a university teacher I must ring every one of my students every single night to check how they’re doing, I shall simply deny this instance of (P1): I am a university teacher, and being a university teacher gives me all sorts of reasons, but

¹ Thanks here to Rae Langton and Neil Dewar. One classic study of this sort of social reflexivity—to which it was After Virtue that first drew my attention—is Erving Goffman’s (1959) The Presentation of Self in Everyday Life.

  

27

not this one. But there won’t be much to interest any philosopher in either of these little exchanges. Again, even when (P1) and (P2) are not obviously false like this, ignorance of them, or a failure to infer (C) from them, can be caused by mere irrationality or dimness; and neither is interesting, not at least to me here. Nor is akrasia. Failing to act on an argument of this form, despite seeing its cogency, is certainly one way of rejecting such an argument. But not one that I will talk about much in this chapter. Role-recalcitrance can be trivial, as when I indisputably have a role that indisputably involves a task, such as potato-peeling or exam-script marking, and I just don’t want to do that task. Role-recalcitrance can be at once both trivial and comical, as with Leporello’s E non voglio più servir, Alfred Doolittle’s Wiv a little bit o’ luck, Kipling’s commissariat camels, Homer’s Thersites (Iliad 2.211–277), or Xanthias the whiny slave in the opening scene of Aristophanes’ Frogs, who complains about the heavy load that he is, with a Pythonesque absurdity, carrying himself even though he is also riding a donkey.² Role-recalcitrance can be comical and serious at once: the biblical tale of the prophet Jonah is a clear example of this.³ And role-recalcitrance can be, all at once, trivial, comical, and arguably a very bad thing. So, for example, with Falstaff ’s dismissal of honour in Henry IV Part I (Act 5, Scene 1), in a scene in which Falstaff ’s concern is simply to avoid the dangerous parts of his own role as a soldier. Falstaff ’s demeanour is certainly funny, but he is also gross and contemptible: before the end of the play, he has mutilated Hotspur’s corpse in an attempt to filch from Prince Harry the honour of killing Hotspur. In what kind of cases, then, might we vindicate my thesis (2) by showing that role-recalcitrance can be significant not trivial, serious not (merely) comical, and ethically interesting? To answer this also gives me an opportunity to address my thesis (1), that role-recalcitrance is a human universal. In Section 2, I shall address both theses at once, by examining some famous remarks of Alasdair MacIntyre’s, in After Virtue (1981), about “heroic societies”.

² Another example of funny role-recalcitrance, from the beginning of the classic travel-book A Short Walk in the Hindu Kush, where the diplomat Hugh Carless cajoles Eric Newby away from his West End job in haute couture to go and climb a mountain in Afghanistan (Newby 1972, 17). “I saw the directors this morning.” “Oh, what did they say?” “That they were keeping me on for the time being but that they make no promises for the future.” “What did you say?” “That I had just had a book accepted for publication and that I was staying on for the time being but I make no promises for the future.” ³ As Matthew Kramer reminds me, Jonah is not the only role-recalcitrant prophet in the Hebrew Bible. It is certainly in Jonah’s case that recalcitrance is most clearly played for laughs; but Hosea, Ezekiel, Jeremiah, Isaiah, and Moses himself are all in one way or another disinclined to take up their prophetic roles. I take it the word “recalcitrance” itself has biblical roots: see Acts 26.14, σκληρόν σοι πρὸς κέντρα λακτίζειν.

28



MacIntyre’s central example of a “heroic society” is the society pictured in the Iliad. (At times he also discusses the worlds of the Norse sagas and of ancient Irish poetry; I know next to nothing about either, so I won’t comment on those examples here.) I agree with MacIntyre that Homeric society⁴ is about as different from our society as any human society could possibly be. My strategy of response to his remarks is therefore a kind of a fortiori. If we can find significant, serious, and ethically interesting role-recalcitrance that far away from us, we can expect to find it pretty well anywhere; and we can. And all of this corroborates my thesis (3), which will come directly into consideration towards the end of the chapter: not only is there role-recalcitrance, it’s a very good thing that there is role-recalcitrance.

2. MacIntyre says this (1981, 126): There is the sharpest of contrasts between the emotivist self of modernity and the self of the heroic age. The self of the heroic age lacks precisely that characteristic which we have already seen that some modern moral philosophers take to be an essential characteristic of selfhood: the capacity to detach oneself from any particular standpoint or point of view, to step backwards . . . and view and judge that standpoint or point of view from the outside. In heroic society there is no “outside” except that of the stranger. A man who tried to withdraw himself from his given position in heroic society would be engaged in the enterprise of trying to make himself disappear.⁵

And a little later (1981, 127, 129): [T]he freedom of choice of values on which modernity prides itself . . . would from the standpoint of a tradition ultimately rooted in heroic societies appear more like the freedom of ghosts—of those whose human substance approached vanishing point—than of men . . . what Homer and the [Norse] sagas show are

⁴ As Alex Barber rightly points out, the phrase is ambiguous. Wherever I use it in this chapter, I primarily mean “the society depicted in Homer’s works”, but of course the depiction has much to tell us about Homeric society in the other sense, i.e. “the society in which Homer lived”. ⁵ One is reminded here more than a little—and no doubt not accidentally—of Sartre’s waiter. What is the Sartrean self once all its roles and masks and mauvaise foi are stripped away? Sartre’s own answer is: un néant, a bare point of will or choice in a hostile world; a “punctual self ” as Charles Taylor (1989, 171) calls it. But, we might say, the difference between Sartre’s waiter and Achilles is that Sartre reifies this nothingness: he takes it to be a néant that appears in the world, not disappears. (Thanks here to Zachary Richards.)

  

29

forms of assertion proper to and required by a certain role. The self becomes what it is in heroic societies only through its role; it is a social creation, not an individual one.

To respond to MacIntyre’s remarks as quoted here, it is natural to begin, as he does, with a notion that is not quite the same as role-recalcitrance; but it is crucially connected to role-recalcitrance, and we shall see the connection more clearly before the end of this section. This is the notion of detachment, distancing, “standing outside”. What kind(s) of detachment can we find in the Iliad? What kinds of “outside”, if any, are available to the society pictured in the Iliad?⁶ I note first that one kind of “outside” that MacIntyre sees to Homeric society is not in fact obviously there at all. “In heroic society there is no ‘outside’ except that of the stranger”: if this means that Troy is the “outside” for the Greeks, and Greece for the Trojans, then it seems mistaken. When they are together—except on the battlefield—Homer’s Greeks and Trojans form a single society, not two. No poet is more scrupulously cosmopolitan than the poet of the Iliad. Beyond, perhaps, the noises they make as they go into battle (Iliad 3.1–9), neither Greeks nor Trojans (nor the Trojans’ allies) have any distinctively national characteristics whatever. The luxury of individual Trojans, especially Paris, is precisely that—their own individual luxury, not generic and culturally stereotyping “oriental luxury” like Cleopatra’s in Shakespeare. Paris’s luxury expresses Paris’s individual weaknesses, not his society’s (Hector does not share his weakness, and Priam has a different weakness—he indulges Paris). Again, Homer makes no attempt to reflect the language-barrier that must presumably have existed between the historical Trojans and Greeks. He does sometimes seem to suggest that Trojan names or terminologies are exotic, but this is done mainly, I think, to sharpen our sense of place, and in any case many of the Trojans have Greek names (Hector, Astyanax, Andromache); nor does Homer go in for anything like the embarrassing “funny accents” that modern war-epics so often use to other the other. As their interactions show so clearly, it is not in the least important, except at the most basic level of plot-construction, that Helen and Achilles are Greek, or Paris and Hector Trojan. What all four alike have is, simply, human characteristics. I share Simone ⁶ We might ask these same questions about two other great ethical thinkers from much later in the Greek tradition: Plato and Aristotle. To pursue either would be a whole other essay. But very briefly, it is a striking fact about both Plato and Aristotle that, in their different ways, both start from a notion of role-determinacy and end up with a notion of role-transcendence. The role-determinacy is more obvious in Plato’s insistence, in the Republic, on his famous definition of justice as ta hautou prattein—“to fulfil one’s role”. The role-transcendence is more obvious in Aristotle’s development of the theme of theôria, which is, as he explicitly tells us, precisely the ability to go beyond the proper role of a human being—and so presumably, at least on a naive reading of the Republic, a case of Platonic injustice. Yet Aristotle too begins with role-determinacy: that of course is the point of his famous ergon argument, and also of the opening hierarchy of tele and pros ta tele at the very start of the Nicomachean Ethics. And Plato too ends up with role-transcendence: that is where the dialectic gets his philosophers to, at least for a while, in the Republic; and in the Theaetetus as well.

30



Weil’s view that this quiet background cosmopolitanism is crucial for the poetic effect of the Iliad; and that it puts Homer in the most striking possible contrast with the court-poet imperialism of Virgil. There is however another way in which the Iliad very strikingly manifests “the capacity to detach oneself from any particular standpoint” and “judge that standpoint from the outside”. For we might say that the Iliad itself provides a kind of “outside” to that society about which it is a fictional narrative. The characters “in the Iliad” are indeed in the Iliad. Homer whose poem is about them, and the audiences who hear of them from Homer, are not in the Iliad but outside it, “detached from any particular standpoint or point of view” within the poem, and viewing and judging any such “standpoint or point of view from the outside”. So— we might say—even if the capacity for detachment that MacIntyre takes to be the hallmark of the modern self were not displayed in the Iliad, still it is displayed by the Iliad. Even if it were not. But in fact the capacity for detachment is also displayed within the poem. And by “displayed”, I don’t just mean “discernible if you look carefully”. I mean deliberately put on prominent show, in some of the most memorable lines in the entire work. One example is the opening of Iliad 13 (1–9), where the Greeks are facing imminent disaster: So Zeus brought Hector’s army among the ships. Then left the humans to their struggle and woe, their unending war. His own bright eyes he turned away, far off, to the horse-plains of distant Thrace; to the hand-to-hand fighters dwelling in Mysia; to the Hippemolgi, who live on horses’ milk; to Abia’s race, the justest people on earth. To Troy his bright eyes turned, now, not at all; for Zeus was sure that none of the immortals would come to help the Trojans; nor the Greeks.⁷

So far from any human standpoints, viewpoints, and roles being inescapable in the Iliad, so far from any alternative to them being unthinkable: every human viewpoint represented within the Iliad is constantly ironized and undercut by the ever-present possibility of a radically other alternative viewpoint: a viewpoint that is, quite literally, Olympian. αἰὼν παῖς ἐστι παίζων, πεσσεύων· παιδὸς ἡ βασιληίη:⁸ Zeus and the other gods can look at and involve themselves with

⁷ All translations in this chapter are my own. ⁸ Heracleitus, Diels & Kranz Fragment 27 B52: “Eternity is a child playing childishly, playing chess with itself; a child’s is the kingdom.” Perhaps Heracleitus had in mind just this Homeric passage, and the general turbid ebb and flow of the Trojan war in the poem.

  

31

whatever parts of the Trojan War they like, for as long—or as short—as they like. So here Zeus, having just (at the end of Book 12) played a central causal role in shifting the balance of the fighting against the Greeks, now simply turns his back on the continuing war, because he thinks that his present intervention settles at least the near future, and because he’d rather look at and think about something else for a bit. (There is at least a hint that the deity turns away out of mere idle boredom.) This is detachment, not human but superhuman detachment, within the Iliad. It is also, pretty clearly, the detachment of the Iliad. For Homer’s gods’ position, relative to his text, is a deliberately ambiguous one. They are both part of the action, making moves (sometimes, as here, important ones) within the war, and also spectators of the action, watching the war (often rather idly) from their eternally safe distance above and beyond human life. Because in this way the gods stand both inside and outside the narrative of the Iliad, we may say, pace Derrida, that in Homer at least there is an hors-texte; it is Olympus. In that respect, Homer flatteringly invites us to think, the gods are like his audience. The Iliad’s hearers too have this Zeus-like freedom. They too can turn away from the recitation whenever they want, command the bard to give them a comfort break, or dismiss him for the evening. Or if the recitation continues, they can if they choose be as sublimely indifferent to the continuing torment of war as Zeus is in these lines. MacIntyre in After Virtue rightly stresses the centrality of narratives for ancient Greek ethics (and for us too). But to stress the importance of narratives is eo ipso to bring out the possibility of stepping outside those narratives. As these lines show, this possibility is nothing foreign to Homer. On the contrary, that such steps away from any narrative are possible (both for Homer’s gods and for his audience⁹) is the point of these very lines.

3. Perhaps MacIntyre would admit this point. He certainly accepts (and I agree, of course) that there are some things that “the poet of the Iliad sees and his characters do not”, that Homer’s “knowledge is . . . at a more general and abstract level than that even of his most insightful characters”, that “the Iliad puts in ⁹ In personal correspondence, Nickolas Pappas raises here the interesting question whether rolerecalcitrance is possible for Homer’s gods. He thinks not, but I am less sure about that: while Homeric deities are certainly constrained quite tightly by something called Moira (fate) or Anagke (Necessity), they are sometimes at least tempted to hybris, to overstepping the mark; as e.g. when Zeus upbraids them in the famous “golden chain” speech (Iliad 8.1–38). And hybris, outrage, is clearly one form of role-recalcitrance—a very important one for the Greeks. Notice too how often Zeus himself is upbraided by other gods, particularly Hera and Athene, for wanting to refuse to go along with the demands of fate. Apparently even Zeus can be at least tempted by recalcitrance.

32



question what neither Achilles nor Priam can put in question” (1981, 128). However, if MacIntyre is to square this admission with his thesis that characters like Achilles cannot step back from their roles, then his argument needs a stable and clear division between what Homer can see and what Homer’s characters can’t. The whole point of the claim that I have just argued for, that Homer’s gods stand “both inside and outside the narrative of the Iliad”, is that no such division can be either clear or stable. Even the gods can be obtuse and blind, even mere mortals—on one occasion even a horse—can have searing moments of insight. This fact is not unrelated to the cardinally important point about Homer’s gods, that in Judaeo-Christian terms none of them, not even Zeus, is God: they are persons or agents just like humans, only stronger, wiser, better at magic, and longer-lived. Thus it is not only with the Olympians that that division breaks down; it is not only Homer’s gods (and audience) who are capable of this kind of detachment from the fray of things. Helen is capable of it too (Iliad 6.354–358): So come and share my couch, Hector my brother, you who have borne the cost of all our wrongs since Zeus laid this bad fate on Paris and me— bitch as I am, blinded and wild as he is— that for ages to come we might be the matter of song.¹⁰

Helen evinces exactly what MacIntyre says there cannot be in a character in Homer: an entirely articulate sense of how to judge her own position detachedly, “from the outside”; a bleakly lucid awareness of how her actions look (will look) sub specie aeterni; how they must look to the gods—and, again, to Homer’s audience. Here too Homer identifies “the aspect of eternity” with the aspect of his own epic. But this time, the speaker who expresses his identification is halfgoddess, half-swan: Helen, the daughter of Zeus and Leda.¹¹ A third character who is also capable of something like this detachment is not even partly divine; he is wholly mortal. This is Hector, whose sentiments at 7.87–91 are strikingly similar to Helen’s as just quoted: One day some man, of an age that’s yet to be, will say, as his ship sails past across the dark sea, “Here is a tomb from a dim posterity: a man whom Hector slew, though young and strong.” So will my deeds endure as the matter of song.

¹⁰ I also quoted this passage—for different reasons—in Chappell 2011. ¹¹ Perhaps it isn’t strictly relevant here; but it is always worth remembering that the other hatchling from the same egg as Helen was—Clytaemnestra.

  

33

The ability to detach, to step outside their particular perspective and see things “from the outside”, from an Olympian viewpoint, or even an authorial one: in Homer this ability is not the exclusive preserve of the poet or his audience, nor of the gods, nor even of semi-divine figures like Helen. It is exemplified even by thoroughly mortal characters like Hector. This capacity for detachment is exemplified too by Achilles. And his case is particularly interesting, because with him the link that I promised earlier, back from the general notion of detachment to the specific notion of role-recalcitrance, is particularly clear. One obvious way in which Achilles exhibits a limited kind of role-recalcitrance is, of course, that from Iliad Book 1 to Iliad Book 19 he is “on strike”: in Book 1 he retreats to his tents in protest at Agamemnon’s insults, and it is only Patroclus’ death in Book 18 that brings him back.¹² There are interesting role-related questions, too, about what Patroclus does or tries to do in Achilles’ absence: we might say that role-recalcitrance—doing less than you should relative to a role— can be contrasted with role-presumption—doing or trying to do more than you should; and perhaps that is what goes wrong for Patroclus.¹³ But it is not just that Achilles “goes on strike”; he also considers a much more complete and definitive form of role-recalcitrance. Here is Achilles on whether to fight or go home at 9.410–416: My mother, silver-footed goddess Thetis, foretells that I confront now double fates, a choice of roads before me to life’s end. The one road is: remain and fight at Troy, never get home, but win eternal fame. By the other road, by return to my dear Phthia, “Your glory is extinguished, but Achilles, you’ll live long years at home, safe and obscure, evade the doom of death, and grey endure.”

The fate that is really in store for Achilles is only fulfilled beyond the Iliad’s last episode, Hector’s funeral. But it is a presupposition of Homer’s art that every reader of his poem knows perfectly well—as do the gods, and as does Achilles ¹² As Iakovos Vasiliou has helpfully suggested to me, one way to read Achilles’ long “strike” is not as limited role-recalcitrance, but as a deliberate enactment of one part of the hero’s role, namely the protection of one’s own honour. But as Iakovos also points out, it might also be said that Achilles is subject to what Aristotle scholars sometimes call “rapid akrasia”: in fact the Iliad is a span from Achilles being overcome, in Book 1, by rage that nearly leads him to kill Agamemnon, to his being overcome, in Book 24, by rage that leads him into repetitively and compulsively abusing Hector’s corpse; and in both cases a god’s intervention is needed to get him “unstuck” from a kind of compulsiveness. ¹³ My thanks to Jessie Munton for pointing this out. (Maybe Aphrodite is guilty of a kind of rolepresumption too, when she tries to join the battle in Iliad 5 and is wounded by Diomedes.)

34



himself—what his fate will be, and that it is, quite determinately, not long-term survival in happy obscurity at home. Achilles’ true fate is stated plainly to him by Thetis at Iliad 18.95–96: the fall of Troy is fated to follow swiftly on Hector’s death, and Achilles’ death to come when Troy’s fall comes. So whatever Thetis may have said¹⁴ as reported in Book 9, Achilles above all knows (and knows exactly) what must be, and what is demanded of him by his fate, by his cosmic and political situation, and by his role as the aristos—the swiftest, strongest, and most proficiently ruthless—of all the Greek warlords. Yet here, at his mother’s prompting, Achilles toys with the alternative. He plays with the possibility of stepping outside his role altogether—not just brooding in his tents on the Aegean shoreline, but leaving the war completely and going home. To take the alternative of nostos (going home) would be Achilles choosing, not just a different way of playing the role of Homeric hero, but a path that is genuinely different from any heroic role. If he took that path he would certainly, in one sense of MacIntyre’s word, “disappear”. He would disappear in just the sense that Odysseus does in fact disappear at the end of the Odyssey, when Athene appears as a dea ex machina to command the suspension of his last heroic war. Athene in effect tells Odysseus that it’s time to retire; and from elsewhere in the Odyssey we have a perfectly clear picture of what coherent agency looks like for someone like him when he is not playing a hero’s role—compare, in particular, the depictions of Odysseus’ son and father, Telemachus and Laertes. So in the Iliad, what really threatens to make Achilles ghostlike and insubstantial is not abandoning the hero’s role, but rather continuing in it and staying in the war. This threatens to make Achilles literally ghostlike and insubstantial—as indeed he is, by the time of Odyssey Book 11 (and as Odysseus is, after his final coming-out-ofretirement adventure, in his marvellous narrative in Inferno Canto 26). This is not the only time in the Iliad that Achilles is seriously tempted by the alternative of detaching himself from his role as war-hero and just going home: compare 1.169–171. If MacIntyre were right in the quotations that I began Section 2 with, it is not clear how Achilles could even be tempted by it. What MacIntyre calls “detachment”, the “step backwards”, and judgement of a “standpoint or point of view from the outside” are all possible in Homer. They are possible not just for Homer himself and his audiences and/or gods, “outside” the poem or its main action, but for Homer’s characters too, “inside” the story. Because this detachment is possible, so is a philosophically interesting kind of role-recalcitrance, of which Achilles is a particularly clear example. It is possible

¹⁴ Where does Thetis make this prophecy to Achilles? Not at Iliad 1.414–427: there, she says that Achilles’ choices are to stay in his own tents at Troy or to fight with the Greek army at Troy, and that the latter choice puts him in more danger of an earlier death. This third option, to leave Troy and go home altogether, she does not there discuss.

  

35

for him to have the thought that, although he is clearly a Homeric hero, and his role as a Homeric hero demands that he stay at Troy and continue with the war, nonetheless he might choose not to: he might simply choose nostos instead. Hence the a fortiori argument that I suggested at the end of Section 1. Homeric society is about as distant from ours as any society can be, and still be recognizably human. So if we can find role-recalcitrance there, we can probably find it anywhere. There are of course many deep differences, as to what it is like to be (a) human, between the Homeric world and our world. Pace MacIntyre, the availability of (philosophically interesting) role-recalcitrance is not one of them. Homeric heroes would be humanly unintelligible to us if this were not available to them—in something rather like the probably related way in which they would be unintelligible if they were as lacking in the capacity for decision and choice as Bruno Snell took them to be.¹⁵ But they are not in the least unintelligible in respect of role-recalcitrance; role-recalcitrance is a notion that is central to the whole poem, and central characters in the poem exhibit a variety of forms of it.

4. The basic schema for role-compliance that I began with was this: (P1) If you occupy R, then you have reason to φ. (P2) You occupy R. _______________ (C)

You have reason to φ.

Philosophically interesting role-recalcitrance can happen when someone denies the truth of (P1). This gets us to some important, and very familiar, ethical disputes about exactly what is entailed by any given role. So, for example, we (and pretty well all ethical agents ever) very often ask each other questions like “What is involved in being a friend?”, “Should school-teachers beat their pupils?”, “Are wives required to submit to their husbands?”, “When should soldiers not obey their superiors’ orders?”, or “Are political leaders permitted, or even required, to hand out perks to their followers?” Questions like these are answered, when they are, by refinements of instances of (P1). The business of so improving our understanding of the roles that we occupy, and of negotiating a happy mean between a stable and a flexible conception of those roles, is a central part of our ethical life together.

¹⁵ On Snell’s arguments about choice and decision in Homer see Bernard Williams 1993, chapter 2.

36



Here, when I say “the roles that we occupy”, I actually mean “the nonpernicious roles that we occupy”: there is another class of roles that we might occupy for which what we want to say is that (P1) is false for them because they generate no reasons at all, since they are bad roles to be in in the first place. Roles do not have to be, intrinsically, positively beneficial or salutary to generate reasons; ethically neutral roles can do that too. But they do need to be not pernicious roles, as “concentration-camp guard” and “Mafioso” and “confidence trickster” are pernicious. (Or if these roles do generate any reasons, they do so against the grain, so to speak. Maybe concentration-camp guards have a rolederived reason to subvert their own role in ways that are only possible because they occupy it.¹⁶) Exactly what it is that makes any role pernicious, or neutral, or beneficial, is a large and interesting question, and our confidence that every role customarily accepted in our society is at least a non-pernicious one is likely to be proportional to our political conservatism; but I won’t try to say more about this here. Important varieties of role-recalcitrance can also result when someone denies the truth of (P2), and here too many of the most interesting resultant responses and disputes are of a very familiar form. Someone who finds themselves in some role that is itself unwanted, or that has unwanted implications, may take action to free themselves of that role, so that they make it that some instance of (P2) is untrue of them. Or someone may say “I do not truly occupy any role that I have not chosen, or that someone else has coerced me into”; though this line of argument seems to face well-known counter-examples, in particular roles like “parent”, “child”, “citizen”, “defendant”, and indeed “fellow shipwreck-survivor”. Or someone may say that “There is some true description of me that does not capture anything that I recognize as a genuine role; I may be a member of the Wingrave family (see Williams 1981), or a Lancastrian, or a Libra, or a woman, but although that is a true description of me, it is not a role-description of me: it is an ethically inert fact of genealogy/geography/genetics about me.” (As we might want to put it: these concepts lack “thickness” for me.) Or again, a role that I do indeed sometimes occupy may, for the moment, be fallow or inert, so that at present, it makes little sense to say that I occupy it. For example, I might be a qualified and professional Californian beach-lifeguard, but currently on holiday in Colorado. ¹⁶ Perhaps, as Rae Langton has suggested to me in correspondence, there can be bad role-related norms, just as there are bad laws: “There are norms for a prison camp guard etc. related to the role, and it is bad that people are bound by those roles. They really are bound, but it is local, just as a bad law is binding in a local way; even if from the wider point of view, such local bindings shouldn’t exist.” So even bad role-reasons have some traction on us, in the particular context that generates them; but how that local traction relates to what ultimately ought to sway us is no easier a question to settle than any question about how to move from the particular context to the viewpoint of universal moral truth. This is of course an enormous topic; in this footnote I can only mark its existence, and say how much I look forward to seeing Rae’s forthcoming work on the topic.

  

37

A final form of role-recalcitrance is to accept the entire argument (P1), (P2), (C) as sound, but add that it has a conclusion too weak to guarantee action. For the sake of generality, I have stated (C) so that it says only “You have reason to φ”. But of course, anyone is at liberty to agree that they have reason to do any number of things, on the grounds of any number of types of considerations including roles, without that necessarily implying any particular deed. (This is one reason why so much of the usual metaethical debate about motivational internalism can get somewhat frustrating: what if moral judgements always generate a reason, but a reason of microscopic strength?) A reason to do something can very often be real, but outweighed or trumped or silenced or defeated by other reasons, and therefore not a decisive reason. Which of these forms of role-recalcitrance does Achilles display in toying with nostos? Homer does not give us conclusive evidence about that, and of course Achilles might be tempted by more than one of these possibilities, only some of which are mutually inconsistent. But the natural assumption, I think, is that he is tempted to take action to make (P2) untrue of himself. After his devastating quarrel with Agamemnon in Book 1 of the Iliad, it is not that Achilles thinks that he still has reason to act as a hero and stay in the war, only now his reason to do this is outweighed by other reasons. It is rather that Achilles has come to think, or at least is now tempted to think, that he has no reason at all, not even an outweighed one, to carry on fighting at Troy. But plainly, the role of Homeric hero does imply that sort of reason; and plainly, up to this point, Achilles remains in that role. Achilles is not in a position to alter the facts about what Homeric heroism implies; but he is in a position to alter the facts about whether Homeric heroism applies (to him). He can get out of the role’s implications simply by resigning the role. This then is what he is tempted to do. The action he is contemplating is, pace MacIntyre, precisely one of stepping back from the key social role of Homeric society: that of the war-hero.

5. One reason to taxonomize these possible ways of defeating a role-compliance argument of the general form (P1), (P2), (C) is because MacIntyre in After Virtue seems, at times, alarmingly close to saying that arguments of something like that form are indefeasible. So, at an earlier point in that marvellous book (1981, 57),¹⁷ MacIntyre reminds us, approvingly, of an argument that A. N. Prior apparently ¹⁷ Despite the fight that I seem to pick with it in this chapter, my praise of After Virtue is absolutely heartfelt. Reading it properly for the very first time, I think on a train from Reading to Oxford in 1988 (a very slow one), was one of the highlights of my early philosophical reading. I go back to it again and again, and every time I do, I find something new and wonderful. I can only hope the present chapter does not seem too inclined to lèse-majesté.

38



sometimes gave in conversations or seminars,¹⁸ the point of which was that roleterms could get us across the Is–Ought Gap. From the purely factual premiss “X is a sea-captain” we can (Prior apparently said) infer, immediately and without further premisses, the evaluative conclusion that “X ought to do whatever a seacaptain ought to do”. If that sounds unpromisingly indeterminate we can, we might suggest,¹⁹ add some more purely factual content to the premiss—“X is a seacaptain, and a sea-captain is responsible for the maintenance of his vessel, and the maintenance of a vessel includes checking for leaks”—and from that infer, immediately and without further premisses, the evaluative conclusion that “X ought to do whatever a sea-captain ought to do: in particular, he ought to maintain his vessel; more particularly still, he ought to check for leaks”. Generalizing, MacIntyre tells us (1981, 57–58) that the sea-captain argument is an instance of a broader class of Is–Ought arguments of which he gives two more examples: From such factual premisses as “This watch is grossly inaccurate and irregular in time-keeping” and “This watch is too heavy to carry about comfortably”, the evaluative conclusion validly follows that “This is a bad watch”. From such factual premisses as “He gets a better yield for this crop per acre than any farmer in the district”, “He has the most effective programme of soil renewal yet known” . . . the evaluative conclusion validly follows that “He is a good farmer.”

Generalizing further, MacIntyre gets to this (1981, 58–59): [For Aristotelian ethicists] “man” stands to “good man” as “watch” stands to “good watch” or “farmer” to “good farmer” in the classical tradition. Aristotle takes it as a starting point for ethical inquiry that the relationship of “man” to “living well” is analogous to that of “harpist” to “playing the harp well” (NE 1095a16). But the use of “man” as a functional concept is far older than Aristotle . . . It is rooted in the forms of social life to which the theorists of the classical tradition give expression. For according to that tradition to be a man is to fill a set of roles each of which has its own point and purpose: member of a family, citizen, soldier, philosopher, servant of God. It is only when man is thought of as an individual prior to and apart from all roles that “man” ceases to be a functional concept.

¹⁸ The sea-captain argument is not in Prior’s Logic and the Basis of Ethics (1960a), though that does contain a rather similar argument about sextons; nor is it in Prior’s ‘The Autonomy of Ethics’ (1960b). Similar arguments do, however, occur in Searle 1964. I am grateful to Professor MacIntyre himself for letting me know by email that “My memory is that this example was used more than once in discussion and was always attributed to Prior.” ¹⁹ I make this suggestion more explicit than Prior does, or MacIntyre.

  

39

Now it cannot be more than a beginning to point out that none of the three arguments that MacIntyre presents here comes anywhere near being formally valid; but it is at least a beginning. In each case, the argument can only go through if we add something like a ceteris paribus or That’s-It clause. So with the watch—the most promising of the three examples—we need to add a premiss that reads, roughly, “ . . . And there are no other countervailing positive characteristics to the watch”; similarly with the farmer, if we substitute negative for positive characteristics. As for the sea-captain: pace Prior and MacIntyre, “X is a sea-captain” plainly does not entail, all on its own, that “X ought to do whatever a sea-captain ought to do”. One can be a sea-captain, yet be in a situation where “sea-captain” is simply not the description under which one ought to act. This can be either because it implies intuitively wrong directives, as with the directive “A sea-captain should always prioritize the well-being of his vessel”, which can be overridden by, say, suddenly arising humanitarian or military emergencies. (Cf. Section 4’s point about how role-compliance arguments can soundly give us reasons, but reasons that are overridden.) Or it can be because a true role-description at present implies no directives at all, as when the sea-captain retires from Falmouth to Gweek for the winter, and from November to March acts mostly under the description “farmer”, and hardly at all under “sea-captain”. (Cf. Section 4’s point about how role-compliance arguments can be defeated by rejecting either the idea that one currently occupies some role (P2), or the idea that a role that one does indeed occupy is currently reason-giving (P1).) With other roles that one might occupy, the non-sequitur is even plainer. As already noted, we might as well say that a drug-dealer “ought to do whatever a drug-dealer ought to do”; or a Mafioso. In sum, before any practical conclusions can be drawn from the fact that someone occupies some role, the role needs to be a non-pernicious one as I put it in Section 4, a good role to be in in the first place. Such functional or role-based arguments as MacIntyre’s three are not, therefore, compelling examples of logically sound or even valid arguments across the Is– Ought gap. To be blunt, they get nowhere near.

6. In their sparkling Modern Epistemology: An Introduction Nicholas Everitt and Alec Fisher (1995) say this: Consider the assertion (M)²⁰ If a woman gives birth to a child, she is its mother. ²⁰ Everitt and Fisher actually label it “(C)”. I re-label it “(M)” to avoid confusion with my own earlier “(C)”.

40

 Until recently, this might have been regarded as an a priori truth, and one which was true solely in virtue of word meanings. Motherhood, it might have been said, is defined in terms of giving birth, so we can know that the proposition is true no matter what empirical discoveries are made. But . . . we now know that the position is not so simple . . . the egg of one woman [can] be fertilised in vitro by a sperm, and then implanted in the womb of a second woman who nine months later gives birth to a child. Is she then the mother of the child? We can of course distinguish between a genetic mother and a birth mother . . . But what does this distinction do to our supposed a priori truth (M)? . . . when (M) seemed to be an a priori truth that was not because we were interpreting “mother” as “birth mother”. The concept of “birth mother” as distinct from “genetic mother” was one which had not even been thought of . . . [U]nforeseen scientific advances can radically change our acceptance of propositions which initially seemed immune to empirical findings. In each case, our willingness to regard the “a priori truth” differently is based on the fact that the new empirical information undermines some very general assumptions which lay behind the old a priori truth. Sometimes, the new information will bring in its train concepts which have no sense in the old picture . . . Sometimes the new information shows that the old “truth” was true if taken in one way but not if taken in another, where the very idea that there are two ways in which that truth might be understood becomes intelligible only in the light of the new empirical information . . . [W]e may each find within our own corpus of beliefs some beliefs which seem to us to be a priori, to be beliefs which we are justified in accepting no matter what empirical findings science may come up with. But if we do take that attitude toward our beliefs, that shows only the limits of our imagination, not that our beliefs will hold true in face of all increase in scientific knowledge. (1995, 112)

Here Everitt and Fisher are expounding, and expanding, Quine’s celebrated attack on the analytic/synthetic distinction. Their target is any supposedly a priori inference “in virtue of the meanings of words”. I hope it is obvious enough why their remarks deserve such full quotation here. Inferences like that from “X gave birth to Y” to “X is Y’s mother” do indeed depend upon “the meanings of words”. But just for that reason, they do not form a stable set of unchangeably, unrevisably logically valid inferences. As a matter of observable historical fact, the meanings of words are changeable, revisable, and in no way logically quarantined. (Not even by stipulation; stipulations too can be reinterpreted, and often are. There is, familiarly enough, no guaranteeing how a rule R1 will be interpreted that consists simply in laying down a rule R2 about how to interpret R1; meta-rules are both possible and, e.g. in law, actual, but they cannot guarantee anything.) And the reasons why the meanings of words change are similarly messy and various, and cannot be neatly

  

41

regimented under any one single category of, say, “rational” or “empirical”. The meaning of “mother” changes—more exactly, what it is possible to mean by “mother” changes—in one way when we come to know the empirical facts about surrogate births; it changes in another way when we come to accept the social possibility of adoptive parents; it changes in a third way when we come to accept the social possibility of transgender parents.²¹ And so on, without any definite or obvious pre-ordained limit; transcending this sort of limit is indeed, in Everitt and Fisher’s words above, a matter of transcending “the limits of our imagination”. (Science-fiction writers do it all the time.) So the reasons why we accept an inference for example from “X gave birth to Y” to “X is Y’s mother”, when we do, are not at bottom logical at all. They are not even semantic. They are sociological; they are about our society, and our own minds’ attunement to that society. It is not “Logic guarantees this inference”; it is “Society guarantees this inference.” And what holds for “mother” holds for “seacaptain”, or “Homeric hero”, too. What reasons, if any, are implied by such roles; whether we find the roles themselves admirable or acceptable, and/or agree that there should even be those roles in our society: this too is not a logical issue, but a sociological and psychological one. And, of course, an ethical one.²²

7. Back in the days when it was still customary for anglophone moral philosophers to discuss Is–Ought arguments at all, such discussions displayed a recurring weakness. This was the general assumption that the only philosophically interesting question about Is–Ought arguments was whether any of them were formally valid. It is hard to see what could have suborned us into thinking that, other than a

²¹ For more about these kinds of revisionary possibility see Chappell, forthcoming. ²² The link between conservatism about word-/meaning-change and what I shall politely call conservatism in ethical matters is a surprisingly pervasive one. So, e.g. N. T. Wright argues online that gay marriage is guaranteed unacceptable by the meanings of the words involved (i.e. “man”, “woman”, “marriage”).

Wright suggests that the semantic—and so it seems canon-law—bipolarity of “man” and “woman” is as divinely ordained as the bipolarity of heavens and earth in Genesis chapter 1. Wright’s argument apparently commits him to thinking that the authors of Genesis meant, by the Hebrew words shamayim and erets in 600 , exactly what we mean today by the English words “heavens” and “earth”. Given the manifest differences between our cosmology and theirs, this looks, to put it mildly, an ambitious claim. His semantic absolutism also leaves him in a somewhat awkward relation to Christianity’s large-scale appropriation and reinterpretation of the Jewish religious vocabulary (“atonement”, “forgiveness”, “righteousness”, “Passover”, “sacrificial lamb”, “unity” (of God), and above all perhaps “Messiah”). So far as I can see, Wright does not comment on these implications.

42



lecture-hall preoccupation with logico-linguistic target-practice. In fact, the recognition that Is–Ought inferences are never underwritten by logic or “the meanings of words” alone, but always also by a social context that makes them seem compelling, or compelling for us, is nothing like the end of the real philosophical interest. It is only the beginning. So with the specific case of role-based arguments like (P1), (P2), (C), the observation that a given role that I occupy compels a particular response from me, just because I occupy it, is an observation not about syntax, nor about semantics, but about Sittlichkeit. But then, we have to consider how exactly to understand such socially mediated inferences; and this is neither a small task, nor a philosophically uninteresting one. The fact that some inference from role to response strikes Achilles as compelling may not be information about the timeless structure of “the language of morals” or “the structure of practical reason”—there being, at any rate timelessly, no such thing. But it is information about Achilles’ language of morals (or at any rate of mores, êthê), and about the structure of Achilles’ practical reason. As too is the fact that Achilles is capable—at least sometimes, at least to some extent—not only of acting within his roles, but also of stepping outside them: of being role-recalcitrant. At least sometimes; at least to some extent. Though in fact, as many readers must have been itching to interject for quite some time now, Achilles was after all only tempted to role-recalcitrance. In the end, of course, Achilles doesn’t yield to this temptation. He doesn’t sail off home; and he does remain at Troy enacting the role of Homeric hero; and like Patroclus and Hector before him, he does get killed for it. In the last analysis, the reason why this happens to Achilles is presumably, as suggested in Section 3, because it is Achilles’ fate. To say this is not to say merely that “This is how Homer intends the story to go”; it is also to say that “This is how Achilles knows that his own story must go”—despite the claims he makes in 9.410–417 about his mother Thetis’ prophecy of his double fates. In Homer’s time such knowledge of one’s own fate was not restricted to fictional characters. It is clear enough that real-world people believed then, and have often believed since, that there was something called “their fate”, and that that fate would find them out, no matter what. For them, being in a narrative meant being trapped in a narrative, as in Somerset Maugham’s “Appointment in Samarra”: There was a merchant in Baghdad who sent his servant to market to buy provisions and in a little while the servant came back, white and trembling, and said, “Master, just now when I was in the marketplace I was jostled by a woman in the crowd and when I turned I saw it was Death that jostled me . . . lend me your horse, and I will ride away from this city and avoid my fate. I will go to Samarra and there Death will not find me.” The merchant lent him his horse, and

  

43

the servant mounted it, and he dug his spurs in its flanks and as fast as the horse could gallop he went. Then the merchant went down to the marketplace and he saw [Death] standing in the crowd and he came to her and said, “Why did you [jostle] my servant when you saw him this morning?” “That,” [Death] said, “was only a start of surprise. I was astonished to see him in Baghdad, for I had an appointment with him tonight in Samarra.”

Perhaps such superstitiously fatalistic outlooks are less common today than in previous eras (though as far as I can see, they are still widespread even in highly scientifically literate societies like our own). Whether or not that is so, even today there are still at least two other kinds of outlook²³ that also cut in on us to tell us, and often it seems with a pretty fatalistic intention, that we are more or less bound to head in some given direction; and at least one of them seems to have a pretty solid backing from science. These are the outlooks, as I shall call them, of biography and biology. For a modern non-superstitious mindset, biography is perhaps the nearest equivalent to what fate is for pre-modern and superstitious minds. Like fate, it is essentially a preoccupation with narrative patterns. Biography tells us, or purports to, how our lives and our choices are narratively framed by our past and the pasts of connected others. It is to biography in this sense that Owen Wingrave’s family appeal, in Henry James’s story, when they tell him that he too—like his father and grandfather—is bound, in more than one sense, to become a soldier. Similarly, in Thomas Mann’s novel, it is to biography that Thomas Buddenbrook appeals to convince his son Hanno that he too must become a Baltic merchant like all the generations before him. It is hard to see what real logical argument for any particular course of action there could be, on the basis of biographical factors of these kinds. But that, of course, is exactly the point of my saying that role-based arguments are not logically but sociologically grounded. In the last analysis, that all the Buddenbrooks for the last 400 years have been merchants is presented by his father as a good reason why Hanno too should become a merchant, not because of logic, but because there is this pattern—and he will enforce its continuation. But the idea that you have a role-based reason to act merely because of a pattern like the Buddenbrooks pattern, or the Wingrave pattern, looks like mere superstition. It would be nice to think that the other modern stand-in for fate, namely biology, is both more “grounded in the facts” and also “less superstitious” than

²³ There is also, of course, philosophical determinism in its numerous varieties. Except among professional philosophers, this view—in any variety—seems to me less common than a belief either in fate, or in biography or in biology as modern-day stand-ins for fate. In any case, I have nothing bespoke to say about determinism here.

44



biography as just described. And perhaps it sometimes is. It is, after all, undeniably a matter of science that we are human animals, and that any individual human animal typically has a familiar roster of characteristics and dispositions, including, in particular, a uterine origin, a carbon-based body which usually displays fairly clear sexual dimorphism, a particular chromosomatic makeup, and a natural lifespan of—in all known cases to date—less than 130 years. For all that, the uses of biology to ground role-based arguments can be—and notoriously often are—no less logically wayward than the uses of biography, or indeed the uses of fate. The idea, for example, that “biology itself ” implies that there is something unnatural-and-so-wrong about in vitro fertilization, or that “biology itself ” grounds arguments for specific “male” and “female” roles in life, and hence has something to say against (inter alia) feminism, gay rights, and transgender rights—such ideas are only possible on the basis of a perverse misunderstanding of what could possibly be meant by “biology itself”. But that, unfortunately, is very far from saying that they are not possible. Here as elsewhere a more intelligent attitude has to begin with a willingness to question the myth of the biological given, and to say instead—as feminists have so often done—that every one of us has the right to seek, as far as we can or wish, not to be constrained by the determinism of the body. This is another way in which it is possible, and philosophically interesting, to take a step back from what our roles might seem to dictate inevitably to us. So it is also another way, among others that should by now be obvious, in which my opening thesis (3) can be vindicated: the thesis that at least some ethically interesting role-recalcitrance is a very good thing.

8. I hope it is readily apparent that the forms of role-recalcitrance that have been central to this chapter are not, or need not be, anything like the “trivial” forms of role-recalcitrance—mere whim, lazy disinclination, failure to see an implication, etc.—that I began with in Section 1. Our conceptions of ourselves, and of the identities and roles under which we choose—or are bound beyond choice—to act, live, and understand ourselves, go much deeper than whim; and they have far more structure and shape than mere (dis)inclination. Part of what gives them this structure and shape is the society that we inhabit, in all its concreteness and particularity, its “thickness”. It is an important truth about the self, one noted by MacIntyre at the very beginning of this chapter, that the society in which it is at home penetrates it all the way through. Whatever else I want to say here about our capacity to “step back”, I do not want to deny at least some readings of MacIntyre’s point that the self is a social creation. It is then as selves who are social creations that we attain our own understanding of these identities and roles as in one way or another exigent for us, as being at least

  

45

sometimes necessary identities and roles. In the case of those roles or identities that go deepest for us, not only do we identify with them. We find it hard, maybe even impossible, to conceive of not identifying with them; typically we act out of them immediately and without needing to decide to; often we find ourselves under a moral incapacity to act any other way. Such reasons are, as Bernard Williams would say, internal for us; and though their being internal for us directly affects our will, our choice, it does not follow that we can simply choose what is and is not internal for us.²⁴ “She is my wife—so I must rescue her”; “He is my friend—so I must listen to his troubles”; “She is my daughter—so I must protect her”; “I am a firefighter—so I must get these people to safety.” Nothing is more psychologically compelling than this sort of practical inference; indeed in many cases the inference will go without saying, so that to rehearse it in any conscious way will be “one thought too many”. Still, as I have pointed out, what is psychologically compelling is one thing, and what is logically compelling is another. It is good to be the kind of person for whom these inferences are compulsory; but that does not make them logically compulsory tout court. To see that, we need only recall some other possible inferences of more or less exactly the same form, and reflect on how important it is that we not accept these: “He is my husband—so I must do what he says”; “Jim is their slave—so I must return Jim to them”; “I am a woman—so I must not press my argument too forcefully in debate”; “Patroclus is my friend—so I must avenge his death by killing Hector.” If what makes such inferences compulsory is not logic or rationality itself, but some shared social world and the “thick concepts” available in that world, together with the quality of character of the agent who does the inferring, then here we have to do with shadings and matters of degree, not of all-or-nothing, absolute distinctions. I have argued elsewhere (Chappell 2020) that what reasons are internal for me is not a fixed and entirely determinate matter, at least partly because what is me is not a fixed and entirely determinate matter: it depends at least to some extent on what I identify myself with, and that can and does change over time, quite often through my own conscious choices. I can come to identify more closely with some role that I occupy, as Nicholas Nickleby does while a teacher at Dotheboys Hall, and so the role’s reasons can become my reasons; but I can also move away from a role, as Dorothea Casaubon moves away from the role of being wife to Edward Casaubon; and so what were her internal reasons become external. I can be right or wrong in such choices (whether or not conscious or one-off) about which reasons to identify myself with (or not): both Nicholas and Dorothea seem to be right in what they eventually choose, but the story of many lives is (also) the story of disastrous choices about what roles to identify with, and of

²⁴ Thanks here to Farbod Akhlaghi.

46



course this too applies to Dorothea. Again, reasons can be internal by prolepsis: that is, they’re not internal for me yet, but I treat them as if they were, and that is a way of making it true that they can (and with any luck will) become internal for me. This happens when we are learning our way into a MacIntyrean practice, and in a sense it happens too with Aristotelian pleasures. I don’t yet want to play this Bach passage with a kind of dreamy nonchalance and with a careful listening-out for the key-changes and the time-signature jumps, but that’s how really good pianists play it, and I want to be like them, a really good pianist; I am not yet able to enjoy playing this passage—it’s too damn difficult—but it is enjoyable to the maestros, and again, I want to be like them. As Laurie Paul (2014) too has pointed out in Transformative Experience, future identities can have a certain authority for me now, even though they are not yet my identities.²⁵ What can also change over time, for me, is what is unthinkable or practically necessary. Perhaps there was a time for Achilles when his identity as war-hero was beyond the reach of (his own) deliberative questioning; but the fact that it was so at some time obviously does not imply that it must always remain so. In the passage I quote in Section 3, Achilles does in fact come to question that identity. From what standpoint, it might be asked, can it be possible for him to come to question that? It is correct, but not very informative, to answer “From his own standpoint”; a more useful answer is to say that he questions it from the standpoint of at least some of his other existing and (at this point) unquestioned identities. A familiar scope-distinction is relevant here.²⁶ The possibility of the Olympian standpoint on one’s own life that I discussed in Sections 2–3 need not be understood as entailing that there exists some one standpoint from which all standpoints can be questioned, and so triggering a straw-man argument against the possibility of role-recalcitrance, to the effect that it cannot ever be possible because it would always require us to adopt this view-from-nowhere standpoint, this Sidgwickian “point of view of the universe”. However, what role-recalcitrance definitely does entail is that, for any standpoint, there exists some other standpoint from which it can be questioned. The ascent to the Olympian viewpoint of which Homer makes so much is a vivid, if perhaps imprecise, way of bringing out this pervasive renegotiability of even our deepest role-commitments. That is why, even in Homer—pace MacIntyre as quoted²⁷—the roles that we live in, and live ²⁵ Thanks here to Jesse Prinz. ²⁶ Thanks for discussion to Senthuran Bhuvanedra. ²⁷ “As quoted”, because in fact I think the moves in After Virtue that I have spent some time attacking here are moves that are themselves not in harmony with MacIntyre’s own larger purposes; for MacIntyre’s own central argument, it seems to me, it would be better to agree that the ultimate ground of even the most promising-looking “functional” arguments across the “Is–Ought gap” is really not logical, but sociological and psychological. And indeed MacIntyre himself makes moves in that direction in some of his most interesting later work: see in particular the 1999 essay “Social Structures and Their Threat to Moral Agency” (in MacIntyre 2006a). (Almost all of what MacIntyre has to say in Ethics and Politics (2006a) and The Tasks of Philosophy (2006b) is, I would say, both of the

  

47

through, do not have to be understood as simply trapping us, into coercive functional obligations that we can never avoid. They can also be understood as modulations of our identities that we ourselves can, at least sometimes, choose to grasp and creatively develop for ourselves. In living out the roles that we find ourselves in, we are very often able also to reflect upon those roles. Part of the way we reflect upon them involves deciding how to use them as themes on which to run variations—or how not to use them. The availability of this negative disjunct is the availability of role-recalcitrance, and once more displays its positive potential in our lives.²⁸ Our interaction with our roles, and our role-reasons and role-obligations, like our interactions with the narratives and scripts that we find interesting and full of potential, need not then be either psychologically or logically coercive; they can also be a matter of improvising, and of creativity.²⁹

References Chappell, Sophie Grace. 2020. ‘Roles and Reasons’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons and Obligations. Abingdon: Routledge. Chappell, Sophie Grace. Forthcoming. ‘Gatekeepers, Engineers, and Welcomers’. In J. Beale and R. Rowland (eds.), Wittgenstein and Contemporary Moral Philosophy (Routledge). Chappell, Timothy. 2011. ‘Glory as an Ethical Idea’. Philosophical Investigations 34 (2): 105–134. Everitt, Nicholas and Fisher, Alec. 1995. Modern Epistemology: An Introduction. New York: McGraw-Hill. Goffman, Erving. 1959. The Presentation of Self in Everyday Life. London: Penguin. MacIntyre, Alasdair. 1981. After Virtue. London: Duckworth.

very greatest philosophical interest and also strikingly unassimilated by contemporary ethical philosophy; and this essay is a particularly clear example. My thanks to Sean Cordell for getting me to reread it; not before time.) ²⁸ Potential: role-recalcitrance is not necessarily good, and role-compliance is not necessarily bad. If role-recalcitrance is like scientific revolutions, and role-compliance like normal science—well, sometimes it is a good idea to carry on with the normal; and revolutions, notoriously, are as often destructive as creative. (Thanks to Miranda Fricker.) ²⁹ My thanks for extremely helpful comments to audiences at the Moral Sciences Club in Cambridge and at the Graduate Seminar at CUNY, among whom respondents whose names I was able to record included Farbod Akhlaghi, Shannon Brink, Senthuran Bhuvanedra, Neil Dewar, Jacopo Domenicucci, Alex Fisher, Miranda Fricker, Pablo Hubacher Haerle, Matthew Kramer, Rae Langton, Cathy Mason, Jessie Munton, Milan Noy, Nickolas Pappas, Jesse Prinz, Zachary Richards, Facundo Rodriguez, Shant Shahrigian, and Iakovos Vasiliou. Thanks also to Alex Barber, Sarah Broadie, Sean Cordell, James Holden, and Alasdair MacIntyre himself for their kind help, encouragement, suggestions, and above all patience.

48



MacIntyre, Alasdair. 2006a. Ethics and Politics: Selected Essays Volume 2. New York: Cambridge University Press. MacIntyre, Alasdair. 2006b. The Tasks of Philosophy: Selected Essays Volume 1. New York: Cambridge University Press. Newby, Eric. 1972. A Short Walk in the Hindu Kush. London: Hodder & Stoughton. Paul, Laurie. 2014. Transformative Experience. Oxford: Oxford University Press. Prior, A. N. 1960a. Logic and the Basis of Ethics. Oxford: Clarendon. Prior, A. N. 1960b. ‘The Autonomy of Ethics’. Australasian Journal of Philosophy 38 (3): 199–206. Searle, John. 1964. ‘How to Derive “Ought” from “Is”’. The Philosophical Review 73 (1): 43–58. Taylor, Charles. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge: Cambridge University Press. Williams, Bernard. 1981. Moral Luck. Cambridge: Cambridge University Press. Williams, Bernard. 1993. Shame and Necessity. Los Angeles: California University Press. Wright, N. T. on gay marriage.

3 Roles and Virtues Which is More Important for Confucian Women? Jing Iris Hu

1. Role Ethics and Relationships In Western ethics, it is typically assumed that when we talk about roles, we mean that a moral agent is filling a role or someone is playing a role. I, as a person or agent, fulfil the responsibilities and expectations the role requires of me, such as being a good daughter, a good teacher, a responsible citizen, and so on. Role ethics, when considered in the light of professional ethics, is frequently understood in terms of the specific ethical codes, moral obligations, and moral expectations the roles carry. I may talk to a colleague about my paper on Confucian philosophy as a fellow philosopher and colleague, and when we have to discuss my next year’s teaching, they may switch roles and talk to me as the department chair. However, according to many advocates of Confucian role ethics, that is not how roles are understood in Confucian ethics. In recent years, Confucian role ethics (CRE) has received increasing attention from both scholars of Chinese philosophy and philosophers interested in virtue ethics, autonomy, and personhood. CRE claims that roles and relationships are intrinsic to people’s moral life not only on a practical level but also in a constitutive sense—that is, we are constituted by our roles and relationships with others (Ames 2011; Rosemont 2016). According to this view, I do not just act as a good daughter, fulfil my role as a good teacher, and take seriously my obligations as a good citizen; instead, I am the good daughter to my own parents, the good teacher to my specific students, and the responsible citizen to my community. I cannot be ‘good’ or ‘responsible’ without these roles and relationships. Indeed, this radical view contends that neither I nor my moral agency can exist independently of these relationships and roles. As Rosemont puts it, ‘We are, in other words, the sum of the roles we live in consonance with our fellows’ (Rosemont 2016, 67; 112). Clearly, it is one thing to say that our ethical and social existence rests on the roles and relationships we inhabit, and another to say that as moral agents, we are nothing more than those roles and relationships. Jing Iris Hu, Roles and Virtues: Which is More Important for Confucian Women? In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Jing Iris Hu 2023. DOI: 10.1093/oso/9780192843562.003.0003

50



CRE, which takes a radical role-ethical approach to Confucian ethics, has been met with revisions and criticisms from commentators such as David Wong (2004), Aaron Stalnaker (2020), and P. J. Ivanhoe (2008). Wong (2004), for example, argues that CRE can be interpreted as a developmental approach; that is, we are constituted continuously by our interactions with others in our moral upbringing and moral education. We are not the sole authors of who we are but co-authors together with those who interact with us and shape our views. Thus, moral agency cannot be reduced to the roles we occupy; but in the long run, who we are is constituted by others as much as ourselves. Wong thus offers an important contribution to this debate by reconciling individual autonomy with the CRE’s radical relational view, which sees individuals as the sum of their roles and relationships. Stalnaker is one of the most recent scholars to criticize Ames and Rosemont for their position, which, he contends, leaves no space for moral agency or personal moral preferences. He criticizes Ames for ‘draw[ing] sharp contrasts between their position and “Western” ones . . . and thereby unnecessarily muddy[ing] the waters regarding the characteristics of both role ethics and virtue ethics, their variations, and their possible interrelations’ (Stalnaker 2020, 96). This exaggerated distinction pushes so-called ‘East’ to the opposite of the deontological understanding of moral agency. This results in the notion that Confucianism—or ‘the East’—understands the self, or personhood, to be entirely relationship based and suggests that virtues are secondary to roles and relationships. Stalnaker (2020) argues, however, that many passages in Classical Confucian texts, including the Mencius and the Book of Odes, are clearly indicative of a predisposition towards virtues, which suggests that virtues take precedence over roles in shaping one’s sense of self. Indeed, one’s preferences can conflict with what one’s roles expect of one, thus proving that moral agency in Early Confucian philosophy is not equivalent to one’s social roles.¹ Even though roles and relationships are, without a doubt, an important aspect of Confucian ethics, Stalnaker criticizes Ames for the simplification of Confucian philosophy and the exaggeration of how ‘relational’ it is. Confucian ethics, according to Stalnaker, places clear emphasis on virtues and moral agents; moreover, these moral agents have their own preferences and judgements that cannot be reduced to their roles. Even though many have shared such worries about CRE, a less radical understanding of Confucian roles and relationships remains viable—namely that roles and relationships are still considered important and are dealt with carefully in Confucian ethics, but that moral agents are not to be reduced to their roles and relationships. Further, I argue in this chapter, with the support of historical texts

¹ Stalnaker (2020) illustrates this point with an example from ‘Cypress boat’ that depicts a young woman who rebels against the marriage that has been imposed upon her.

  

51

written by Confucian women, that roles and virtues should be understood as complementing each other. Whether CRE is the best approach to interpreting Confucian ethics should not be understood as a debate about the superiority of roles against virtues, or vice versa. A debate that assumes a dichotomy between roles and virtues, which is not present in historical Confucian texts, bears little significance to our understanding of the moral function of roles. Instead, CRE draws our attention to the moral and normative aspects of roles and, further, explains how relationships help create and sustain a moral self. We should focus our investigation on the special ethical function of roles, and how, sometimes, it is with the help of virtues and rituals that one successfully manages one’s roles and relationships. This leads to the methodology of this chapter and the reasons why historical texts by Confucian women authors are especially relevant to the discussion of role ethics. For example, a worry about CRE is that allowing relationships and roles to dictate moral lives can be dangerous, especially for those vulnerable to the abuses of power. Paul Goldin (2016) discusses a few historical texts where women were caught in moral dilemmas unique to women, such as having to choose between loyalty to their father and loyalty to their husband. Goldin notes, ‘The demands of interpersonal relationships occasionally thrust men into agonizing moral dilemmas too, but women were inherently more vulnerable’ (Goldin 2016, 32). Moreover, the philosophical resources provided to these women were rare, frequently contradictory, and even recommend suicide as a solution (Goldin 2016; 2001), which appears to suggest that there is no theoretical way to reconcile the conflicting roles of women. Thus, if we understand Confucian ethics as an ethical system in which we are indistinguishable from our roles, we won’t be able to account for cases where one’s disposition and desire depart from what one’s roles would expect, as Stalnaker’s example about the young woman rebelling against her marriage arrangement would show. On the other hand, if we understand Confucian ethics as one that heavily relies on our roles, we would need to discuss methods and means, such as looking up to moral exemplars, citing rituals, and using virtues as moral guidance, to deal with moral dilemmas, conflicting situations, and abuse from those with power.² In this chapter, I provide additional evidence from the texts written by Confucian women such as Ban Zhao and the Song Sisters, and argue that moral agents, especially female agents, frequently use

² Sarah Mattice argues that CRE can provide some resources in cases of domestic violence, for instance, by publicly humiliating the perpetrator, thereby establishing that such abuse is not accepted in this community (Mattice 2016). I, however, remain sceptical of the efficacy of measures such as social pressure when it comes to correcting abusive behaviour within the family. More importantly, I question how helpful such measures are to the victims of abuse. Before more comprehensive discussions are offered on this topic, I remain reluctant to conclude that CRE, rather than Confucian virtue ethics or Confucian reliance on rituals (li), is more helpful to people vulnerable to abuse.

52



virtues as their moral guides instead of roles. Roles are not sufficient by themselves to provide ethical guidance for historical Confucian women. In addition, one of the allies of the relational and role ethics reading of Confucian philosophy is a feminist philosophy that discusses the self as relational—that is, as a situated self (Lai 2016). Given its special focus on roles and relationships, Confucian philosophy recognizes the particularity of each agent within particular sets of relationships; each agent is unique due to their relationships. For example, I am not defined by my general rational capacity but by the fact that I am the person having a father–daughter relationship with my father, a sister-in-law relationship with my brother’s wife, and so on. These are not abstract relationships. Confucian philosophy can be used as a resource to argue against the abstract, autonomous, rational moral agent—or ‘Man of Reason’ (Lai 2016)—that the deontological tradition assumes. However, it is worth noting that many of the Confucian thinkers that the supporters of CRE presently discuss are male, including Confucius, Mencius, Xunzi, Zhuxi, and Wang Yangming. These Confucian thinkers are sometimes criticized within the Confucian tradition and blamed for the suppression and/or neglect of the plight of women, though as Lisa Rosenlee points out, some of these accusations may not be entirely fair (Rosenlee 2016; 2012). It is therefore important to address the concern that women’s voices and plights may not have been sufficiently considered by advocates of CRE or by Confucian philosophers in general. I will argue that it is methodologically important to take a close look at the philosophical accounts of Confucian women and the texts written for women. Women’s plight and moral lives have been used as evidence by Goldin (2016) and Stalnaker (2020) to argue against CRE; for example, as Goldin (2016) argues, because women are more vulnerable to abuses of power and moral dilemmas and can be ‘consumed’ by their roles, they may have a perspective on Confucian ethics that is less favourable to reconciliation with CRE than that of their male peers. However, none of the existing literature has considered historical Confucian women authors’ writing and their view on the topic of CRE. This neglect would have serious theoretical consequences and this chapter aims to rectify this methodological problem. For example, as we can see from the preface of Ban Zhao’s writing, from a young age, pre-modern Confucian women could anticipate a drastic change in roles and relationships—from daughter to wife and daughter-in-law—as they grew older. This involved changes to their living environment and a shift in the structure of their key social relationship—once married, they no longer live with their biological family in the home where they grew up. On a daily basis, they had to manage their relationships with their mother-in-law and father-in-law instead of their own mother and father. Such a stark change in social relationships and roles is much rarer for men. Since a woman’s childhood and adolescent education essentially prepared her for such a shift, I argue that in comparison with men,

  

53

women had more need of a virtue-driven ethic to sit alongside a role-driven one, since that enabled them to cultivate the virtues necessary for being a virtuous wife and daughter-in-law pre-emptively. Virtues can also equip them to navigate the complex roles they must take on within a new family structure that included not only their husbands, but also their parents- and siblings-in-law.

2. Why Should We Talk About the In-laws? Despite the Confucian’s focus on roles and relationships, women are only mentioned once in the five paradigmatic relationships (Mencius 3A.4 and see below). Furthermore, many wonder whether some of the roles and ethical achievements are gender specific in practice, though in theory they are open to all genders in Confucian thoughts. Karyn Lai’s account of the situated self is of particular relevance here. She points out a problem in Confucian philosophy: ‘in failing adequately to account for gender as an important dimension in relationships, it [the current methodology] may contribute to inequalities for women and also possibly for men’ (Lai 2016, 119). Many of the Analects’ conversations on friendship, just like its depiction of ‘junzi’ (the exemplary person, 君子), are applicable only to upper class men, as Erica Brindley (2009) points out. Lai further notes that in the Analects 15.10, friendship is discussed in the context of a person’s taking up official duties, which is an activity limited to men (Analects 15.10; Ames and Rosemont 1999). This is not an option for women. Norman Kutcher’s analysis of friendship in the Confucian historical context illustrates this point: he writes, ‘[h]aving a good friend should make one a better son, brother, or official’ (Kutcher 2000, 1616). Without ever being explicitly stated, attention is given to the way that men relate to each other in close companionship, but there is no mention of the friendships between women. Those interested in promoting the contemporary relevance of Confucian relationships must address its silence regarding the impact of gender in close personal relationships (Lai 2016, 121). For example, when discussing the essential relationships in the Confucian tradition in Mencius, five relationships are listed: [The sage-king] appointed Xie to be the Minister of Education to teach people human relationships: close affection between father and son, rightness between superior and minister, differentiation between husband and wife, proper order between young and old, and trustworthiness between friends. (Mencius 3A.4, adapted from the translation by Lau 2003, 115–117)

In this passage from the Mencius, female roles—the subservient counterpart to ‘husband’ no less—are only mentioned once. Women’s roles as mothers, daughters, friends, and sisters-in-laws are not included here, even though these

54



relationships may be the focus of their moral lives. Moreover, we do not see relationships among women represented in these five relationships. Even though one may say that friendship—the last relationship—could include female friendships, Lai and Brindley astutely observe that the activities associated with the kinds of friendships discussed are, in fact, gender specific to men (Lai 2016; Brindley 2009). Moreover, this passage does not capture the wide range of relationships between Confucian women and men. To Lai and Brindley’s observations, I’d like to add that mother–son and mother–daughter relationships are also not mentioned—and they cannot simply be reduced to the paradigm of the ‘father– son’ relationship for the same reason that relationships between sisters and other siblings cannot be modelled after brother–brother relationships: each sibling has a unique social role assigned to them. For example, a younger sister would have a very different relationship with her older brother and his wife than with a younger brother in pre-modern China. Furthermore, the familial role in the household, which consumes a large part of women’s lives, is also left out in these five relationships—in the Lessons for Women, Ban Zhao devotes two of her seven lessons towards teaching her daughters and other women how to be respectful and obedient to their parents-, brothers-, and sisters-in-law. Without considering these important relationships, the Confucian idea of a ‘relational self ’ or rolebased ethics rests on only a small fraction of all human relationships, reflecting a male-centric perspective on which relationships are important. It would, therefore, be hypocritical for a philosophical approach based on a male-centric perspective on relationships to claim that it can contribute to feminist theory today without first looking at the works of women thinkers who wrote on relationships in their own tradition.

3. Lessons for Women and the Analects for Women In the next few sections, I will examine two works by Confucian women, Lessons for Women (LFW) by Ban Zhao (45–117 ) and the Analects for Women (AFW) by the Song sisters (?~820/825). In the preface of her writing, Ban Zhao clearly indicates that she is worried about the transition of roles that her daughters will experience after they marry, and her writing is intended to help them with these changes. As a matter of fact, it brings her great distress, ‘worries and frustration’, that her young daughters do not know the proper rituals of a married woman. This worry motivates Ban Zhao to write down the seven lessons: It, however, grieves me to see you, my daughters, who have just reached the age of marriage, to not have been gradually taught more nor have you learned proper rituals of being a married woman. I fear that you may lose face with other

  

55

households and bring shame upon your family clan. I am now seriously ill; my life is uncertain. As I think of you in such an [untrained] state, I am distraught by worries and frustration (chou zhang) whenever this thought comes to my mind. (LFW, Preface; Pang-White 2018, 41)³

As mentioned in the introduction, most young women, in pre-modern China, would leave their natal homes and go to live with their husband and his family when they came of age. Though men also experienced the addition of new relationships, ‘gaining’ new in-law relationships through their marriage, their primary relationship as a son did not change. They did not have to leave their households, and their relationships with their in-laws were also of less practical importance than those of their wives. In Ban Zhao’s Lessons for Women, quite a few passages discuss women’s virtues and suggest a view of virtues that is based on dispositions instead of particular social relationships. For example, she writes, ‘A woman should be humble, yielding, reverent, and respectful. She should place others before herself. If she does something good, let her not boast. If she does something bad, let her not deny it . . . This is to act as the lowly, the weak, the one beneath others (Lesson 1).’ One may argue that the characteristics she attributes to women are still about relationships—after all, according to Ban Zhao, women are supposed to act weak (as opposed to strong) and lowly (instead of high) so that they can better maintain their relationships. However, this advice sounds more like a general guiding principle, which she believes matches up with women’s nature, than specific rules for particular relationships. As a matter of fact, the first lesson she composed for her daughters is titled ‘The Lowly and the Weak,’ and it lectures young women not only to act as lowly and weak but also to recognize their place in the world as such. She starts the passage by identifying the three responsibilities of women: namely being lowly and weak, carrying out household work, and continuing ancestral religious rites. In case Ban Zhao’s emphasis on women’s nature and virtue are not made clear enough in the previous passage, let us take a look at Lesson 7, titled ‘Harmony with Younger Brothers- and Sisters-in-Law’: If a woman is virtuous, beautiful, humble, and compliant, she will be able to rely on righteousness to deepen their [her younger brothers and sisters-in-law] good wills, and uphold kindness to bind their assistance, which will enable her beautiful virtue to be known while keeping her flaws and mistakes hidden. Thereupon, her father- and mother-in-law will be proud of her goodness; her husband will praise her virtue. Her repute will shine in her district and

³ All translations of the LFW and AFW are from the translation of Ann Pang-White (2018).

56

 neighborhood, and the luster of her name will even extend to her own father and mother. (LFW, Lesson 7; Pang-White 2018, 64)

In this passage, we see clear evidence that when navigating the relationship with younger family members, Ban Zhao believes the virtues of beauty, humility, and compliance are at the core of a good woman, who can subsequently play the role of a good wife, a good daughter-in-law, and a good daughter. Her pursuit of virtues would precede, though not be independent of, her consideration of the roles. Even though the roles such as a daughter-in-law, a wife, a neighbour, are still an important part of her moral consideration, the virtues are a more reliable source of moral guidance in complex moral situations. Her righteousness and kindness are the key virtues that help her navigate complex household relationships and help her to gain not only the affection but also the respect of her in-laws. These relationships do not precede or replace virtues; on the contrary, Ban Zhao’s book is geared towards her unmarried daughters, and she therefore instructs that they should acquire and practise these virtues before marrying. In other words, they must practise these virtues before actually taking up the roles of a married woman and before even knowing what kind of family they are marrying into and who they are marrying. Ban Zhao seems to be suggesting that becoming a virtuous woman is the only path to a successful marriage and life. Interestingly, Ban Zhao adds that it is never right for a woman to rely on her role as a senior member of the household to manage her relationship with the younger members of the household, such as the younger brothers and sisters of the husband. In a typical Confucian household, the wife of a man would be considered senior to the younger siblings of the man, but this seniority should not be used to justify her poor behaviour to the younger siblings of the husband. In other words, her role as the wife of an older brother can be abused. Ban Zhao recognizes and warns her daughters of this possibility and the consequences of such abuse; she thus makes the point that virtues are to be stressed and roles cannot provide us with reliable moral guidance. More specifically, one should never relax one’s manners towards in-laws simply because one enjoys the husband’s affection. According to Ban Zhao, such a mistake would destroy one’s relationships and lead to a life of disgrace and shame: If a woman is stupid and unintelligent (chun), she will rely on her name as an older sister-in-law, acting haughtily toward her younger brothers-in-law; likewise, she will treat her younger sisters-in-law arrogantly because of her husband’s favor. Once arrogance is expressed, whence comes harmony? (LFW, Lesson 7; Pang-White 2018, 64)

She continues, ‘Once kindness and righteousness are violated, whence arrives (zhen) good reputation?’ Violating the two virtues of kindness and righteousness

  

57

can only result in a woman’s public humiliation, provoke the anger of her in-laws and husband, and cause her own parents to feel embarrassment at her behaviour and reputation (LFW, Lesson 7). In addition, when discussing how best to serve one’s parents-, sisters-, and brothers-in-law, Ban Zhao remarks that only a few virtues, such as humility and compliance, are at the core in dealing with and managing these relationships (LFW, Lessons 6 and 7). It is thus clear that it is insufficient to rely on the seniority relationship as guidance for how one should behave, since this can easily lead to abuse. Roles, especially those that place one in a potentially more powerful position, are not a self-sufficient source of moral guidelines needed to navigate one’s relationships successfully. For example, a woman cannot rely on her seniority over a younger sister-in-law and neglect the necessary virtues of a good sister-in-law. In Ban Zhao’s view, to become successful as a Confucian woman, one must use virtues as guidelines and never resort to the supposed moral authority that their roles might lend. In the lesson ‘Women’s Conduct’, Ban Zhao also outlines four different conducts that she deems essential to all women; the foremost is virtue and the other three are speech, appearance, and work. Indeed, it is hard to discern any role that is more important than virtue in Ban Zhao’s texts, and it is clear that the radical interpretation put forth by Rosemont and Ames—that we are the sum of our roles and nothing more—cannot be sustained in Ban Zhao’s case. She clearly thinks that these women have their own moral disposition, temperament, roles to fulfil, and relationships to manage. A person should not rely on the power or authority of her role and her close relationship with others (e.g. her husband) to manage her relationships with others; instead, she should act on the virtues of kindness and righteousness. Lesson 7 also supports the idea that virtues precede roles, as Ban Zhao attributes the success of managing relationships to virtues instead of vice versa. Of course, a few passages appear to discuss how to manage relationships; however, the verdict from Ban Zhao seems to be that you must be virtuous so you can carry out your roles well. Your virtues will manifest themselves in the roles. In this section, we have seen that, in Ban Zhao’s view, roles are not sufficient in providing moral guidance. When she discusses a woman’s relationship with her husband (Lesson 2), and a woman’s relationship with in-laws (Lessons 3 and 7), she repeatedly emphasizes that virtues are the only reliable source of moral guidance. As Ban Zhao’s texts indicate, Confucian ethics do not necessarily treat roles and relationships as the sole or sufficient source of moral guidance—instead, they need to be supplemented by the pursuit of virtues.

4. Content of the Women’s Works A close survey of the titles of the Confucian Four Books for Women shows us that virtues, such as diligence, intelligence, one-mindedness, benevolence, loyalty,

58



righteousness, frugality, watchfulness, prudence, and filial piety, are used in titles many times. There are also sections titled after particular relationships and roles; however, these are normally the last few sections of the text, and they follow sections that talk about activities such as establishing one’s person, maintaining virtuous nature, self-cultivation, learning the work, etc. For example, in LFW, the first lesson is about the nature of women, and the third is about ‘Respect and Compliance’, the fourth is about ‘Women’s Conduct’, and the fifth is about ‘OneMindedness’. The sixth is titled ‘Conceding Obedience’, and it concerns how to manage one’s relationship with parents-in-law, while the seventh is about managing relationships with younger brothers- and sisters-in-law. This quick survey (Table 3.1) tells us that, for these Confucian thinkers, virtues—rather than roles—are considered of great importance, and perhaps are even more fundamental. Even though dealing with relationships is an important concern for these thinkers, their texts suggest, quite clearly, that one must first think about and learn about virtues. Evidently, role ethics does not explain Confucian women’s writings and their lived experiences as much as we might desire.

Table 3.1. Section titles divided by content in early Confucian texts, including Biographies of Women, Lessons for Women, and the Analects for Women. Author(s)

Biographies of Women, Liu Xiang (c.77–76 )

Lessons for Women, Ban Zhao (c.45–117 )

Analects for Women,Song Ruoxin (?–820) and Song Ruozhao (?–825)

Virtues/ Conducts

2 Virtuous Intelligence 3 Humaneness and Wisdom 4 Chastity and Obedience 5 Integrity and Righteousness 6 Penetrating Rhetorical Skill 7 The Evil and Spoiled

1 The Lowly and the Weak 3 Respect and Compliance 4 Women’s Conduct 5 One-Mindedness 6 Conceding Obedience

1 Establishing One’s Person 2 Learning the Work 3 Learning the Rituals 4 Rising Early 9 Managing the Household 10 Hosting Guests 11 Harmony and Gentleness 12 Guarding One’s Integrity

Roles

1 Model Motherhood

2 Husband and Wife 7 Harmony with Younger Brothersand Sisters-in-Law

5 Serving One’s Parents 6 Serving Parents-in-Law 7 Serving One’s Husband 8 Instructing Boys and Girls

Note: Section titles are divided according to whether the passage primarily concerns women’s virtues and/or conduct, or roles and relationships. Evidently, most of the sections have to do with virtues and conducts.

  

59

5. Confucian Familial Roles In this section, I will briefly examine Confucian familial roles and argue that they are actually both familial and professional roles in today’s sense. In other words, being a good daughter-in-law is not only a familial role but also a job. This is perhaps why some of the role requirements for women are very specific and akin to what we might call a professional ethical code. This observation once again debunks the idea that Confucian roles are drastically different from how roles are perceived or understood in Western ethics. One may argue that the roles Ban Zhao discusses are mostly familial roles— none of them are public roles such as the roles in the Ruler–Minister relationship. Of course, with rare exceptions, pre-modern Confucian women were, in theory, limited to the inner court or the inner domain of a household.⁴ I should mention to readers unfamiliar with Chinese philosophy that Confucian philosophy offers great philosophical resources to women because of its relational view of moral agency—women and men are not defined by their biological features or reasoning capacities. The implication is that in Confucian philosophy, women are never considered to be limited in any way by their natural endowments. This is especially inspiring when compared to philosophical theories that base our understanding of personhood on capacities such as narrowly defined rationality. On the other hand, Confucian women are limited by social activities and education. As some philosophers point out, the division that separates the realm of women from that of men in Confucian philosophy is not feminine versus masculine or emotion versus rationality; instead, it is the distinction between the inner and outer domains (Rosenlee 2012; Pang-White 2018). This pair of concepts does not strictly correlate with the public domain and the domestic domain either. The division of nei and wai, or the inner and outer realm, do not always strictly correlate with gender. They are a pair of spatial binary terms that have their own unique characteristics in Chinese texts. Nei literally means ‘in’, or ‘inner’; wai means ‘out’ or ‘outer’. In early Confucian texts, the nei–wai distinction has many different meanings: it could mean, as in the Mencius, the inner virtues and their outer exhibition; or, as in Xunzi and Guanzi, this distinction may signify the boundary between a civilized area and a barbaric, chaotic one that does not maintain proper social distinctions, such as the ones between men and women. This distinction of nei–wai is a functional definition with a somewhat flexible boundary—it is ‘a relational category that describes a series of nested hierarchies ⁴ I emphasize this is a theoretical limitation because in practice there are times women are responsible for a large range of responsibilities that we would nowadays consider as public, rather than domestic. See Rosenlee (2012; 2016) for a more detailed discussion on how the In/Out (Nei/Wai) distinction in Chinese philosophy is fluid throughout history. Also, see Rosenlee (2012) for a detailed discussion on how Confucian philosophy is frequently wrongly seen as a scapegoat for the suppression of women.

60



whose boundary changes with context’ (Ko 1994, 144–145). With this ideal in mind, we understand that women are limited to the inner domain, and the inner domain is both where they cultivate filial love and where their performance is evaluated. For them, familial roles are more like a combination of family roles and professional roles. Ban Zhao’s readership, I argue, could explain, at least partially, why she believes that virtues precede roles and moral agency precedes the exercising of roles. The young women who are unmarried are studying and preparing themselves for roles they do not have yet. They were not born into these in-law relationships, and nor are they related by blood to the family they will live with after marriage. If we follow the Confucian explanation for why we should care for our elderly parents and mourn for them for an extended period of time after they pass away—because they carried us in their arms for three years before we could walk—we can see why a woman’s filial piety is not directly owed to her parents-in-law. The parents-inlaw didn’t carry the women in their arms when they were infants. A woman’s relationship to her in-law’s family is thus not only a familial relationship but also her work and occupation. These roles and relationships are both her familial and professional obligations. For example, it is made clear that household work is a married woman’s responsibility, and in both Ban Zhao’s Lessons for Women and the Song sisters’ Analects for Women there are highly specific instructions for women regarding their appearance, their clothing and hair, whether to serve tea or soup, and whether to rise up early in the morning: Every young woman should get used to the routine of getting up and putting on clothes at the fifth watch of the night when the roosters are crowing. After she has washed her face and rinsed her mouth, she shall groom and dress herself at will. Thereupon, she will gather firewood to start the fire [for the stove], go to the kitchen early, scrub the pots, wash cooking utensils, boil water, and make tea. According to the family’s condition of having plenty or only simple means, she will steam or boil food, arrange vegetables, quick-fry fermented beans, grind (chong) ginger, readily add them to the dishes, and make food suitably sweet, plain, or fragrant. She will also place bowls and plates in an orderly manner and set up the table. All three meals (can), whether during the day or at night, should be prepared similarly. Because she rises early at dawn, she can complete all things without problems. (AFW, ‘Rising Early’, Section 4; Pang-White 2018, 91)

The passage from AFW and a number of passages from LFW (e.g. Lesson 4 ‘Women’s Conduct’ and Lesson 6 ‘Conceding Obedience’) provide detailed expectations regarding what women should do. While Stalnaker’s observation of the difference between the early Confucians’ attention to roles versus the contemporary view of roles still stands (‘early Confucians do not limit their reflections to what we today call professional ethics, which tends to focus on the duties of expert

  

61

professionals such as lawyers and doctors to their clients or patients’, Stalnaker 2020, 97), we do see similarities between the Confucian code of conduct for women and modern professional role ethics. Contrary to the claim by supporters of role ethics that Confucians tend to focus on familial affection and respect for these relationships, female Confucian writers appear to care just as much about conduct and virtues as they do about relationships and roles. I think this showcases the fact that Confucian women, in dealing with their daily lives, were confined to the Nei domain (see Rosenlee 2012, 2016), and so their roles were also, in a sense, their jobs. Stalnaker observes that ‘the early Rú analyze all of human community and society in terms of roles, which are in turn defined by the relationships that constitute them, with familial roles at the centre of their attention, although they also attend to at least some quasi-professional roles, such as ruler and minister’ (Stalnaker 2020, 97). Actually, for pre-modern women, their entire lives are defined by familial roles, and they do not have professional roles, or, if we put it in another way, their family roles are their professional roles, since it is their responsibilities to manage nei affairs. During their lifetime, men experience an expansion of different kinds of roles—familial relationships, such as father–son or brother–brother relationships, are considered the ethical foundation of their social relationships. Conversely, given the lack of roles for women in the wai domain, familial roles are all that women will know. It is thus more important for them to ensure that these roles do not consume them and do not become abusive.

6. Establish the Person In the previous sections, I have demonstrated that virtues are stressed in one’s moral life and moral cultivation in both LFW and AFW. The focus on virtues may in fact be a quite commonly held view in early Confucian texts in addition to these two texts by women authors (e.g. Mencius and The Book of Odes, see Stalnaker 2020). As for Ban Zhao and the Song sisters, while roles and relationships remain important moral considerations, they all consider virtues to be a more reliable moral guide. Granted, their arguments are within the Confucian philosophical contexts of a situated and concrete agency—meaning that the moral self is cocreated and sustained by their relationships with others. This brings us to the last puzzle of this chapter: What is the relationship between virtues and roles? How do we reconcile the focus on roles and relationship (or some version of CRE) with a more virtue-focused approach? A more comprehensive study would be necessary in order to adequately address these questions, but here I will discuss a concept that is particularly relevant to them and to the Confucian idea of a moral self, namely the idea of ‘Establish[ing] the Person’, which is discussed in the first passage of AFW by the Song sisters. The

62



AFW was composed by the Song sisters during the Tang dynasty (618–906 ).⁵ This is much later than the composition of the Classic Confucian texts. However, there are a few passages that make it very clear that one needs to cultivate one’s own moral dispositions. To them, the foremost thing for a woman to consider is to establish herself morally and socially, instead of fitting into her roles. Ban Zhao started her text with what she considered to be the nature of women and the normative aspect of women (lowly and weak); in the Analects for Women, the Song sisters also begin with a section about the nature or the personhood of a woman. In Establish the Person, they focus first on cultivating virtues such as tranquillity and chastity, in order to become pure and honoured. The acquiring of these virtues is subsequently associated with several related rules of conduct, such as how a person should carry herself when experiencing joy or when talking: Every young woman should first learn how to establish her person (lishen). To establish her person, she must cultivate tranquility and chastity. If she is tranquil, her body will be pure. If she is chaste, her person will be honored. When walking, she should not turn her head and look back. When speaking, she should not open her lips too wide (xuan). When sitting, she should not shake her knees. When standing, she should not sway her skirt. When happy, she should not laugh aloud. When angry, she should not raise her voice. [Men and women] should manage the inner sphere and the outer sphere respectively . . . Only by being established in propriety and rectitude can a woman be [truly] a person. (AFW Section 1)

I should note that the phrase—establish the person—is not simply about cultivating the virtues, it also has the connotation of establishing one’s moral autonomy. The term li shen literally means ‘standing up/establishing oneself ’. In the text above, clearly one is expected both to acquire certain inner qualities, such as tranquillity and chastity, and to conduct oneself externally in certain manners in order to be considered an ‘established person’. As a matter of fact, in the Confucian texts, this term has the connotation of establishing oneself morally and gaining a strong moral identity. For example, in order to explain the necessity of learning the book of Propriety, Confucius states that ‘If you do not learn the book of Propriety, your character cannot be established’ (Analects, 16.13; Chinese Text Project 2006–2022).⁶ Furthermore, in the famous passage that describes the stages of a person at different ages, Confucius states that when he was in his thirties, he was able to establish himself (li shen), or ⁵ The exact dates of when the Analects for Women is written is unknown, however, we do know the Song sisters—Song Ruoxin and Song Ruozhao passed away in 820 and 825 . ⁶ This translation is based on the English text at the Chinese Text Project, 2006–2022, which uses a reprint of James Legge’s 1861 The Chinese Classics, volume 1, printed and distributed by the translator.

  

63

‘took a stance’ (Analects 2.4; Ames and Rosemont 1999, 76). The most famous passage relating to the idea of establishing oneself is, ‘Now the man of perfect virtue, wishing to be established himself, seeks also to establish others; wishing to be enlarged himself, he seeks also to enlarge others’ (Analects 6.30; Chinese Text Project 2006–2022). As we can see from these passages, just like in AFW, several passages in the Analects highlight the importance of one being able to take a stance and establish oneself in one’s moral learning. Unlike Rosemont’s (2016, 38–39) claim that ideas such as ‘autonomy’ are completely irrelevant to early Chinese thought, autonomy is discussed in the first section in AFW as well as in many passages of the Analects (see above).⁷ Of course, some may object to my claim and argue that the idea of li shen or establishing has little to do with autonomy—this would be true if we understand autonomy in a Kantian deontological sense as a product of human dignity narrowly construed in the rationalist tradition. However, many have offered alternative ways of approaching concepts such as autonomy. Catriona MacKenzie and Natalie Stoljar, for instance, famously discuss relational autonomy from a feminist perspective (Mackenzie and Stoljar 2000; Stoljar 2018), while Wong argues for the understanding of autonomy as an achievement concept about one’s ability to reliably apply certain global virtue traits (Wong 2004). And we also need to consider how moral agency and moral authority are discussed not only in the Analects, Mencius, and Xunzi but also in the works of female authors. How do we reconcile this idea with the Confucian stress on social relationships and roles? It seems that Confucians view roles and relationship as inevitable—we are all born into a network of relationships and are expected to sustain these relationships. We are also expected to not only cultivate our own virtues but also nourish the lives of those with whom we interact. In addition to the inevitability of roles and relationships, Confucians are also aware of associated dangers: of becoming a people pleaser, of the ‘thieves of virtues’ (‘village worthies’ in the Analects) being swayed by the opinions and preferences of others, or, worse, of altogether surrendering one’s moral reasoning to others who are more senior or considered authorities. This is why establishing oneself and setting clear goals are important parts of anyone’s moral education. Moreover, this is also why, without considering the relational nature of Confucian ethics as the context, some even misunderstand Confucian ethics as too individualistic due to its clear focus on one’s own moral cultivation (Slote 2020). In conclusion, the focus on establishing one’s person gives us a lot to think about, especially given the Confucian understanding of moral autonomy and moral reasoning, which is beyond the scope of this chapter.

⁷ It is highly likely that the Song sisters take the idea of li shen from the Classical texts The Analects because they think this idea is particularly relevant and important to their women readers.

64



7. Conclusion In this chapter, I have examined the attractive understanding of Confucian ethics as a form of role ethics. By taking into account Confucian women’s writings, specifically LFW by Ban Zhao and AFW by the Song sisters, and the plight of women as articulated by Paul Goldin, I have argued against the understanding that we are constituted by our roles and relationships. I have shown that our moral agency cannot be reduced to our roles and demonstrated that virtues and conduct are important topics in the works of Confucian women. I have also demonstrated that if we want to interpret Confucian ethics as role ethics, as Ames and Rosemont have done, we must successfully address two questions: the first is whether Confucian philosophy provides philosophical resources to address the moral dilemmas and conflicts that accompany roles and relationships; the second is whether role ethics provides philosophical resources to guide a person through drastic role/relationship changes such as those experienced by pre-modern Confucian women when they marry. However, perhaps the most interesting question to ask here—in addition to suggesting that a radical understanding of Confucian ethics such as CRE is not reflective of women’s writing and lives—is that of how we can reconcile the apparent importance of roles and relationships in Confucian ethics with Confucian understanding of moral cultivation and living a moral life. How does a woman establish herself morally through her roles or by rebelling against the roles she is expected to take on? I invite further investigations of these questions.

References Ames, Roger T. 2011. Confucian Role Ethics. Hong Kong: Chinese University Press. Ames, Roger T., and Henry Rosemont, Jr. (trans.). 1999. The Analects of Confucius: A Philosophical Translation. New York: Ballantine Books. Brindley, Erica. 2009. ‘ “Why Use an Ox-Cleaver to Carve a Chicken?” The Sociology of the Junzi Ideal in the Lunyu’. Philosophy East and West 59 (1): 47–70. Chinese Text Project 2006–2022. Analects. . Goldin, Paul R. 2001. The Culture of Sex in Ancient China. Honolulu: University of Hawaii Press. Goldin, Paul R. 2016. ‘Women and Moral Dilemmas in Early Chinese Narrative’. In The Bloomsbury Research Handbook of Chinese Philosophy and Gender. Ann PangWhite (ed.), 25–36. London: Bloomsbury. Ivanhoe, Philip J. 2008. ‘The Shade of Confucius: Social Roles, Ethical Theory, and the Self ’. In Polishing the Chinese Mirror: Essays in Honor of Henry Rosemont, Jr. Marthe Chandler and Ronnie Littlejohn (eds.), pp. 34–49. New York: Global Scholarly Publications.

  

65

Ko, Dorothy. 1994. Teachers of the Inner Chambers: Women and Culture in Seventeenth-Century China. Stanford, Calif.: Stanford University Press. Kutcher, Norman. 2000. ‘The Fifth Relationship: Dangerous Friendships in the Confucian Context’. The American Historical Review 105 (5): 1615–1629. Lai, Karyn. 2016. ‘Close Personal Relationships and the Situated Self: The Confucian’. In The Bloomsbury Research Handbook of Chinese Philosophy and Gender. Ann A. Pang-White (ed.), 1st edn, pp. 111–126. Bloomsbury Research Handbooks in Asian Philosophy. London: Bloomsbury. Lau, D. C. 2003. Mencius, A Bilingual Edition. Revised edition. Hong Kong: The Chinese University Press. Mackenzie, Catriona, and Natalie Stoljar (eds.). 2000. Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self. New York: Oxford University Press. Mattice, Sarah A. 2016. ‘Confucian Role Ethics in the 21st Century: Domestic Violence, Same-Sex Marriage, and Christian Family Values’. In Feminist Encounters with Confucius. Mathew Foust and Sor-Hoon Tan (eds.), pp. 198–225. Leiden: Brill. Pang-White, Ann. 2018. The Confucian Four Books for Women: A New Translation of the Nü Sishu and the Commentary of Wang Xiang. New York: Oxford University Press. Rosemont, Henry. 2016. ‘Rights-Bearing Individuals and Role-Bearing Persons’. In Confucian Role Ethics: A Moral Vision for the 21st Century. Roger Ames and Henry Rosemont (eds.), pp. 33–57. Taiwan: V&R Academic National Taiwan University Press. Rosenlee, Li-Hsiang Lisa. 2012. Confucianism and Women: A Philosophical Interpretation. Albany, NY: SUNY Press. Rosenlee, Li-Hsiang Lisa. 2016. ‘Multiculturalism and Feminism Revisited: A Hybridized Confucian Care Ethics’. In The Bloomsbury Research Handbook of Chinese Philosophy and Gender. Ann Pang-White (ed.), pp. 157–172. London: Bloomsbury. Slote, Michael. 2020. Between Psychology and Philosophy: East–West Themes and Beyond. Palgrave Studies in Comparative East–West Philosophy. Cham: Springer International Publishing. Stalnaker, Aaron. 2020. ‘Roles and Virtues: Early Confucians on Social Order and the Different Aspects of Ethics’. In Perspectives in Role Ethics: Virtues, Reasons, and Obligation. Tim Dare and Christine Swanton (eds.), pp. 95–122. New York: Routledge. Stoljar, Natalie. 2018. ‘Feminist Perspectives on Autonomy’. In The Stanford Encyclopedia of Philosophy. Edward N. Zalta (ed.), Winter 2018. Metaphysics Research Lab, Stanford University. . Wong, David B. 2004. ‘Relational and Autonomous Selves’. Journal of Chinese Philosophy 31 (4): 419–432.

4 John Dewey’s Analysis of Moral Agency as Pragmatist Role Ethics John Simons

1. Introduction In his analysis of moral agency, John Dewey, one of the founders of American Pragmatism, recognizes that moral obligations arise naturally in, and are constitutive of, the role relationships that are necessary to a society’s existence. For him, the regulation of the expectations that people have of one another’s conduct in role relationships is the purpose of morality, its raison d’être, and moral agency is realized in the enactment of social roles. This argument may seem novel to those familiar with Dewey’s views on moral agency, because neither he nor his exegetists present them from the standpoint of role ethics. Yet the immanence of that standpoint in his work is evident in the relationship between his own naturalistic assumptions and the sociology of self-formation that he presupposes, one developed by his colleague and fellow founder-pragmatist, George Herbert Mead. The aim of this chapter is to show that the rendering of Dewey’s analysis of moral agency as his account of pragmatist role ethics makes its resources and their significance more obvious than they are in standard accounts, and more evidently a challenge to alternative analyses. This rendering presents role ethics as part of the core subject matter of moral philosophy rather than as one of its specialist domains. Section 2 of the chapter shows that, as Dewey recognized, any tenable theory of normative ethics needs to be grounded in role ethics because the normative regulation of role relationships is one of the existential needs of society. The subject of Section 3 is George Herbert Mead’s explanation, presupposed by Dewey, of how organisms acquire selves and the capacity for reflective participation in normatively regulated role relationships. Section 4 reviews contributions to the concept of ‘role’ in moral philosophy inferred from Mead’s sociology by some British philosophers in the 1960s. Section 5 presents the essentials of Dewey’s analysis from a role-centred standpoint. Section 6 concludes by suggesting that by clarifying its advantages over traditional theories and their derivatives, the new rendering may make his theory an attractive foundation for future studies of the

John Simons, John Dewey’s Analysis of Moral Agency as Pragmatist Role Ethics In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © John Simons 2023. DOI: 10.1093/oso/9780192843562.003.0004

  ’     

67

ethical features of social roles, and help role ethics to become seen as a major focus of moral philosophy.

2. Why Normative Ethics is Necessarily Role Ethics In an essay published seven decades ago, D. F. Aberle and co-authors (1950) specify eight existential needs of any human society, needs that must be met somehow if the society is to remain viable. They define a society as any selfmaintaining group (from a small tribe to a nation state) with the following characteristics: it occupies a territory, it is capable of existing longer than the lifespan of its members, and its members’ socialized offspring can replace at least some of the offspring who die (1950, 101). Four of these existential needs are especially relevant to the purposes of this section. One such need is the division of labour into an integrated set of social roles necessary to the individual and collective survival and welfare of the society, and that are occupied by people competent and motivated to perform them. The viability of even the simplest societies depends on the performance of a variety of roles, including those entailed in the bearing and rearing of children, the provision of food and shelter, and the management of the society’s affairs. Owing to the interdependence of human beings, the performance of their roles requires the collaboration of occupants of counterpart roles, and the recognition of differences in power between occupants of dominant and subordinate roles. As the sociologist Ian Burkitt points out, the individual’s capacity for agency is dependent on both impersonal interdependencies (for example, the interdependence of government official and citizen) and personal interdependencies (for example, the interdependence of family members) (2016, 332). Secondly, society has an existential need for the normative regulation of how its goals are to be met in role relationships—such goals as the maintenance of the society’s economy, the minimization of conflict, and the control of force and fraud. It is because role relationships are normatively regulated that participants can usually be reasonably confident that their expectations of one another’s conduct in the relationships will be met. Thirdly, a related need is the normative regulation of affect. Members’ affective states need to be mutually comprehensible and communicable. Some of them (such as guilt and shame) are necessary, others (such as lust and rage) need to be controlled. Finally, every society must also meet the existential need of socializing the offspring of members: the conversion of helpless human organisms into persons or selves who can and will perform their various social roles adequately. A commonly accepted explanation of this process, an explanation presupposed by Dewey, is one developed by George Herbert Mead, a colleague of Dewey’s and a fellow founder-pragmatist.

68



3. Mead’s Contribution to Dewey’s Ethics Dewey himself was not a novice in the application of the behavioural and social sciences to ethics. He was already a forerunner in the application of interpersonal psychology to ethics when, in 1891, he appointed Mead to the Department of Philosophy at the University of Michigan. Drawing on Jay Martin’s (2002) biography of Dewey, Nicholas Pagan (2008) notes his early application of interpersonal psychology to ethics in Outlines of a Critical Theory of Ethics (1891) and The Study of Ethics: A Syllabus (1894). Another work, often mentioned in contemporary accounts of his moral philosophy, is Human Nature and Conduct: An Introduction to Social Psychology (1922). According to the Mead scholar Gary Cook, most of Mead’s work can be seen as ‘the detailed exploratory extension of Deweyan pragmatism in areas that Dewey himself mapped out only on a rather large scale’ (Cook 2006, 76). Another wellinformed commentator on both philosophers, Hans Joas, notes that Dewey’s concept of specifically human communication is deeply grounded in his early idealist writings and not due simply to Mead’s influence. Joas adds that while Mead is certainly the more important for those interested in a theory of intersubjectivity, the practical implications of intersubjectivity are ‘much better and more comprehensively elaborated’ in Dewey’s pragmatism (1990, 171). In an earlier publication, Joas notes that there is evidence in students’ lecture notes that Mead drew heavily on Dewey’s ethics and referred students to it for explication of lines of thought to which Mead had made only passing reference (Joas 1985 [1980], 122). On the other hand, as Daniel Huebner explains, while Dewey took the lead in the practical application of pragmatist ideas, he and Mead collaborated closely and both were committed to the application of their ideas to social reform (Huebner 2020). The two philosophers had a high regard for each other. Cook says Mead makes it clear in his publications that he regards Dewey as ‘the pragmatist par excellence’, who saw pragmatism as a philosophical enterprise in which an ‘initial fund of Hegelian ideas is naturalized and put to work as a vehicle for addressing concrete human problems’ (Cook 2006, 76). Referring to Mead’s ‘social interpretation of the world’, Dewey says (in the published version of his address at Mead’s funeral), ‘I know that his ideas on this subject worked a revolution in my own thinking, though I was slow in grasping anything like its full implications’ (Dewey 1931, 313). Despite his contribution to Dewey’s analysis of moral agency, Mead is not mentioned in most mainstream contributions to pragmatist moral philosophy, and his importance to Dewey’s analysis is largely unrecognized by the latter’s exegetists and commentators on his work. One reason is that in his major work on ethics, Dewey does not refer to his use of Mead’s ideas or make his use of them explicit. Another reason is the long-standing reluctance of moral philosophers to consider the sociological presuppositions of their theories. Alasdair MacIntyre

  ’     

69

(2007, 23) complains that the failure of philosophers ‘since Moore’ to spell out the sociological presuppositions of their philosophies is a common shortcoming, and that in consequence it is not clear how their philosophies are, or at least can be, embodied in the real social world. As will become clear in what follows, the rendering of Dewey’s analysis of moral agency as his account of pragmatist role ethics reveals the function and importance of its Meadian presuppositions in explaining its social embodiment.

Mead’s Theory of Self-formation The classic source of Mead’s theorizing is a book published after his death and compiled from lecture notes and unpublished manuscripts: Mind, Self, and Society from the Standpoint of a Social Behaviorist (1934). In this he explains the transformation of the human organism into a person or self as a process in which, by means of verbal expressions and other significant symbols, the mind is colonized by the culture to which it is exposed and becomes imprinted with the community’s shared expectations of conduct in role relationships. In the process, the organism acquires from its interaction with existing selves (care-givers, peers, imaginary peers, players in a game, and so on) the capacity for intersubjective communication with them in role relationships, and, by becoming able to adopt their standpoint, a consciousness of its own unique selfhood. New entrants to society thereby become equipped to judge the situation-specific conduct expected of themselves and others as occupants of the numerous role relationships—parent and child, customer and shopkeeper, lecturer and student, citizen and fellow citizen, and so on—in which they may participate. In short, they become exponents of a culture and able to interact with one another with the predictability that makes social life possible. For Mead, the essential mechanism of human cooperation is the capacity of the participant in a role relationship to, as he put it, ‘take the role of the other’. This taking the rôle of the other, an expression I have so often used, is . . . of importance in the development of co-operative activity. The immediate effect of such rôle-taking lies in the control which the individual is able to exercise over his own response. The control of the action of an individual in a co-operative process can take place in the conduct of the individual himself if he can take the rôle of the other. (Mead 1934, 254)

In becoming a mature self, the individual not only internalizes the attitudes involved in numerous particular role relationships, she also internalizes generalizations of the broader ideas and attitudes of specific groups towards their projects, such as those specific to a university or a company. These generalizations,

70



which Mead described as ‘generalized others’, merge to become the generalized other. This endows the mature self with a worldview that encompasses her own and others’ roles and their motivations, functions, and interrelationships. For Mead, it ‘is in the form of the generalized other . . . that the community exercises control over the conduct of its individual members; for it is in this form that the social process or community enters as a determining factor into the individual’s thinking’ (1934, 155). When interacting with others, the agent attempts to choose and perform her roles in ways that maintain consonance between her own construal of the self she is enacting and her interpretation of its construal by others. But that self is not a single entity. As Mead, Dewey, and another founder-pragmatist, William James, observe, the self is a composite of constituent selves. James maintains that there are as many of these selves—some compatible with one another, others not—as there are different groups of people whose opinion the individual cares about (James 1890, 294). For example, a doctor is likely to treat her patients in a way intended to engender in them a construal of her that validates her self-construal as a competent professional. When interacting with members of another group, she may seek a validation of another of her constituent selves as, say, a fervent Christian. It is the agent’s responsibility to choose and perform her roles and to decide how best to do so in a situation-specific manner. The fact that the self is both a conforming product of society and a decision-maker is recognized by Mead in his distinction between the self as ‘I’ and the self as ‘Me’. The ‘Me’ is the unique self that has been formed by the internalization of the attitudes of the generalized other. It differs from all other selves but has to have the attitudes of the generalized other in order to function as a member of the community. The ‘I’ is the self in decision-making mode. It determines the self ’s response to the attitudes of the ‘Me’, and may change them, for example when choosing between roles that make incompatible demands on her conduct. In changing the attitudes of her own self, and responding unconventionally to the attitudes of others, she may change their attitudes. In Mead’s view, individuals continually change the attitudes of society in such ways (1934, 182).¹ James also distinguishes between the ‘I’ and the ‘Me’, but for him, the ‘I’ is the self which attends to her own thoughts and dispositions, including her moral sensibility and conscience. This innermost self ‘presides over the perception of sensations, and by giving or withholding its assent influences the movements they tend to arouse’ (James 1890, 297–298). Merging Mead’s and James’s views, the constitution of the self at any moment can be seen as the outcome of past ¹ Two social psychologists, Jack Martin and Alex Gillespie (2010), offer a useful ‘neoMeadean’—developmental—account of the stages by which individuals achieve the detachment and reflective intersubjectivity required for the intelligent exercise of agency in their roles.

  ’     

71

dialogues between the ‘I’ and the ‘Me’ on the extent to which the self ’s choice and performance of its social roles should be consistent with, or modify, the dispositions of the existing self. The social psychologists George McCall and J. L. Simmons say that both Mead and James ‘took over Kant’s definition of the “I” as the essentially unknowable active agent of the personality—that which does the thinking, the knowing, the planning, the acting’ (1978 [1966], 52–53). Although Mead’s sociology has been subjected to strong criticism by philosophers and social scientists over the years, most critics have proposed improvements rather than questioned its essential features, and these remain widely endorsed. His sociology became the foundation of a branch of sociology known as symbolic interactionism, the focus of which is on how people’s acts are based on the meanings they attribute to objects and events, meanings that are derived from social interaction and modified by interpretation (Society for the Study of Symbolic Interaction 2017). One of the products to which research in symbolic interaction and related fields has contributed is a measure of the extent to which the individual construes herself as possessing a set of attributes that are culturally recognized as indicative of moral praiseworthiness (Aquino and Reed 2002). One version of the measure has been shown able to predict the extent of the morally praiseworthy choice and performance of role in a range of situations (Boegershausen et al. 2015). Another version has been shown to be able to predict the reported experience of guilt or shame when the individual perceives a mismatch between her own and others’ construal of her role-specific conduct as morally praiseworthy (Stets and Carter 2011; Stets 2015). Among qualitative studies in symbolic interaction, probably the best known is Erving Goffman’s The Presentation of Self in Everyday Life (1959), in which he illustrates how agents attempt to perform their social roles in a way that achieves consonance between their own construal of their role-specific selves and the construal of them that they attribute to observers and occupants of counterpart roles. All these studies support Mead’s theory of self-formation, the theory presupposed by Dewey.

4. Some Implications of Mead’s Role Concept for Moral Philosophy Before turning to Dewey’s analysis, it will be instructive to consider the extension of Mead’s role concept into moral philosophy by two British philosophers in the 1960s, Dorothy Emmet and Bernard Mayo. Emmet made an innovative contribution to the subject with her book, Rules, Roles and Relations (1966). In it she observes that social action depends on the existence of mutual reciprocal expectations about how people are likely to act, ‘and on these expectations not being too often disappointed’ (1966, 7). She explains

72



why the concept of ‘role’ is important to moral philosophers as well as sociologists. The latter use it when describing situations that are structured in part by the prevailing rules of acceptable and unacceptable conduct. Moral philosophers too need to recognize that how people view their roles, and conflicts between their roles, largely determines what they think they ought to do. The concept of ‘role’ says Emmet, ‘may be a bridge notion between myself as an individual, with my proper name and my personal responsibility, and “my station and its duties” in the institutional world of the society in which I have to live’ (Emmet 1966, 15).² Partly in response to a criticism of him by Emmet in her book, Mayo (1968) published an illuminating essay on Mead’s concept of ‘role’ when applied in moral philosophy. The essay clarifies implications of the ‘role’ concept and the ‘I’/‘Me’ distinction that are still not widely grasped.³ Referring to the role of bus conductor as an example, Mayo points out that no set of statements of beliefs and actions can stand as a description of what the individual does as a bus conductor without the qualifiers ‘as’, ‘qua’, ‘in his capacity as’. Moreover, the conductor’s actions in that capacity cannot be understood except in relation to those of the occupants of other, counterpart, roles, such as passenger, driver, manager. ‘The concept of role is irreducibly sociological; it cannot be reduced to elements of individual behaviour or belief ’ (1968, 52). Nor need its use be restricted to familiar relationships, such as those between parent and child or lawyer and client. As Mayo insists, all social relationships are role relationships, even the fleeting relationship between someone helping a fellow passenger avoid slipping from a bus.⁴ Mayo explains that, as an institution, a role shapes the beliefs and conduct of its performers, because a performer’s knowledge of what she is doing involves knowledge of what she is doing it as and the conduct expected of her in that role. She also knows that those in counterpart roles know that she is performing a role and therefore the conduct to expect of her, and they know that she knows that they know, et cetera. The role is not constituted by their expectations. Rather, their expectations are determined by the role they take her to be playing. In sum, a description of someone’s conduct as performing a role is a condensed explanation of it ‘in terms of an open-ended series of expectations by, and of, others’ (1968, 58). This, and the interdependencies of human beings noted in Section 2, mean that such activities as those that distinguish the artist or the vegetarian are also performances of a role. If they were not, their actions would not be intelligible, to ² The book contains a brief summary of Mead’s sociology of self-construction (1966, 156–157) and refers once to John Dewey (29), though only to claim that his influence had promoted the false assumption that critical scientific intelligence and liberal values go together. The phrase ‘my station and its duties’ is the title of a chapter in Ethical Studies (1876), a book by the idealist philosopher F. H. Bradley. ³ Mayo makes no reference to Mead or Dewey. He seems to be well aware of the former’s theory. ⁴ Which is why, contrary to the assumption of some writers, there is no intelligible distinction to be made between ‘role morality’ and ‘ordinary morality’. Role morality is ordinary morality.

  ’     

73

themselves or anyone else. Sometimes someone’s actions are unintelligible to observers precisely because it is not clear what role they are performing. Is that person with the screaming child a parent or an abductor? Is that man on a ladder a window cleaner or a burglar? Actions are also puzzling when roles are encountered for the first time, for example when they are the roles of participants in an unfamiliar religious ceremony or an unfamiliar game, or when they are newly created and are in the process of being institutionalized. To support his claim that moral imperatives are role requirements, Mayo argues that when such motives as ‘on principle’, ‘as a matter of right or wrong’, or ‘from a sense of duty’ are used to interpret an agent’s conduct, the interpretation is of what the agent recognizes herself to be doing something as: it is an interpretation of her understanding of the role she is performing. ‘If I do Y as a matter of moral principle, then so far as witnesses are concerned I expect them not merely to note my doing Y, but to note it—again, to take it—as done in fulfilment of expectations, and moreover in the expectation of recognition as having fulfilled expectations, the expectations being shared between me and them’ (1968, 60, emphases in original). It is essential to distinguish between the consensus of mutual expectations that constitute the role as a social institution and how it is understood in the moral thinking of the individual (1968, 62). How the agent (Mead’s ‘I’) understands the consensus may lead her to resist the demands of a role as a social institution, and also determines how she chooses between those roles that make conflicting demands. As will become clear in the next section, the capacity and willingness to make such choices is seen by Dewey as what distinguishes reflective from conformist attitudes to customary morality, a capacity essential to moral progress.

5. Dewey’s Ethics as Role Ethics The principal sources drawn on here to show that Dewey is a role theorist are two of his later works: the chapters for which he was responsible in the second edition of Ethics (1932), of which he was the co-author with James Hayden Tufts, and the article ‘Three Independent Factors in Morals’ (1930). According to Dewey specialists, these two works best represent his treatment of ethics in his later works (Pappas 2008, 5, 310 nn. 7 and 8). The 1932 book will be cited as E in what follows. Dewey barely used the term ‘role’ and he did not adopt a specifically rolecentred standpoint for his analysis of moral agency, but that standpoint is clearly immanent in the analysis. He also made several statements which show the fundamental importance of role relationships to his ethics. Explaining what motivates people to acknowledge obligations to others, even when doing so is irksome, he points out that obligations are inherent in the relationships people sustain with one another. As examples he refers to the parental relation that

74



engenders the parents’ obligations to their child, the friendship relation that engenders the obligation that friends have to one another, and the relation of citizens to a just state that explains why they respond to its demands. ‘If we generalize such instances, we reach the conclusion that Right, law, duty, arise from the relations which human beings intimately sustain to one another, and that their authoritative force springs from the very nature of the relation that binds people together’ (E, 237–238, capitalization in original). Later he points out that many of the duties recognized in common law arose out of recurring economic relations between people, such as those between landlord and tenant, vendor and purchaser. Such duties, he says, were what the Romans called offices. ‘It is as a parent, not just as an isolated individual, that a man or woman imposes obligations on children; these grow out of the office or function the parent sustains, not out of mere personal will’ (E, 249). Similarly, the holder of a public office, such as legislator, does not exercise authority as a private possession, but as the representative of an office in a relationship in which many participate. ‘In principle, therefore, Right expresses the way in which the good of a number of persons, held together by intrinsic ties, becomes efficacious in the regulation of the members of a community’ (E, 249). There can be little doubt then that for Dewey, offices are what Mead, and contemporary sociologists, describe as social roles, and that Dewey’s ethics is grounded in the recognition that obligations arise naturally in, and are constitutive of, the role relationships that obtain between members of a society, including those that are necessary to its existence. As Dewey himself points out, the facts on which morality depends ‘are those which arise out of active connections of human beings with one another, the consequences of their mutually intertwined activities in the life of desire, belief, judgement, satisfaction, and dissatisfaction’ (Dewey 1922, 329). In sum, although Dewey does not use its terminology, role ethics is the foundation of his analysis of moral agency and is immanent in his account of it. This fact, which has been ignored by his exegetists, is the justification of the version of his account that follows, a version which makes clear his presupposition of Mead’s theory of self-formation as the colonization of the mind by the culture to which it is exposed, including its shared expectations of conduct in role relationships. The account of Dewey’s analysis as his account of pragmatist role ethics is faithful to his own but clarifies the analysis and extends its implications.

The Agent as Enactor of Self in Social Roles For Dewey, the agent’s character is the self ’s set of stable dispositions to respond to similar stimuli in a similar way (E, 181). These dispositions are the underlying causes of acts and of their resemblance to one another. But dispositions can be

  ’     

75

stable without being consistent with one another, and in fact, ‘Divided and inconsistent interest is common’ (E, 281). He argues in an earlier book that it is a mistake to suppose that there is one ready-made self behind activities. Except when it is immobilized by routine activity, the self is always in a process of being produced and can include a number of inconsistent selves that are unharmonized in their dispositions. ‘There are complex, unstable, opposing attitudes, habits, impulses which gradually come to terms with one another, and assume a certain consistency of figuration, even though only by means of a distribution of inconsistencies which keeps them in water-tight compartments, giving them separate turns or tricks in action’ (Dewey 1922, 137–138). Note the similarity here of Dewey’s conception to James’s proposal (described above in Section 3) that the self is a combination of role-specific or group-specific selves, each of which responds appropriately to the different expectations of those with whom the agent interacts in the different social groups to which she and they belong.⁵ Also relevant here is Mead’s explanation (also described in Section 3) of the agent’s motivation to display different versions of the self: the desire to choose and perform roles in ways that maintain consonance between a preferred construal of own self (or the one to which the agent aspires), and the perception of how the realization of that self is construed by occupants of counterpart roles. While Mead’s ‘I’/‘Me’ distinction is not used explicitly by Dewey, it is implicit in his conception of the agent as the possessor and arbiter of a multi-versioned self that is formed and can be changed by the agent’s choice of acts and their consequences. When it is awake, he says, the self is always active. The function of a stimulus to the self from the environment is not to initiate activity, but to alter it in response to whatever a change in the environment signifies for the self ’s dispositions. This means that when acts have moral significance, they are realizations of the self ’s disposition to observe or neglect moral obligations. Thus, Dewey asserts, ‘It is not too much to say that the key to a correct theory of morality is recognition of the essential unity of the self and its acts, if the latter have any moral significance; while errors in theory arise as soon as the self and acts (and their consequences) are separated from each other, and moral worth is attributed to one more than to the other’ (E, 318–319, emphasis in original). Those whose self-construal places a high value on acting benevolently and the avoidance of cruelty will act accordingly. ‘Benevolence or cruelty is not something which a man has, as he may have dollars in his pocket-book; it is something which he is, and since his being is active, these qualities are modes of activity, not forces which produce them’ (E, 322, emphases in original). The unity of self and acts is implicitly recognized, says Dewey, when we hold someone responsible for her

⁵ Alasdair MacIntyre (1999) takes a similar view in an article in which he complains about the malign effects of the compartmentalization of roles in modern societies.

76



conduct, for to do so is to acknowledge the intimate and internal connection between an act and the actor’s self or character. The unity of the self and its acts entails the unity of the self and its enactment in its choice and performance of social roles. It is a choice that will entail the enactment of a role-specific self, though the choice and performance will be influenced by other constituent selves and their relative importance to the self as a whole. For example, role-specific selves as woman or man or as citizen often have pervasive effects on the choice of other roles and performance in them. The choice will also depend on the particular circumstances in which it is made, including the agent’s interpretation of the normative expectations embodied in the chosen role and her perception of the expectations of occupants of counterpart roles. In making their choices, agents may be governed by customary or by reflective morality. Dewey stresses the importance of this distinction. Customary morality locates its rules and standards of conduct in the community’s established practices, whereas reflective morality ‘appeals to conscience, reason, or to some principle which includes thought’ (E, 171). It is not a distinction between ancient and modern moralities, he notes, since some degree of reflective thought must have been present since earliest times, while in modern societies, conduct is often governed by convention, even when the need for critical judgement is strongest.

The Agent’s Response to Moral Perplexity Dewey points out that the morally responsible agent’s preferred self is revealed by the way she deals with occurrences of moral perplexity if these have consequences for her future self. The perplexity may be posed by the need to make a choice between roles that make conflicting demands or between different ways of responding to the conduct or needs of an occupant of a counterpart role. Moral perplexity may also occur when the habits of conduct that comprise the customary performance of a social role are made inadequate by new circumstances or by a change in how existing circumstances are construed. Whole communities or groups can find themselves faced by novel opportunities and requirements that challenge habits and beliefs formed in the past (E, 175). For example, in recent decades high proportions of women in modern societies have encountered the need to choose between childbearing and career opportunities. When deliberating over such choices, each alternative may appeal to a different element in the constitution of the self. Thus, while moral deliberation over a choice is, superficially, a weighing of the values of different ends, ‘Below the surface, it is a process of discovering what sort of being a person most wants to become’ (E, 317). Deliberation is initiated by the individual’s affective response to situations and acts and must end with a conclusion that is experienced affectively if it is to stir the agent to act on it. Dewey echoes David Hume and Adam Smith in maintaining

  ’     

77

that it is sympathy which evokes these affective responses and that prompts deliberation and action, the sympathy that functions by ‘rendering vivid the interests of others and urging us to give them the same weight as those which touch our own honor, purse, and power’ (E, 298). But while the valuing of acts that accompanies the immediate response to them determines their importance to the agent, it is an error to suppose that this immediate response excludes the need for reflection, when in fact, ‘The direct valuing which accompanies immediate sensitive responsiveness to acts has its complement and expansion in valuations which are deliberate, reflective’ (E, 299). In reaching a judgement, the agent must engage in moral theorizing. This is not the struggle between desire and moral belief that is evoked by the temptation to do something wrong (E, 174). Moral theorizing is the reflective process in which an agent must engage when the problem is that of deciding what one ought to do in morally perplexing circumstances. Dewey sees no fundamental difference between systematic moral theory and the reflection that an individual engages in when attempting to find general principles to direct and justify her conduct. ‘For what is called moral theory is but a more conscious and systematic raising of the question which occupies the mind of anyone who in the face of moral conflict and doubt seeks a way out through reflection’ (E, 173). Systematic moral theory may help the individual to clarify the difficulty that such problems pose and to think about them more systematically than she might do otherwise, but it cannot offer a practical solution, ‘it cannot take the place of [a] personal decision, which must be made in every case of moral perplexity’ (E, 176). Dewey rejects as untenable the traditional idea that the aim of moral philosophy is to establish some ultimate principle as a formula for moral judgement, a formula independent of history, culture, material conditions, and individual circumstances. To adopt such an aim is to take what he describes as ‘the spectator view’ of the task of moral philosophy. Instead, for him the proper starting point is the correct understanding of human experience from the agent’s standpoint. Dewey’s own understanding of it, one grounded in a Darwinian conception of the nature of humankind, sees experience as everything that occurs in the continuing intercourse of the living being with the continually changing physical and social environment on which it is dependent and with which it continually seeks successful adjustment (Dewey 1998 [1917], 48]). He argues that moral theorizing is subject to the influence of not one, but three different sources of normativity—sources of influence on the agent’s view on what action ought to be taken.⁶ These are sources of influence on the agent’s choice of role and its enactment. The proximate source in each case is the agent’s consciousness of the criterion of morally praiseworthy action that is distinctive of one ⁶ Dewey does not use the term ‘sources of normativity’. On one occasion he describes them as ‘independent variables in moral action’ (1930, 199).

78



of the types of traditional moral theory, types that are usually considered to offer rival conceptions of this criterion (E, 193–196). While each of the three sources of normativity is soundly based, the fact that they differ in their origin and mode of operation means that they can exercise divergent influences on the formation of judgement. In consequence, ‘it is characteristic of any situation properly called moral that one is ignorant of the end and of good consequences, of the right and just approach, of the direction of virtuous conduct, and one must search for them’ (Dewey 1930, 199). In place of an ultimate end as a criterion by which the agent’s response to the influence of each of the sources should be assessed, he offers a criterion of the moral praiseworthiness of the agent’s intentions in relation to each of them. A description of the three sources and their attendant assessment criteria follows.

Desires, Intentions, and the Good One of the three sources of normativity specified by Dewey is a conception of the Good: the desired or intended objective of the choice and performance of a role. Here the task of moral theory is to help the responsible moral agent identify objectives that are consonant with the welfare of others. He pays particular attention to utilitarianism. With Jeremy Bentham’s version of this doctrine in mind, he notes that if the happiness of the majority were the supreme moral principle, even an act that was morally repugnant to the agent’s character would be good if its overall effect was considered benign. That, he points out, would be to ignore the fact that it is the agent’s character which determines the extent to which her actions are directed towards the wellbeing of others. It is a different matter when it is the effect of laws and institutions on the general welfare that is being estimated. For that purpose, the precise consideration of consequences is required. Dewey is more positive about John Stuart Mill’s version of utilitarianism. His version maintains that laws and social arrangements should attempt to achieve as close a correspondence as possible between the happiness of the individual and the interests of society as a whole, and that establishing the same association in the mind of the individual should be the aim of formal and informal education. In Dewey’s view, making the wellbeing of society the end or purpose of desire would not guide conduct but merely arouse a diffused sentimental state (E, 270). The criterion of the moral praiseworthiness of the objects of desire should be a characteristic of the desire itself. His proposal is that the objective of the chosen role and its performance should be one ‘which at the same time brings satisfaction to others, or which at least harmonizes with their well-being in that it does not inflict suffering upon them’ (E, 271). To identify objects that comply with the criterion, the reflective agent needs to adopt the standpoint of an ideal observer and consider proposed objects ‘through the eyes of this impartial and far-seeing

  ’     

79

objective judge’ (E, 269). Here he has in mind the concept of ‘impartial spectator’ proposed by Adam Smith (1976 [1759], 113]), whose influence he acknowledges on several occasions in Ethics.

Role Obligations and the Right Another source of normativity is a concept of the Right that should govern the intentions and ends of the agent’s choice and performance of roles. The Right is ‘that which accords with law and the commands of duty’, and which may, according to its advocates, be threatened by desires (E, 233). Kant is the bestknown exponent of the type of traditional theory in which this concept is central. For Dewey, a concept of the Right that requires the suppression of desires and intentions inherent in the nature of human beings is untenable. Kant’s identification of the Good Will as the only moral good implies that the self, independently of the consequences of its acts, is the supreme and exclusive moral end. This is to treat acts and their consequences as mere means for maintaining the good self (E, 315–316). Thus, however bad the consequences of an act, if it maintains the goodness of the self, it is good. What is required, says Dewey, is a concept that preserves the distinction between the Good and the Right without such implications. As noted in Section 2, he finds it in his understanding of the nature of role relationships, which leads him to the conclusion that ‘Right, law, duty arise from the relations which human beings intimately sustain to one another, and that their authoritative force springs from the very nature of the relation that binds people together’ (E, 237–238). The obligations should be seen, not as Kant saw them—a product of reason—but as the claims made by society on behalf of its members on how they ought to behave in their relationships with one another. The ultimate source of the authority of these obligations is the socially shared conviction of their rightness. The proximate source is that conviction in the consciousness of the agent. That is why, for example, a shoplifter is censured: in failing to pay for her purchases, she has failed to observe the shared expectations of right conduct that govern the role relationship between customer and shopkeeper and that are essential to its effectiveness. But role obligations should always be seen as subject to revision or rejection, in light of enquiry and reflection. Dewey reminds readers that ‘Some persons persecuted as moral rebels in one period have been hailed as moral heroes later’ (E, 251). Those who reject a customary view of a particular role obligation may have a duty to not observe it, but they need to show that an alternative would better serve the common good (E, 252). Dewey’s criterion of the rightfulness of the agent’s choice and performance of a role is whether the obligations embedded in it are regarded by the agent as serving the common good and are therefore as binding on her own judgement and action

80



as it is on those from whose observance of it she benefits (E, 250–251). This criterion would not be met, for example, by an agent who benefits from but does not observe the obligation embedded in the role of citizen to avoid causing undeserved harm to others. The wrongdoer ‘contradicts, not as Kant would have it, some abstract law of reason, but the principle of reciprocity when he refuses to extend to others the goods which he seeks for himself ’ (E, 251–252).

Praise, Blame, and Virtue The third source of normativity is the desire for conduct to be found praiseworthy rather than blameworthy by others for the virtuous dispositions revealed by the agent’s choice and performance of roles—‘the dispositions which are socially commended and encouraged constituting the excellencies of character which are to be cultivated; vices and defects being those traits which are condemned’ (E, 195). Because praise and blame are spontaneous and instinctive, they differ in principle from the good and the right as a source of normativity. But their effect may be to endorse judgements of rightness or influence deliberation about ends (Dewey 1930, 203). From the standpoint of conventional morality, traits counting as virtues include kindness, honesty, temperance, and many others. Dewey points out that what these virtues mean in practice is defined by socially approved forms of conduct in the roles in which the putative virtues are realized, forms which vary between societies and change over time. Such virtues as public-spiritedness, regard for life, faithfulness to others, and fairness have a permanent value because no community could survive in which they were not approved. But the view of what constitutes a realization of these virtues in conduct differs between communities and between different periods. In some communities, even regard for human life does not extend to all female infants or the aged (E, 281). There are also of course major variations in the conduct described as showing fairness. Further, when they are treated as separate from one another, the traits described as virtues have variable meanings. For example, when treated as separate virtues, courage may be seen as merely stoical resistance, and conscientiousness an obsession to avoid falling into error. When it is partial, the virtue of concern for others may refer to a concern restricted to the interests of friends and relatives, rather than a disposition to treat the interests of occupants of all counterpart roles, not with the same intensity, but with an equal and even measure of value, whether they are ‘friends or strangers, fellow citizens or foreigners’ (E, 282–283). Because a virtue cannot be defined by a permanent and uniform feature of conduct, Dewey proposes that it should be defined by whether the agent’s enactment of a role in pursuit of a morally praiseworthy interest reveals a disposition with

  ’     

81

the following three qualities: whole-heartedness (total commitment), continuity and persistence, and impartiality. It is dispositions with these qualities that are the virtues which ought to be praised. When virtues are defined in this way, they ‘interpenetrate one another’ and ‘this unity is involved in the very idea of integrity of character’ (E, 283).

Conflicting Influences on Moral Judgement Dewey maintains that because each of the sources of normativity represented by the traditional moral theories has a function in moral theorizing, uncertainty and conflicting ideas and the need for individual choice are inherent features of moral judgement. In an address to France’s Société Française de Philosophie in 1930, he describes the difference between his view of the traditional theories and the views taken by their advocates, and the implications of the difference for moral judgement. The traditional theories postulate a single principle as the explanation of moral life, and exclude the possibility of uncertainty or conflict when considering the moral quality of a proposed act. According to these theories, ‘Conflict is, in effect, between good and evil, justice and injustice, duty and caprice, virtue and vice, and is not an inherent part of the good, the obligatory, and the virtuous’ (1930, 199). In fact, argues Dewey, the sources of normativity that are the subjects of the theories, while independent in their origins, are intertwined in their operation and often divergent in their influence. He summarizes their different origins as follows. ‘Good, I have said, is ascribable to deliberation about desires and intentions; right, obligation, is dependent upon requirements that have social authority and force; virtues are dependent on approbation’ (1930, 203). Dewey concedes that traditional theories also touch on all three sources, but maintains that each takes one of them as central and treats the others as secondary (1930, 209). The three sources of influence on conduct can converge in their effects. An example would be when in their role as citizens, agents act on the socially approved right and obligation to protest against an injustice, regard doing so as contributing to an outcome that has good consequences, and know that the exercise of the right is considered praiseworthy and virtuous by respected others. But it is where these different influences are intertwined and cut across one another, as they often do, says Dewey, that problems arise. As an example of what he had in mind, consider the dilemma of Huckleberry Finn (HF) when faced with a choice between two role-specific selves: that of a self who, because he has assimilated the worldview of a slave-owning society, believes that he should surrender an escaped slave, Jim, to the authorities, and that of a self, who as a friend of Jim, believes he should protect him. Although HF decides to protect Jim, he feels bad about having done so, but acknowledges that if he had betrayed his

82



friend, he would feel just as bad (Twain 1884, 156). For him, Good and Right pulled in opposite directions. Jim’s poignant praise of HF’s virtue as a friend seemed to tip the balance.⁷ In Dewey’s view it is absurd to ignore such competing pressures. He proposes that it would be better to recognize that no single formula can resolve all moral difficulties, and to admit that human beings can only do their best to accommodate the disparate sources of normativity when deciding how to respond to them in a particular situation. This would not be a loss, he says. ‘By turning our attention from rules and rigid standards, it would lead us to take fuller consideration of the specific elements which necessarily enter into every situation where we must act’ (1930, 204).

Social Conditions and Moral Agency Dewey was well aware that moral deliberations and the capacity to act on them are heavily influenced by the agent’s social and material environment. Gil Richard Musolf (2013) describes him as the progenitor of a variety of contemporary contributions to a literature that have ‘not only illustrated domination in a bureaucratic, corporate, capitalist world—one saturated in racism, sexism, and classism—but have also suggested strategies to reduce domination without succumbing to utopian grandiosity’ (Musolf 2013, 87). Dewey believed that domination by the ruling groups of the capitalist system reduced workers to a slave-like condition, while at the same time manipulating them to see the rulers’ interests as their own. In other words, as occupants of the dominant roles, members of the ruling groups act to ensure that their interests prevail in their relationships with occupants of the subordinate roles. Musolf quotes Dewey’s assertion that the master–slave relationship exists ‘in any situation which effectively places one person in subjugation to another—children subject to their parents, wives to their husbands, subjects to their rulers, laborers to their employers’.⁸ In this and other respects, Dewey’s views were similar to those of Marx (Musolf 2013, 101). If he were writing now, Dewey would find little reason to change the claim he made in 1932 that almost all ethical problems are the result of such developments as the invasion of domestic life by new modes of industry, the control of opportunities for finding work by concentrations of capital, and the modifications of social conditions by new inventions. These developments have pervasive effects on the social roles available in a society and the relationships of their occupants with occupants of counterpart roles. A common outcome has been widening ⁷ Although it is vividly portrayed in Twain’s text, HF’s struggle and its significance are commonly ignored by commentators. ⁸ The quotation is from Dewey’s Lectures in China 1919–1920.

  ’     

83

differences in power and wealth between occupants of dominant and occupants of subordinate roles.⁹ There have been related changes in the character of role occupants, because ‘social conditions enter integrally and intrinsically into the formation of character, that is, the make-up of desires, purposes, judgments of approval and disapproval’ (E, 380). The commercial activities of capitalist society, with their emphasis on ruthless competition for private gain, permeate the lives of most people and often undermine the appeal of their own convictions (E, 286). Dewey observes that in democratic societies, reflection on the social conditions that people favour and disfavour is practical and ‘morally indispensable’ (E, 353). He championed the concept of a society that is able and willing to adopt an experimental attitude to itself, one in which there is a general willingness to reexamine current convictions about any aspect of social life in light of its consequences for the wellbeing of those affected, ‘even if that course entails the effort to change by concerted effort existing institutions, and to direct existing tendencies to new ends’ (E, 366, emphasis in original). It would need to be a genuinely democratic society in which members had been educated and socialized in ways that would realize its potential, in their personal lives and when deciding which social policies to support.

6. Conclusion In the view of distinguished advocates of his work, John Dewey’s analysis of moral agency is a fitting replacement for traditional alternatives, and rivals the best of those derived from them. Yet that estimate of the value of his work is uncommon, to judge from the comparative rarity of allusions to his ideas in mainstream discourses on moral theory. One reason for this is that Dewey’s own presentation of his analysis and most presentations of it by his exegetists do not make sufficiently clear its advantages over rivals that lack its resources and the potency of its naturalism, its grounding in the view that the social world is part of the natural world. As the contemporary pragmatist philosopher Joseph Margolis observes, the naturalistic resources of pragmatism make it well placed to counter the claims of its rivals (Margolis 2006, 9). Arguably, the role-centred account of Dewey’s analysis makes these advantages very clear, including those over other theories of role ethics. Section 4 shows that a tenable analysis of moral agency needs to acknowledge the necessary interdependence of sociology and moral philosophy. By making explicit the contributions of George Herbert Mead and William James, the rolecentred rendering of Dewey’s analysis shows that it is functionally dependent on ⁹ As Lonnie Athens notes, the same people or members of the same group often occupy the dominant roles in multiple role relationships (2012, 436–438).

84



the sociology it presupposes. As well as clarifying its advantages over traditional theories and their derivatives, the new rendering may make his theory an attractive foundation for future studies of the ethical features of social roles, and perhaps the reconsideration of some previous studies. Such developments and the intuitive plausibility of Dewey’s analysis in its role-centred form may help role ethics to become seen as a major focus of moral philosophy instead of one of its specialist domains.¹⁰

References Aberle, D. F., A. K. Cohen, A. K. Davis, M. J. Levy, Jr., and F. X. Sutton. 1950. ‘The Functional Prerequisites of a Society’. Ethics 60 (2): 100–111. Aquino, Karl and Americus Reed. 2002. ‘The Self-importance of Moral Identity’. Journal of Personality and Social Psychology 83 (6): 1423–1440. Athens, Lonnie. 2012. ‘Mead’s Analysis of Social Conflict: A Radical Interactionist’s Critique’. The American Sociologist 43 (4): 428–447. Boegershausen, Johannes, Karl Aquino, and Americus Reed. 2015. ‘Moral Identity’. Current Opinion in Psychology 6: 162–166. Bradley, F. H. 1876. Ethical Studies. First edition. New York: Stechert & Co. Burkitt, Ian. 2016. ‘Relational Agency: Relational Sociology, Agency and Interaction’. European Journal of Social Theory 19 (3): 322–339. Cook, Gary A. 2006. ‘George Herbert Mead’. In A Companion to Pragmatism. John R. Shook and Joseph Margolis (eds.), 67–78. Oxford: Blackwell. Dewey, John. 1891. Outlines of a Critical Theory of Ethics. Ann Arbor, Mich.: Register Publishing Company. Dewey, John. 1894. The Study of Ethics: A Syllabus. Ann Arbor, Mich.: The Inland Press. Dewey, John. 1922. Human Nature and Conduct: An Introduction to Social Psychology. Modern Library Editions. New York: Random House. Dewey, John. 1930. ‘Three Independent Factors in Morals’. Educational Theory 16 (3): 197–209. Dewey, John. 1931. ‘George Herbert Mead’. The Journal of Philosophy 28 (12): 309–314. Dewey, John. 1998 [1917]. ‘The Need for a Recovery of Philosophy’. Republished in The Essential Dewey. Volume 1: Pragmatism, Education, Democracy. Larry A. Hickman and Thomas M. Alexander (eds), 46–70. Bloomington and Indianapolis: Indiana University Press.

¹⁰ I thank the following for helpful comments on earlier drafts: Sean Cordell, Hallvard Lillehammer, and Alan Ryan. The chapter draws on Simons (2021).

  ’     

85

Dewey, John and James H. Tufts. 1932. Ethics. Second edition. New York: Henry Holt. Emmet, Dorothy. 1966. Rules, Roles and Relations. New York: St. Martin’s Press. Goffman, Erving. 1959. The Presentation of Self in Everyday Life. London: Penguin Books. Huebner, Daniel. 2020. ‘Mead, Dewey, and Their Influence in the Social Sciences’. In The Oxford Handbook of Dewey. Stephen Fesmire (ed.), New York: Oxford University Press. James, William. 1890. The Principles of Psychology. Volume 1. London: Macmillan. Joas, Hans. 1985. G. H. Mead: A Contemporary Re-examination of His Thought. Raymond Meyer (trans.), Cambridge: Polity. Originally published in 1980 as Praktische Intersubjekivivitat. Die Entwicklung des Werkes von George Herbert Mead. Suhrkamp-Verlag. Joas, Hans. 1990. ‘The Creativity of Action and the Intersubjectivity of Reason: Mead’s Pragmatism and Social Theory’. Transactions of the Charles S. Peirce Society 26 (2): 165–194. McCall, George and J. L. Simmons. 1978 [1966]. Identities and Interactions: An Examination of Human Associations in Everyday Life. Revised edition. New York: The Free Press. MacIntyre, Alasdair. 1999. ‘Social Structures and Their Threats to Moral Agency’. Philosophy 74 (289): 311–329. MacIntyre, Alasdair. 2007. After Virtue: A Study in Moral Theory. Third edition. Notre Dame, Ind.: University of Notre Dame Press. Margolis, Joseph. 2006. ‘Introduction: Pragmatism, Retrospective, and Prospective’. In A Companion to Pragmatism. John R. Shook and Joseph Margolis (eds.), Oxford: Blackwell. Martin, Jack and Alex Gillespie. 2010. ‘A Neo-Meadian Approach to Human Agency: Relating the Social and the Psychological in the Ontogenesis of PerspectiveCoordinating Persons’. Integrative Psychological & Behavioral Science 44 (3): 252–272. Martin, Jay. 2002. The Education of John Dewey. New York: Columbia University Press. Mayo, Bernard. 1968. ‘The Moral Agent’. In The Human Agent: Royal Institute of Philosophy Lectures Volume 1. London: Macmillan. Mead, George Herbert. 1934. Mind, Self, and Society: From the Standpoint of a Social Behaviorist. Charles W. Morris (ed.), Chicago: University of Chicago Press. Musolf, Gil Richard. 2013. ‘Radical Interaction on the Rise’. In Lonnie Athens and Norman K. Denzin (eds.), Studies in Symbolic Interaction 41: 83–121. Bingley: Emerald Publishing Limited. Pagan, Nicholas O. 2008. ‘Configuring the Moral Self: Aristotle and Dewey’. Foundations of Science 13: 239–250. Pappas, Gregory Fernando. 2008. John Dewey’s Ethics: Democracy as Experience. Bloomington, Ind.: Indiana University Press.

86



Simons, John. 2021. Moral Agency Analysed as Self-Enactment in Social Roles: A Productive Recast of Dewey’s Pragmatist Analysis. PhD thesis, University of London. Smith, Adam. 1976 [1759]. The Theory of Moral Sentiments. D.D. Raphael and A.L. Macfie (eds.), Indianapolis: Liberty Fund. Stets, Jan. 2015. ‘Understanding the Moral Person: Identity, Behavior, and Emotion’. Topoi 34 (2): 441–452. Stets, Jan E. and Michael J. Carter. 2011. ‘The Moral Self: Applying Identity Theory’. Social Psychology Quarterly 74 (2): 192–215. Society for the Study of Symbolic Interaction. https://sites.google.com/site/ sssinteraction/ Accessed June 2019. Twain, Mark. 1884. The Adventures of Huckleberry Finn. Reprinted by Planet PDF, . Last accessed July 2019.

5 A Good Doctor but a Bad Person? A Puzzle for Role Ethics from Løgstrup Robert Stern

1. Introduction In this chapter, I want to consider the following challenge to social role ethics: 1. Role ethics holds (a) that you can be a good occupant of a role by fulfilling the obligations associated with the role, and (b) that this is sufficient to make you a good person (at least to some degree) provided that the institution itself is a good one 2. But a person can count as having fulfilled their role obligations while being a moral monster 3. So role ethics is not an ethics. I will show how this challenge is suggested by what K. E. Løgstrup says concerning our social norms and roles, but also how he might provide the resources with which to meet it. I will first present the challenge in an intuitive way (§2), before considering how Løgstrup offers a plausible account of our social roles out of which this problem then arises (§3); but I will then consider whether he can also offer a way for the role ethicist to escape the problem (§4). In the final section (§5), I will look at the wider implications of Løgstrup’s view for some of the standard issues in role ethics.¹

2. A Challenge to Role Ethics Let me begin with the first premise in the argument above, in order to explain what it is claiming and why it might seem plausible.

¹ The relation between Løgstrup and role ethics has been largely confined to consideration of his possible contribution to nursing practice, which will not be my focus here. See Martinsen 1993; 2006; and 2012. Robert Stern, A Good Doctor but a Bad Person? A Puzzle for Role Ethics from Løgstrup In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Robert Stern 2023. DOI: 10.1093/oso/9780192843562.003.0005

88



As its name suggests, role ethics conceives of itself as an ethics, which distinguishes itself from ‘the big three’ of Kantianism, utilitarianism, and virtue ethics by focusing on the way in which being guided by our roles can lead to the good life, in both senses of that word ‘good’: namely, a flourishing life for the individual, and a life that also involves being a virtuous or moral person who does the right thing for the right reasons, where these two forms of goodness might be said to be connected in an Aristotelian manner. Within the history of recent Western philosophy, role ethics is most obviously associated with the Hegelian tradition,² with roots in Hegel’s notion of Sittlichkeit or ‘ethical life’ which then influences the British Idealists, particularly F. H. Bradley.³ These thinkers themselves refer the idea back to the Greeks, while there is doubtless an influence on Hegel from the Lutheran tradition of ‘creation ordinances’ which structures society into the three spheres of family, church, and state, with their appropriate roles and responsibilities.⁴ For both Hegel and Bradley, it is clear that a major part of the attractiveness of this approach is that it avoids the formalism and abstractness of the Kantian emphasis on a single ‘supreme principle of morality’, an approach which is said to be flawed because it takes us too far away from the concrete ethical relations of social life: without these relations, we have too little content to go on, while at the same time the principle of universalizability which Kant offers us instead is too formal and empty to provide any basis on which to ground these relations.⁵ The general lesson that Hegelian ethics therefore takes from the empty formalism objection to Kantian ethics is that ethical life should be treated more like a concrete social practice, rather than attempting to derive it from scratch based on some supreme principle such as universalizability: from inside this practice, we find ourselves embedded within various social norms, and we should reason ethically on the basis of those norms, rather than try to step outside them entirely. If we do try to step outside them, we will lose our ethical bearings altogether, like ² Outside Western philosophy, role ethics is often associated with Confucianism: for some helpful discussion, see Stalnaker 2020; Cottine 2020; and Hu, this volume. ³ For Hegel, see Hegel 1991, Part Three, and for Bradley, see Bradley 1927, Essay V—though it should be noted that this chapter on ‘My Station and Its Duties’ is not the culmination of the discussion, and is followed by a chapter on ‘Ideal Ethics’ which importantly qualifies some of the claims made in the earlier chapter based on problems raised in pp. 202–206. For some general discussion which connects Hegel to Bradley on a variety of ethical issues, see Mander 2016. ⁴ For further discussion, see Bayer 1998 and Wingren 1957. This tradition is reflected in the Book of Common Prayer catechism ‘what is thy duty towards thy Neighbour?’, which ends: ‘to do my duty in that state of life, unto which it shall please God to call me’. Various commentators have plausibly suggested that this could be the source of Bradley’s famous phrase ‘my station and its duties’: see e.g. Hardimon 1994, 338 n. 9; Chappell 2020, 205. Cf. also Hegel 1991, §150, 193: ‘In an ethical community . . . in order to be virtuous [a person] must simply do what is presented, expressly stated, and known to him within his situation [Verhältnissen]’. ⁵ Hegel’s critique of Kant occurs in four main places: Hegel 1999; Hegel 1977, §§429–437, pp. 256–262; Hegel 1991, §§133–136, pp. 161–164; Hegel 1995 vol III, 458–461; it also occurs briefly Hegel 2010, §§53–54, pp. 102–103. Bradley’s critique can be found in Bradley 1927, Essay IV. For further discussion of this issue, which has been very extensive, see e.g. Wood 1989; Hoy 1989; Freyenhagen 2011. I consider the issue further in Stern 2015, 139–156.

       ?

89

the enquirer who begins by first setting aside all their beliefs, or (to use the famous Neurathian image) the sailor who seeks to rebuild his ship at sea by throwing away all the planks. Thus, as Bradley puts it: ‘There is here no need to ask and by some scientific process find out what is moral, for morality exists all round us, and faces us, if need be, with a categorical imperative, while it surrounds us on the other side with an atmosphere of love’ (Bradley 1927, 187 (my emphasis)).⁶ And this is then where roles come in, because ethical life is partly constituted by various roles in which we find ourselves, as children, parents, employers, employees, doctors, teachers, politicians, and so on, and these roles then have associated obligations and responsibilities which are taken to govern them. Thus, as Bradley puts it in articulating the key role ethics idea of ‘my station and its duties’: To know what a man is (as we have seen) you must not take him in isolation. He is one of a people, he was born in a family, he lives in a certain society, in a certain state. What he has to do depends on what his place is, what his function is, and that all comes from his station in the organism. Are there then such organisms in which he lives, and if so, what is their nature? Here we come to questions which must be answered in full by any complete system of Ethics, but which we can not enter on. We must content ourselves by pointing out that there are such facts as the family, then in a middle position a man’s own profession and society, and, over all, the larger community of the state. Leaving out of sight the question of a society wider than the state, we must say that a man’s life with its moral duties is in the main filled up by his station in that system of wholes which the state is, and that this, partly by its laws and institutions, and still more by its spirit, gives him the life which he does live and ought to live. (Bradley 1927, 173–174; my emphasis)

Putting a human being in a social context therefore makes clear the way in which we are all of us already bound together through roles, some of which we can enter into voluntarily (such as employment) and some of which we cannot (such as the child’s relation to their parent, or a citizen’s relation to the country in which they are born). On this account, then, acting as required within one’s role promises to explain both how the individual can flourish (by providing the social context that makes this flourishing possible) and how the ethical individual is to act (by providing a content that is missing on the Kantian account, and which might be said to be equally missing in utilitarianism with its unquantifiable notion of maximizing happiness, and virtue ethics with its indeterminate conception of the virtuous agent). ⁶ Cf. Hegel 1977, §437, 262: ‘I have to think neither of making laws or of testing them. All such thinking on my part would upset that relation, since, if I liked, I could in fact just as well make the opposite conform to my indeterminate tautological knowledge and make that the law.’

90



However, the role ethicist can readily accept two limitations to this picture. First, they can allow that there may well be obligations which, while they have a social context, are not to be associated with social roles as such, even though the Hegelian might still want to view them as part of ethical life more broadly. For example, I may have an obligation not to steal your property because (on this account) property ownership is a social practice within our society, and so I would be violating various socially enforced norms if I were to do so, but this does not seem to be linked back to any particular social role that I occupy—unlike, for example, the police officer’s obligation to arrest me if I do decide to steal. Second, and relatedly, the role ethicist can allow that following one’s role obligations is likely only ever to be part of what it takes to be a good person in the moral sense, as there will be other obligations one needs to fulfil in order to complete the picture but which cannot be linked to a role, such as the obligation not to steal, or not to lie, and so on. Finally, the role ethicist should not be taken to be claiming that following the role obligations of just any role within just any society or institution can make one’s life good, in either of the two senses: for, if the role or society or institution is itself sufficiently ethically defective, then fulfilling the obligations associated with these roles may well be damaging to the individual involved, in terms of harming both their self-realization and their moral standing. The Hegelian will thus only make their claims in relation to the rational state, which, as they will be at pains to point out, is not necessarily the same as the existing state,⁷ which could indeed be corrupt enough to harm the individual who follows their roles within it, both as regards their own well-being and as a moral individual (where of course, again in an Aristotelian manner, these harms might well be said to be related). I trust this discussion has been sufficient to flesh out our first premise: ‘Role ethics holds (a) that you can be a good occupant of a role by fulfilling the obligations associated with the role, and (b) that this is sufficient to make you a good person (at least to some degree) provided that the institution itself is a good one.’ We can now turn to the second, which raises the concerns I now want to discuss: ‘But a person can count as having fulfilled their role obligations while being a moral monster.’ To explain this claim, let me begin with an example. Suppose you are a doctor, working in a hospital. Within that institution, you will have various obligations associated with that role, such as treating the illnesses of your patients, working effectively with your colleagues, keeping up your professional qualifications, attending various key meetings, and so on. Imagine you are extremely good at doing all these things—so good, indeed, that you are regularly promoted as a result. But now imagine that the reason you do all these things is not because you are moved by the interests and well-being of your

⁷ Cf. Hegel 2010, §6, 33–34, and Bradley 1927, 203–204.

       ?

91

patients, your colleagues, or indeed the institution itself, but because you welcome the high salary that comes with doing what is required of you, or because you fear censure or demotion if you do not, or because you crave the admiration of others for what you have done, or some combination of these motivations, which are perfectly sufficient in your case to keep you doing everything that is required of you in an exemplary fashion—hence all the promotions you have received. Now, it seems intuitive to say that if this was all true of you, then from a moral perspective, you are a monster—a monster of egoism, pride, arrogance, selfinterest, and so on. For, from a moral point of view, while you may perform right actions (such as curing the illnesses of your patients), you are doing so for all the wrong reasons. You have clearly fallen far short of the moral ideal. But (and of course this is the challenge), it is not so clear that you have fallen short of what it is to be good from the point of view of role ethics, for you have still fulfilled all the obligations associated with your role, by curing your patients, working with your colleagues, keeping up your professional training, and so on. That this is sufficient from this perspective is suggested by the fact that you have been promoted as a result, as from the perspective of the institution, and thus from the perspective of role ethics which focuses on one’s role in that institution, all that seems to matter is that these obligations have been met—it does not matter from this perspective what your motivations in meeting them happened to be. This then contrasts importantly with more general moral obligations, as it is commonly supposed that to have met one’s moral obligations, one must have not only done an action that is right or good, but also have done so for the right reasons: one cannot have met the obligation not to lie (for example), unless one’s motives in not lying involved some recognition of its wrongness, rather than merely fear of sanction or the desire to look good in the eyes of others. This is why there can be ‘wrong reasons’ objections to the claim that a moral obligation has been met—but it appears that there is no such corresponding ‘wrong reasons’ objection to meeting our role obligations: provided one has done what is required, one’s reasons for doing so seem irrelevant (as they might be to the doctor’s annual appraisal committee, for example). Moreover, perhaps even more surprisingly, the same blindness to your motivation seems to be true of us as individuals who engage with you in your role as a doctor. That is, if I have been operated on by you, arguably all that matters to me is that you did your job and hence performed your surgery effectively: I would not regard you any the less qua doctor if I came to find out that your motivations were as outlined above, as long as you treated me successfully, which seems sufficient to count as having fulfilled your role obligations, which in turn seems sufficient to make you a good doctor in my eyes. This again suggests that one can be taken to successfully fulfil one’s role obligations and hence be a good occupant of that role, while being a moral monster. Of course, I would be unlikely to want to have much

92



to do with you at a personal level—but I could still think you are excellent as a doctor, as you have done everything that is required of you in the role. Finally, if this second premise seems correct, we are led to our conclusion: namely, role ethics cannot really claim to be an ethics in any genuine sense, as one can count as good within the parameters of role ethics, while being a moral monster. This of course suggests that those parameters are set too narrowly for role ethics to really be an ethics at all, given what it seems to treat as sufficient for goodness in this context. And it should be clear that this claim is more radical than the more mundane worry often raised for role ethics, namely that the obligations associated with a role might clash with the requirements of more general morality, which sometimes seems implicitly taken to be the central ethical problem of roles: for example, police officers are permitted to lie in interrogations. Here we seem to have uncovered an even more fundamental difficulty.

3. Løgstrup on the Ethical Demand and Social Roles Having identified this issue, it would also be helpful to have some explanation of why it is that the problem arises: what is it about roles and their place in institutions that means someone can count as having perfectly well fulfilled their role obligations, while still being a moral monster? Of course, this isn’t a mystery when the institutions themselves are morally corrupt, or the roles themselves are ethically abhorrent—as in the case of the conscientious concentration camp guard, or the conscientious assassin; but the role of doctor that we have been considering is clearly within a good institution and is unproblematic qua role, and a similar argument could be given regarding other cases, such as being a teacher, a politician, a journalist, and so on. To offer an explanation which I find convincing, I will now turn to Løgstrup—who may also provide the role ethicist with a way out of the challenge, as I will explore in §4. Løgstrup’s major work of 1956, The Ethical Demand, begins by reflecting on the commandment to ‘love thy neighbour’, arguing that this commandment only makes sense given our vulnerability and hence dependence on others, to whom we are therefore required to respond with a love that responds appropriately to their needs. Thus, the power that the vulnerability of others gives us must be responsive to an ethical demand, which says that ‘the other’s life should be cared for in a way that best serves the other’ (Løgstrup 2020b, 48). The demand itself is then further characterized by Løgstrup in four key ways: as silent, radical, one-sided, and unfulfillable. The demand is silent because it does not just involve doing what the other person asks of you, and cannot just be read off social norms or rules, but requires you to reflect on what is best for the other person and take responsibility on yourself for the actions that follow in a way that makes it ‘isolating’; it is radical because the other person has no right to make the demand and it is

       ?

93

non-contractual, while it requires you to act selflessly and perhaps against your interests, though it should not be thought of as ‘limitless’; it is one-sided because you cannot demand something in return for the care you give to the other as a recompense or ‘payback’, meaning it is non-reciprocal in this sense; and it is unfulfillable because insofar as you feel that care for the other is a demand or obligation, you have already failed in the love for the other that such care truly involves, for to act out of love is not to act out of any such duty or obligation. Now, in terms of the issues being discussed in this chapter, it is highly significant that in presenting his position above, Løgstrup does so initially by contrasting the ethical demand with the ordinary social norms that govern our lives, for (he suggests) in the case of these norms, several key features of the ethical demand do not apply. First, while the ethical demand is ‘silent’ in that it does not lay down any prior obligations to be followed in this situation, interactions involving social roles usually do. Thus, for example, as a doctor there are certain responsibilities I have towards you and some I do not, and there are also certain procedures I am allowed to follow and others I am not, all of which structures how I am meant to behave towards you in a partially predetermined manner. Second, while the ethical demand is ‘isolating’ as it is up to you as an individual to decide how best to act in the situation rather than following prior structures, social roles are not wholly isolating in this way, as there will be norms of behaviour associated with the role for you to follow, and if what you do had the wrong results, you can pass responsibility on to those norms. Thus, as a doctor there will be certain methods I am expected to comply with in carrying out the operation, which I will have been trained to adopt; so if things go wrong, I can blame those methods and the professional training that led me to follow them. And third, Løgstrup argues, as we have seen above in §2, while the ethical demand is ‘radical’ because it requires you to act selflessly for the good of the other person, one can act in a social role (even in a ‘caring profession’ like being a doctor) from a variety of motivations, including self-interested ones. In these three respects, therefore, the ethical demand differs from the kinds of demands associated with social roles and their norms. As well as identifying these differences, Løgstrup also provides an explanation of why it is that social norms, including those associated with roles, differ from the ethical demand in this way, by offering a kind of genealogical account of how such norms arise, and why they are required. His account is provided in chapter 3 of The Ethical Demand, particularly section 3, which is entitled ‘The Protection Provided by Legal [retslige], Moral, and Conventional Laws [love]’. He begins by observing that the ethical demand arises out of our dependence on each other, in a way that therefore gives us power over one another, as we have seen. However, he notes, in itself there is nothing to prevent the ethical demand being ‘disregarded’ or ‘disdained’ [ringeagtet] by individuals, so that faced with our self-assertion and desire to promote ourselves, ‘[t]he radical demand provides no protection against

94



this, since through our disobedience to it we render it powerless’ (Løgstrup 2020b, 47). However, as we are inevitably dependent on others, ‘[o]ne human being is therefore abandoned to exploitation by another’ (Løgstrup 2020b, 47) in a way that cannot be avoided. Løgstrup then asks: ‘Does that mean that the individual is abandoned to being exploited by another human being in an unlimited way? Does the individual only have the choice of being either delivered up to the other person’s mercy—or more likely mercilessness—on the one hand, or on the other hand retaliating by exploiting the abandonment of the other to them in a kind of war of all against all?’ He then responds to his question as follows: No, it has not come to this, thanks to law, morality, and convention. They protect us against one another, each in their own way. In their respective jurisdictions, they impose limits on the individual’s exploitation of the fact that the other human being is at the individual’s mercy. They prevent quite specific forms of violence. (Løgstrup 2020b, 48)

And he then adds in a footnote: ‘In that sense they are rational’, where he identifies this as ‘their actual rational function’ (Løgstrup 2020b, 48 n. 6). For Løgstrup, then, in this respect the social structure arises because the ethical demand on its own is too precarious (so to speak) as a basis on which to ensure that we will act for the good of the other; we need norms of a different order, that will make it less likely that the vulnerability of the other will be open to our self-assertion and self-promotion. How does this work? According to Løgstrup, the advantage of the social norms is that they are better ‘accommodated’ to our nature, for as we have seen in their characterization above, they do not require us to be selfless in our attitude to the other in order to be followed successfully, unlike the ethical demand itself, where it is this selflessness that is so hard for us to achieve, and which thus makes it so precarious.⁸ This is the case for various reasons, which hinge on the difference between the ethical demand and the social norms that we noted earlier. First, the social norms are not ‘silent’, and therefore incorporate guidance or instructions on how we should act, so we are not required to work this out for ourselves using selfless concern for the other person as our guide, and in this respect can follow the social norms without much reflection and out of habit. Second, the social norms are therefore not ‘radical’, as it is perfectly appropriate to follow a social norm for all sorts of motives, such as fear of sanctions or the desire to be well thought of by others, where this is not appropriate for the ethical demand; this then means that we can use such motivations to enforce them and make sure they are followed. Third, social norms are generally more contractual and so are not ⁸ In The Ethical Demand, Løgstrup seemed to suggest that such selflessness is impossible, but he was later persuaded to clarify his view, and allowed that it can happen: for further discussion of this issue, see Stern 2019, 90.

       ?

95

one-sided, so built into them is an assumption of reciprocity, which means that they cater more easily than the ethical demand to our self-concern (for example, a doctor can expect to be rewarded for their efforts). And finally, social norms are fulfillable in the sense that we can feel them as demands or constraints on us, without at the same time undermining them, which is not the case with the ethical demand itself, which as a demand to love the other should not be felt as a demand at all. In all of these respects, therefore, the social norms are closer to our natures, and thus are more likely to provide a stable social structure within which the needs of others are satisfied, albeit without any fundamental concern for the care of those others. Thus, as Løgstrup concludes this section: ‘But all of this—habit, the irrelevant motives, and the sanctions—explains why the requirements of the social norms are largely met, and why the protection they thereby provide is, relatively speaking, so effective’ (Løgstrup 2020b, 49). Because we are vulnerable beings, with what Løgstrup takes to be a realistic estimate of the moral frailties of other fellow creatures, we should therefore welcome the social norms, as making possible for us a more stable form of social life, by reducing the likelihood that our inevitable and inescapable vulnerabilities to others are exploited by them, thereby enabling a more secure existence for us all.⁹ Up to this point, it may thus seem that Løgstrup has shown how the perspective of care embodied in the ethical demand and the perspective of the social norms embodied in our various roles differ fundamentally from one another, as highlighted in our example of the doctor. As an institution, the hospital is just tasked with getting its patients treated as successfully as possible, where it cannot rely on all its doctors being motivated by concern for their patients, and so puts in place a variety of other incentives (from the reward of a higher salary, to the threat of sanctions) in order to make sure this is achieved—and therefore does not concern itself with which of these possible motives its doctors act. Likewise, as patients of the hospital, who are aware that this is how it works and why, we also do not concern ourselves with the motives of the doctors who treat us, provided they meet the obligations imposed on them by the institution in a reliable manner. I might like a doctor who cares for me more than one who cures me to boost their success rate and hence salary, but if the latter does a better job than the former, I will think they are the better doctor, and be more satisfied with them as a result. Thus, in offering an account of why this level of social institutions and roles are needed, and how this differs from the level of the ethical demand, we have an explanation for the phenomenon we identified in §2, which seems plausible as a philosophical genealogy. It also seems to explain why role ethics faces the challenge we identified above, for the place of these institutional obligations is ⁹ Cf. also Løgstrup 2020b, 19, where Løgstrup presents the ‘neutral possibility of social convention’ as follows: ‘We cannot bear the two alternatives, of care or of ruin, which belong to given life itself, and so to avoid these alternatives we have tacitly agreed to adopt convention as a form of existence instead.’

96



precisely to replace the motivations of the ethical demand with a different set of motivations, otherwise they wouldn’t be serving their Hobbesian function, of protecting us when the selfless motivations of the ethical demand fail. And it perhaps might be said, this feature of role ethics is reflected in the very terminology of ‘roles’ itself, which we take up in the manner of actors who can perform their part from a variety of underlying motivations, which ultimately do not matter to us provided the performance is convincing on the surface.

4. Refraction: Løgstrup on the Relation Between the Ethical Demand and Social Roles Having used Løgstrup to show how the difficulty faced by role ethics has emerged, why this is to be expected, and why it fits with our practices and intuitions to a large extent, I now nonetheless want to show how considerations that are also to be found in Løgstrup might also be used to address this difficulty, and to resolve the problem we have raised. For, while stressing how social norms differ from the ethical demand, and why they are needed and are valuable, Løgstrup does not think that such norms can ever be entirely unrelated to the ethical demand, so that the perspective of the ethical demand will still be needed in a way that connects to our social roles, for three main reasons. First, while, as we have seen, social norms are more ‘articulate’ and less ‘silent’ than the ethical demand, in providing more determinate procedures for us to follow, nonetheless they rarely tell us what to do in all circumstances, so that some degree of judgement or discretion will still be needed, which will then require us to have some sense of the ethical demand if we are to know how to act. Secondly, while we can follow social norms with a variety of motivations, when these norms fall short and do not tell us how to behave, we will then need the motivations associated with the ethical demand in order to act correctly. And thirdly, the ethical demand provides a kind of critical purchase on the social norms, as our aim in constructing such norms and roles within institutions should be to ensure that care for our needs as dependent creatures is catered for as much as possible, and so structure our roles, norms, and practices accordingly. For Løgstrup, therefore, a social structure that tried to forget or do without the normativity of the ethical demand would become corrupted, and it cannot ever lose touch with it altogether, as our social norms are never so determinate that we can do without any concern for the individual as such, so that the ethical demand and its motivations will always come back into play in some form, even if it is also mediated by other social considerations.¹⁰ ¹⁰ Cf. Løgstrup 2020b, 53: ‘There are also social norms that can be fulfilled to the benefit of the other person only if we are indeed moved by consideration for the person. The connection between the radical demand and the social norm can therefore be very intimate—and the social norm on its own can be very inadequate.’

       ?

97

Given therefore that both these levels are needed—the ethical demand on the one hand, and the social norms on the other—how should we think of them as being related? Løgstrup introduces a useful metaphor to explain the relation, which is that of refraction: namely, just as white light becomes broken up into the various colours of the spectrum, so the ethical demand finds itself dispersed across the range of social relations, as different ways in which it is articulated through those different relations: In other words, the demand is, as it were, refracted through a prism in a variety of ways. In the first place are the various distinctive relationships in which we stand to one another—as spouses, parents and children, employer and employee, teacher and student, and so on. The demand is refracted through the spiritual content of these relationships . . . All these various relationships have in common that they do not oppose the demand, but urge us to do the same as it does. For in each one of them—each in its own way—the one human being holds something of the other human being’s life in their hands. Each one of them is a distinctive form of the fact out of which the radical demand arises. (Løgstrup 2020b, 91)¹¹

Thus, while we cannot reduce the one to the other, and while we need to retain social norms, we should nonetheless see properly constituted social roles as to some degree relating to and embodying the ethical demand, though not mirroring it precisely. What I think Løgstrup has in mind can be illustrated by an example Løgstrup himself gives, which focuses on the role of being a parent (Løgstrup 2020b, 53–54). To some extent, parents can perform some aspects of their role in relation to their children just by following the rules and regulations associated with that role, such as the requirement to provide certain material goods or some degree of education, and this might be sufficient to make them an adequate parent; nonetheless a parent that simply acted in this way towards their child would be failing them if they did not engage with their child as a particular individual in a caring way, to whom their responsibilities have something of the characteristics of the ethical demand, and this will also be needed for them to count as a ‘good parent’. This is because whatever the obligations which might be treated as essential to the role from an institutional perspective, that institution and those obligations are there to enable the child to be cared for and so they refract the ethical demand; but a parent who did not understand that would be unable to respond properly when the obligations fail to provide specific guidance and so remain relatively ‘silent’,

¹¹ Cf. also Løgstrup 2020a, 8: ‘The ethical demand is refracted as through prisms by all the different and particular relationships in which we stand to one another as spouses, parents and children, teachers and students, employers and workers, as they are all forms of the fundamental condition whereby the ethical demand receives its content.’

98



which is inevitable as the social norms and rules around parenting can never be made so precise that the parent could successfully relate to their child by just following them, in an entirely loveless manner. Thus, for example, Løgstrup contrasts the obligations parents might have to ensure their child has a basic education, which perhaps could be fulfilled by just following certain requirements set down by law, and the obligation they also have ‘of enabling the child to become independent from them as parents; this latter can only be carried out through judgement on their part—and cannot be accomplished by following a number of precise directives on how one should act, but rather cannot be accomplished without love for the child’ (Løgstrup 2020b, 54). Similarly, to return to my own example of the doctor: even though it is the case that the doctor may successfully perform the operation simply by following the procedures and with no concern for my well-being as such, nonetheless when these procedures remain indeterminate or do not cover the situation at hand, the doctor must be able to respond by being motivated to decide what is best for me as a patient; so to be a good doctor on Løgstrup’s account, they must still be able to respond with care, otherwise they will not really grasp the point of these procedures, or be able to act appropriately when they break down or do not cover the situation at hand, and so when the doctor needs to judge for themselves about how to behave and precisely how best to fulfil their obligation. This in turn will mean the doctor must ‘internalize’ the norms governing the role, in the sense of grasping the goal for which those norms are instituted, and so see the former as ‘refracted’ in the latter in the way that Løgstrup suggests. We will still need a way to enforce those obligations, in order to ensure that those who practise medicine have other incentives to cure their patients even when they fail to care for them; but the role ethicist can still hold that these doctors are defective exemplars of this role, who will be found wanting in certain key situations, if this is the only way in which they relate to their role obligations. If this is correct, then we can see how in the work of Løgstrup, we can find considerations that can help with our initial concern, even though his work also makes that concern vivid in the first place. For, in the case of any reasonably complex social role, such as being a parent, doctor, teacher, and so on, the role ethicist can argue on Løgstrupian grounds that to be a good representative of that role will always require more than having fulfilled the obligations associated with that role, as the agent must be able to see the deeper structure of care which that institution is designed to protect, in order for them to know how to fulfil those obligations when they remain partially silent, and so require judgement to follow, which is always likely to be the case with a role beyond the most basic. What this brings out is therefore the following. On the one hand, unlike in the moral case, a person can count as having fulfilled their role obligations if they do so for a variety of motives, including self-interest and so on, and thus in this sense there is no ‘wrong reasons’ objection to how we fulfil our role obligations.

       ?

99

Nonetheless, on the other hand, merely having fulfilled one’s obligations does not make one a good occupant of the role. For, a person who fulfils their obligations for these wrong reasons is unlikely to fulfil their obligations in a reliable manner, which will require instead that they have internalized the reasons why the institution has adopted these obligations, as judgement will be required to fulfil those obligations in many cases, which requires some understanding of their purpose, and thus the ethical demand which underpins them. And while an agent might coherently set out to internalize the norms of a role on purely instrumental grounds (for example, to make themselves more reliable at fulfilling their role obligations and thus more likely to gain promotion), an agent cannot succeed unless this instrumental attitude has been lost during this process, as otherwise the norm would not have been properly internalized, and the relation of the individual to the role would remain problematic for that reason. Thus, in terms of our original argument: 1. Role ethics holds (a) that you can be a good occupant of a role by fulfilling the obligations associated with the role, and (b) that this is sufficient to make you a good person (at least to some degree) provided that the institution itself is a good one 2. But a person can count as having fulfilled their role obligations while being a moral monster 3. So role ethics is not an ethics. What our discussion now suggests is the following response: while Løgstrupian reasons can be given for 2, as role obligations can indeed count as having been fulfilled even if they are done for all sorts of immoral reasons, nonetheless premise 1 turned out to be too simplistic. For, to be a good occupant of a role, one has to be someone who can be expected to reliably fulfil the obligations associated with it— and this will require being motivated by considerations of a more recognizably moral kind that are associated with the ethical demand, so that role ethics can claim to be an ethics after all, at least as far as this argument goes.

5. Implications In conclusion, I now want to briefly consider the implications of the Løgstrupian picture for more general issues that are often discussed in the literature on role ethics, where I will focus on three, to which I think we can now give a fresh response. First, one common point for discussion is how far roles provide us with distinctive ethical obligations which can be justified in their own terms and so treated as foundational, or whether they must be grounded on some more general ethical basis. This can then be presented as a kind of dilemma for role ethics: if role

100



ethics is distinct from ethics more generally, the worry is that it can then be seen as problematic from that more general perspective; but if it is grounded in that general perspective, it may seem redundant.¹² Løgstrup’s approach may offer a way to avoid this dilemma, however. For on the one hand, we have seen how role ethics can avoid the challenge of redundancy, as on Løgstrup’s account we need an account of the social roles that understands how they differ from the ethical demand, given their distinctive function; on the other hand, we have seen how such roles also need to relate to the ethical demand, which means they cannot ultimately become cut off entirely from this ethical space or treated as selfstanding. As Løgstrup puts it, social norms must ‘not oppose the demand, but urge us to do the same as it does’ (Løgstrup 2020b, 91), as the ethical demand and so are not independent of it in that sense, but nonetheless their relation to us as norms may differ from how we relate to the ethical demand and so they can have a distinctive place in our lives. A second issue concerns exactly what we classify as ‘roles’, and on what basis. On some accounts, they are closely tied to institutional structures that have been designed for certain purposes which we join voluntarily in a contractual manner, such as various sorts of professional roles. However, on other accounts this is said to be too narrow, on the grounds that for example being a parent or child can be said to be a role, even if it is not carried out within any contractually based institution, and has not been consciously designed. On the other hand, there is a worry that if we include being a parent or a child as being a role, then it is hard to see why we should not also include friends; but this seems to be going too far from the kind of professional roles from which we began. So the worry is, does the idea of a social role form a coherent class? One advantage of Løgstrup’s position may be that it can explain why social roles form a continuum, but with relatively clear boundaries, by asking the question: can one count as fulfilling the obligations involved even on the basis of the kinds of motives of self-interest that we have discussed—where for reasons given, the more institutional the set-up, the more likely this is to be the case, as from an institutional point of view, what matters is we act as required, not why we do so. We can thus expect professional roles to pass the test, but also other social roles which still have institutional aspects, like being a parent or a child: society still has an interest in making it the case that parents and children behave in certain ways which it is important to enforce through various motivational mechanisms, including self-interest, for example ensuring that parents enroll their children in school, or provide them with adequate meals, and ¹² See e.g. Walker and Ivanhoe, who put the problem as follows: ‘Either Role ethical requirements conflict with general moral requirements or they do not. If they do conflict, then we are stuck with deep moral dilemmas in which actors are torn between their ethical duties in their identified roles (as soldiers, teachers, lawyers, physicians, mothers etc.) and their ethical requirements as human beings. If they do not conflict, then there is no distinctive ethical arena for roles—all ethical responses are merely those of the good human being’ (2007, 26).

       ?

101

while the parent/child relationship has not itself been socially designed, its reinforcement in various ways can still be said to stem from various social purposes we have found for this relationship. However, these relationships are closer to the margins of role ethics than the professional cases, as other aspects of the relationship cannot be handled in this way—for example, the love that a parent should feel for their child cannot be treated as an obligation one can fulfil out of self-interest, or indeed as an obligation at all, as if one feels obliged to love one has already failed. And some relationships seem to fall outside this range of possible self-interested motivations altogether, such as friendship: here, whatever one does for one’s friend would seem to fail that relationship unless the motives were based on concern for one’s friend and one’s relationship to them, and not simply one’s individual interests (although of course that does not preclude one’s interests being furthered by this relationship). Løgstrup thus seems able to provide us with an attractive way of carving up the phenomena into coherent groups. Finally, another puzzle concerning role ethics concerns the source or explanation of the obligatoriness of roles: does it come from the explicit or implicit contracts that might be said to underpin some roles, or the social sanctions that can be applied to roles, or the way certain roles constitute our self-identity, or just the way in which role obligations relate to moral obligations more generally, and so is their obligatoriness to be explained by whatever one says about the sources of obligation at that more basic level? One feature that makes Løgstrup’s approach attractive is that rather than having to decide between these options in a way that then often turns out to be too simplistic, he can explain how we might find all these sources of obligation at play at different levels, depending on exactly how we relate to the role. Thus, to the doctor who has no intrinsic concern for the wellbeing of their patients, the obligations attaching to the role will be socially enforced in a variety of ways, or they might accept the obligation on a contractual basis. But for the doctor who by contrast has ‘internalized’ the role, they may feel bound to respond to their patients because of their self-identification with the role. And at a further level, once the role has been internalized such that the underpinning ethical demand is salient to the doctor, the obligation to care will arise not from the role as such but from the perceived needs of their patient—such that if they truly care for the patient, it will not be felt as an obligation at all. Løgstrup’s complex account of the nature of social roles and their obligations can therefore help us understand why this issue has no neat or single answer. In this way, as in the others, Løgstrup may be seen to offer a challenge to role ethics, but in a way that ultimately might help it find answers to that challenge and other issues that have arisen for this tradition in our ethical thought.¹³

¹³ I am very grateful to the following for helpful comments on previous versions of this chapter: Chris Bennett, Sean Cordell, Kayleigh Doherty, Max Hayward, and James Lewis.

102



References Bayer, Oswald. 1998. ‘Nature and Institution: Luther’s Doctrine of the Three Estates’. Christine Helmer (trans.), Lutheran Quarterly 7: 125–159. Bradley, F. H. 1927. Ethical Studies, 2nd edition. Oxford: Oxford University Press. Chappell, Sophie Grace. 2020. ‘Roles and Reasons’. In Perspectives in Role Ethics: Virtues, Reasons, and Obligation. Tim Dare and Christine Swanton (eds.), pp. 194–217. Abingdon: Routledge. Cottine, Cheryl. 2020. ‘That’s What Friends Are For: A Confucian Perspective on the Moral Significance of Friendship’. In Perspectives in Role Ethics: Virtues, Reasons, and Obligation. Tim Dare and Christine Swanton (eds.), pp. 123–142. Abingdon: Routledge. Freyenhagen, Fabian. 2011. ‘Empty, Useless, and Dangerous? Recent Kantian Replies to the Empty Formalism Objection’. Hegel Bulletin 32: 163–186. Hardimon, Michael. 1994. ‘Role Obligations’. The Journal of Philosophy 91: 333–363. Hegel, G. W. F. 1977. The Phenomenology of Spirit. A. V. Miller (trans.), Oxford: Oxford University Press. Hegel, G. W. F. 1991. Elements of the Philosophy of Right. H. B. Nisbet (trans.), Cambridge: Cambridge University Press. Hegel, G. W. F. 1995. Lectures on the History of Philosophy, 3 vols. E. S. Haldane and Frances H. Simson (trans.); reprint edition. Lincoln, Neb.: University of Nebraska Press. Hegel, G. W. F. 1999. ‘On the Scientific Way of Treating Natural Law’. In Hegel, Political Writings. H. B. Nisbet (trans.), pp. 102–180. Cambridge: Cambridge University Press. Hegel, G. W. F. 2010. Encyclopedia of the Philosophical Sciences in Basic Outline: Part I: Science of Logic. Klaus Brinkmann and Daniel O. Dahlstrom (trans.), Cambridge: Cambridge University Press. Hoy, David Couzens. 1989. ‘Hegel’s Critique of Kantian Morality’. History of Philosophy Quarterly 6: 207–232. Hu, Jing Iris. This volume. ‘Roles and Virtues: Which is More Important for Confucian Women?’ Løgstrup, K. E. 2020a. Ethical Concepts and Problems. Kees van Kooten Niekerk and Kristian-Alberto Lykke Cobos (trans.), Oxford: Oxford University Press. Løgstrup, K. E. 2020b. The Ethical Demand. Bjørn Rabjerg and Robert Stern (trans.), Oxford: Oxford University Press. Mander, W. J. 2016. Idealist Ethics. Oxford: Oxford University Press. Martinsen, Kari. 1993. Fra Marx til Løgstrup: Om Etikk og Sanselighet i Sykepleien. Oslo: TANO. Martinsen, Kari. 2006. Care and Vulnerability. Linn Elise Kjerland (trans.), Oslo: Akribe.

       ?

103

Martinsen, Kari. 2012. Løgstrup og Sygeplejen. Aarhus: Klim. Stalnaker, Aaron. 2020. ‘Roles and Virtues: Early Confucians on Social Order and the Different Aspects of Ethics’. In Perspectives in Role Ethics: Virtues, Reasons, and Obligation. Tim Dare and Christine Swanton (eds.), pp. 96–122. Abingdon: Routledge. Stern, Robert. 2015. Kantian Ethics: Value, Agency, and Obligation. Oxford: Oxford University Press. Stern, Robert. 2019. The Radical Demand in Løgstrup’s Ethics. Oxford: Oxford University Press. Walker, Rebecca L. and Philip J. Ivanhoe. 2007. ‘Introduction’. In Working Virtue: Virtue Ethics and Contemporary Moral Problems. Rebecca L. Walker and Philip J. Ivanhoe (eds.), Oxford: Oxford University Press. Wingren, Gustav. 1957. The Christian’s Calling: Luther on Vocation, Carl C. Rasmussen (trans.). Philadelphia: Muhlenburg Press. Wood, Allen W. 1989. ‘The Emptiness of the Moral Will’. The Monist 72: 454–483.

PART TWO

T H E N AT UR E A N D N O R MAT I VI TY O F R OLE S

6 All Together Now When Is a Role Obligation Morally Binding? Erin Taylor

1. Introduction It is natural to think that pairs of sentences such as those below do not merely state two unconnected truths about the agents involved. Rather, in each case, the truth of (b) seems to be in some sense explained by the truth of (a). (1a) Maura is Ethan’s mother. (1b) Maura has an obligation to provide for Ethan. (2a) Aisha is William’s oncologist. (2b) Aisha has an obligation to inform William of the risks of chemotherapy. (3a) Pedro is Jim’s lawyer. (3b) Pedro has an obligation to file timely briefs on Jim’s behalf. Competing theories of the moral status of social roles, and their associated obligations, aim to explicate the nature and significance of this explanatory connection. In this chapter I defend a novel account of the connection between social roles and their associated demands. In particular, I argue that many of the obligations or norms that attach to social roles morally bind us by default. These norms are morally binding since they are ultimately explicable in terms of conventional moral duties and those conventional duties enjoy a default but defeasible good standing.

2. Conventional Norms and Moral Bindingness Some Terminology and Distinctions I will assume that we can agree about a list of paradigm cases of social roles— doctor, patient, teacher, parent, government minister, lawyer, professor, etc. I treat such roles as constituted at least in part by (social) institutions. I’ll view Erin Taylor, All Together Now: When Is a Role Obligation Morally Binding? In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Erin Taylor 2023. DOI: 10.1093/oso/9780192843562.003.0006

108



institutions as sets of conventional norms that assign occupants to roles, to which rights and duties are attached. The family, the professions, the political system, and the military are institutions in this sense. Social roles help guide the behavior of their occupants and structure one’s expectations of others. As the above examples indicate, institutions differ with respect to their roles (family member, educator), and sub-roles (sister, professor), and in how flexible their compliance conditions are. Minimally, a convention is a social practice that people follow at least partly because others do, too (Lewis 1969; Burge 1975; Marmor 1996; Miller 2001). They are thus social regularities that we are typically able to correctly predict and describe. For example, if you had to bet whether guests will talk loudly over the couple’s vows during a wedding, you’d probably bet that they wouldn’t, and you’d probably be right about that. But conventions may also have a normative component. A conventional norm in my sense is a standard of conventional practice to which participants hold each other responsible. Driving on the correct side of the road, educating one’s children, following hospital guidelines, and not talking loudly over wedding vows are all conventional norms in this sense. The existence of a conventional norm is an empirical matter, describable by social science. But conventional norms may additionally be morally justified or binding (pro tanto). I will call these conventional moral duties.¹

Locating the Project Before we begin to explore my treatment of conventional moral duties, it might be helpful to stress what I am not aiming to do here. First, I am not classifying such roles as conventional because I think that their moral bindingness can be wholly explicated in terms of non-moral convention alone (and indeed I see no reason to believe such reduction is possible). Rather, I call them conventional because what counts as complying with their associated norms depends at least partly on contingent social facts. It is plausible that they are constituted by the rights and duties that attach to them, and the contours of such norms depend on contingent patterns of mutual expectations and social behavior. For example, the detail of physicians’ duties evolves over time; and what counts as complying with duties to kith and kin may expand or contract with the availability of social services, the timing of the school year, etc.

¹ I won’t classify a conventional norm as a conventional moral duty if its being morally binding is not properly related to its role in regulating social practice, i.e. in a way that involves essential explanatory appeal to its status as a conventional norm. I say more about what this amounts to at the beginning of section 4.

  

109

Second, although my focus here is on the moral bindingness of social conventions qua social conventions, I do not wish to deny that people will often also have moral reasons to act as social roles demand that are in some sense independent of that conventional nature. For example, it might be that Maura could be subject to some moral obligations to provide for Ethan just because Ethan needs looking after, that Maura is especially well situated to care for him, etc. in a way that could in principle be true even if they were not related as parent and child. Furthermore, in addition to such convention-independent moral demands, occupants of social roles may also have independently persuasive non-moral reasons for acting as conventions demand. So, for example, citizens may pay taxes to avoid going to jail, doctors may care for patients to garner prestige, etc. My point is only that a good moral theory should explain that characteristic pattern of connections between role and obligation that we find in the examples above. To sum up this section, we can view social roles as generated by institutions, and constituted by sets of conventional norms. Thus, the question “When is a role obligation morally binding?” is an instance of a wider question: “When is a conventional norm morally binding?” Let us now turn to this question.

3. Why We Can’t Have Nice Things Consider some paradigm cases of conventional norms: Drivers ought to obey the traffic signs. Doctors ought to follow hospital consent procedures. Parents ought to celebrate their young child’s birthday. Football referees ought to master the details of the offside rule.

Are such norms also conventional moral duties? Let’s consider some reasons why it would be nice for us as reflective moral agents if they were.

Moral Accuracy First, we have a general interest in the moral appearances matching moral reality. Since it often seems to us that we should, for example, keep our promises or abide by professional standards, it would advance this interest if we were right. Conventions must by their nature structure both expectation and performance. If conventions were morally binding, we would expect a high degree of fit between the way things seem morally and the way they are, at the level both of belief and of action. Conventions could thus help moral agents respond properly to moral reasons.

110



Moral Knowledge Second, it would be good if conventional norms were morally binding since that would give us a way to reliably track what we ought to do. Our accurate moral beliefs would be stable and difficult to lose. They would result from a faculty that reliably guided us with respect to moral matters, that constituted an admirable feature of our epistemic character, and that prevented our successful actions and beliefs being a matter of luck. Quite plausibly they would amount to moral knowledge.

Moral Understanding Third, we value being right about the moral explanation of why we should do X rather than Y. We want more than mere reliable true belief. We also want to have access to satisfying explanations of why X is the right thing to do. Since we often cite facts about the occupancy of social roles or the demands of conventions in giving such moral explanations, we value those norms in fact being relevant. “We ought to make sure that we sort something out for his birthday because we are his parents” doesn’t only tell us a putative fact about what we ought to do, but also about why we ought to do it.

Moral Efficacy Fourth, if conventions were morally binding we would have a shot at living a successful moral life, without first having to complete the difficult task of resolving foundational philosophical disagreements. Contrast (a) knowing who morally ought to plan the birthday via grasping the demands of relevant social roles and conventions, with (b) knowing who ought to do it because one can formulate and defend an account of the ultimate moral grounds, and then derive particular contextually embedded courses of action from those.

Moral Neutrality Fifth, there is widespread agreement among competing moral theories as to the moral necessity of many institutions. This means that moral explanations given in terms of conventional norms will often be ones that moral theorists and agents can employ to coordinate action successfully. This will be the case even when theorists and agents disagree with each other in some quite fundamental ways about the ultimate ground of support for those norms.

  

111

Moral Transparency Sixth, conventions are fixed by mutual expectations, and so (by their nature) must be shareable, learnable, and comparatively easy to identify and follow. They can be taken as objects of study by the social sciences. If conventions were morally binding, then we would thus have access to a broadly naturalized epistemology of moral demands. Foreseeably, this will inform moral reflection. For example, reading social surveys might help us understand when we are demanding too much of each other as parents. As theorists, we might learn from organizational behavior studies that when seeking consent from a group of people it is important to attend to certain extant forms of social relations. Failure to do so might, for example, predictably lead to “polarization cascades” of the kind that arise with respect to, for example, vaccine participation, fluoridization, or mask-wearing during pandemics. All of the above suggests that it would be a pleasing result if conventional norms were also thereby morally binding. Unfortunately, a number of competing considerations seem to suggest that this is not the case.

Evil Conventions Consider first the existence of morally unacceptable conventions. Human history gives us many examples of conventional norms that any plausible moral theory would condemn. These examples establish that some such norms fail to bind us morally. Such conventions can be deeply interwoven with social structures that have permissible variants. However, it is often difficult to extract morally acceptable conventional norms from the evil systemic structures that house them. We find good and bad conventions entwined in such a way that no simple and direct excision of the bad ones seems open to us.

Fundamental Conflict Second, even if some social roles are morally acceptable, their associated norms may bring us into conflict with other basic values. Perhaps there is nothing intrinsically bad about parents organizing birthday celebrations, in the way that slaveholding, caste hierarchies, or footbinding practices are widely thought to be evil in themselves. It might still be true that we should spend our time and money dealing with famine instead, just as proponents of Singer-style beneficence would insist. Similarly, theists and libertarians might concede that school dances or compulsory taxation for beneficent causes do not deserve moral condemnation in the way that evil practices do. Nevertheless, they might think that the value of God’s commands or freedom make such social practices morally unjustified.

112



Formulating Principles Third, one potential remedy to the above two problems has proven difficult to obtain. This is to winnow out the morally acceptable conventions in a way that avoids conflict with other values by identifying a set of general moral principles. These would identify the conventional moral duties and explain why they have that status.² Unfortunately, no version of this principle-based approach commands the field. Theorists have (a) struggled to formulate principles that apply to all and only the extant paradigm cases and/or (b) provided principles that fail in (often strange and unrealistic) counterfactual or hypothetical cases. To take a prominent example, Rawls’s Principle of Fairness has spawned an expanding range of criticism, responses, and modifications. Even if some future variant of one of these principles turns out to make intensionally correct predictions, the initial controversy and resultant complexity undermines its explanatory potential. The outputs of such a methodology often seem gerrymandered and highly unnatural. This in turn seems to undercut the potential benefits of morally binding conventions outlined above. Even if some complex, unintuitive principle in fact encapsulates all and only the morally binding obligations, this threatens our moral knowledge and understanding. The history of seemingly plausible principles that turn out to fall to unusual cases and bizarre counterexamples means that we have excellent higher-order evidence that it is not correct, and that we shouldn’t place too much faith in it when the moral stakes are high. We have strong inductive grounds for doubting that we have locked on to a generalization that correctly classifies the cases, even if we have done so (Williamson 2002).

Explanation and Action-Guidingness Fourth, even if philosophical reflection eventually allows us to formulate unified, non-gerrymandered, and plausible looking principles, we would still face a serious problem. Attempts to identify the morally binding conventions seem responsive to two conflicting standards. On the one hand, as moral theorists, we want the principles that single out the good cases to be as unified and general as possible, while covering all of the cases. This relates to our interest in explaining why acting

² By “principle” I have in mind an exceptionless, lawlike universal generalization that (i) tells people who meet a certain condition that (ii) they are to do a certain thing in a way that (iii) is detailed enough to guide action and (iv) helps explain why they ought to do that thing. The Principle of Reparation for parental duties to children, the Principle of Fairness for political obligation, and Scanlon’s Principle of Fidelity are potential examples. I won’t count general explications of the rightness or wrongness of action in terms of candidate values—“that form of government is unfair,” “beneficence demands that you pay more tax,” etc.—as principle-based in this sense.

  

113

in the light of conventional guidance is permissible or wrong. Our interest in providing moral explanation pushes us towards comparatively abstract and general moral “laws.” But as moral agents, we are interested in what to do here, now. This interest in action-guidingness brings with it a concern for the fine-grained, highly nuanced, and contextually specific detail of moral situations. The general principles provided by explanatory accounts do not themselves tell us how they ought to be applied in the morally puzzling situations in which we actually find ourselves. These conflicting desiderata for a principle-based theory of conventional moral duties make it difficult to see how its output could ever do the job that we want it to do.

The Norms-as-Theorems Model The attempt to trace back the moral authority of conventions to fundamental principles can be motivated just by reflection upon evil conventions and normative conflict. Some methodological approaches to moral theory seem to make the attempt even more indispensable, however, and its failure even more costly. Such approaches seem to treat non-bindingness as a kind of default state, that only a vindicatory explanation in terms of the moral fundamentals can serve to overthrow. We can draw an analogy between this model and the proof-based approaches employed in the formal sciences. Even a seemingly plausible nonaxiomatic mathematical principle may be treated as unknown and potentially untrue if we lack a derivation of it from more basic principles, which themselves are thought to trace back to a sparse and explanatory set of axioms. Similarly, for proponents of what we could call the norms-as-theorems model, we can only be confident in the moral guidance that conventions provide if we have derived them from more fundamental moral principles. These norms have only a provisional or candidate status before such a vindicatory program is completed. This philosophical methodology seems bound to deprive us of much of the moral knowledge, understanding, and guidance that we had hoped conventions would provide. As Bernard Williams famously suggests, it threatens us with a loss of moral knowledge. This threat remains even if our conventional norms are in fact morally binding just as we believe (Williams 1985). Similarly, it seems likely that moral efficacy and moral neutrality (as described above) would also be undermined by (i) the methodological demand for theorem-hood, together with (ii) the disciplinary failure to provide it in widely recognized ways. To sum up, it would be nice for us if conventional norms were also thereby morally binding. But any such view faces a range of thorny-looking problems. In the remainder of this chapter, I will set out a treatment of conventional norms that may help us reap the benefits without paying the unpleasant costs. Even if my attempt here fails, I hope it might suggest a new angle of attack.

114



4. A Minimal Theory of Convention The picture I have in mind is rather complicated in detail, so let me first give a broad-brush overview of the strategy. I am not aiming to give any kind of reduction of moral norms or values to conventions (and I see no reason to believe such reduction is possible). Rather, I will suggest that reflection on the characteristic features of social conventions qua conventions helps motivate an attractive account of conventional moral duties. We can start from features of convention that competing approaches to morality can agree upon, then try to build up from there to an account that secures the desirable features of conventional norms while minimizing the problematic ones. In particular, I will argue that the public, shared, holistic, and morally essential aspects of social conventions favor a view on which their associated norms bind us morally by default, unless overthrown by a certain form of demonstration of their ill-standing. I begin by outlining a minimal theory of convention. A minimal theory in this sense is not wholly platitudinous or uncontroversial, but it should involve claims that many of us are likely to accept or at least judge as reasonable. Following David Lewis, I will think of convention as (minimally) a social practice to which people conform at least partly because others do, too. In a given cultural context conventions might include: driving on a designated side of the road; using certain forms of words to take on binding commitments; giftgiving on your child’s birthday. Recall that a conventional norm is a standard of conventional practice to which participants hold each other responsible for conforming. Participants in conventions typically include, for example: People who can reasonably be held responsible for acting in a way responsive to the convention; People who are the beneficiaries of the norms, in a way specified by the convention; People who help enforce compliance of the norms. Thus, in the case of driving conventions, participants might include licensed and unlicensed drivers, traffic police, legislators, and so on. In the case of conventions governing dissertation supervision the obligation to aid students in certain stipulated ways might devolve to designated faculty members, be enforced by Department Chairs, etc. I do not intend anything overly complicated or controversial here. Conventional norms are individuated by their applicability and success conditions. The applicability conditions of such norms are the conditions under which certain participants become subject to the norm. For example, the condition under which I become a dissertation supervisor might be that I am a faculty member with a certain style of contract who has been assigned that task by my Department Chair. The success conditions of a conventional norm are given by the content of

  

115

its demands. If adequately discharging a supervisory role requires that PhD supervisors read and comment on a student’s work, then this will be part of the success conditions for that norm. The possibility of evil conventions seems enough to show that conventional norms are not always morally binding. Building on my discussion from section 3, I will call a conventional norm “morally binding” when: a. Participants who meet the applicability conditions morally ought (at least pro tanto) to act as the convention demands b. A convention is at least instrumental in generating such a moral requirement c. A convention plays a part in specifying the applicability and/or success conditions of those requirements. What is it for a convention to be instrumental in generating a moral requirement? I have the following kind of thing in mind. Suppose for the sake of argument that ultimate moral grounds such as utility or rational autonomy, in combination with general descriptive facts about human life, could by themselves explain why people ought to be allowed to vote for a representative government once they reach some suitably advanced age. Even if this were true, it would still be the case that before we can determine whether we ought to allow 17-year-old Sarfraz to vote, we would need to determine the facts about which arbitrary voting age was selected in our society. The convention that we have in fact adopted would thus be instrumental in generating the more precise, action-guiding norm. From the discussion so far we can extract some theses about convention that I hope are comparatively uncontroversial: (Co-conformity) A convention is (minimally) a social practice to which people conform at least partly because others do, too. (Individuation) Particular conventions can be individuated by their applicability and success conditions. (Distinctness from the Moral) Conventional norms are not by nature identical with moral norms. (Norm Imperfection) Many actual and potential conventional norms are morally unacceptable. (Moral Significance) It matters whether there are any conventional norms that are morally binding in the sense that I set out above.

But these are not the only theses about convention that might command wide agreement. Another concerns the interrelatedness of social practices. It seems very plausible that conventions are parts that function within whole systems of

116



institutions. Token conventional norms are constitutively interrelated with others in at least two senses. First, the norm cannot be adequately characterized without citing other conventions. For example, in describing the norms applying to poll workers on election day, one needs to cite other conventions related to citizenship, criminal convictions, voter identification, security of ballot papers, what counts as a signature, etc. Second, the moral bindingness of a conventional norm often depends on the moral status of others. Thus, the question of whether the poll worker is morally bound (pro tanto) to prevent Sarfraz from voting in certain circumstances may well turn on whether a discretionary police order to close down polling booths earlier than advertised was itself morally justified. If the order was motivated by a need to respond urgently to an imminent threat of harm to voters, then the poll worker may be morally bound to do so. Things are different if the police officer is immorally and illegally attempting to subvert free elections. Similarly, the question of whether a parent is permitted to rely on the school bus to drop off their child right at the entrance will not be independent from the question of whether the bus driver is morally permitted to use their discretion about where the child is dropped off, what procedures are in place in the event of medical emergencies en route, what counts as a morally authoritative order from a traffic cop, and so forth. Unsurprisingly, questions about the moral bindingness of conventional norms thus display the kinds of local epistemic holism familiar from other areas of enquiry. Settling the question of who burgled my house is not fully independent of settling the questions of whether people have unique fingerprints. Likewise, the question of whether someone failed morally as a parent is sensitive to evidence about the responsibilities of the bus driver, school nurse, governor, police officer, and so on. Assuming that the many individual conventions making up modern social life form an interrelated web in this way, we have: (Interrelatedness) Conventional norms are functional parts of interrelated systems. Identifying their nature and justification typically requires reference to other elements of that system of conventions. This reference may relate to the applicability conditions, success conditions, or explanation of their moral bindingness.³

This principle in turn aids in the formulation of another important thesis: (Moral Necessity) In order to secure the provision of a range of important human goods, we require a web of interrelated institutions. Institutions, which

³ I will talk as if conventions form a single unified web. It doesn’t affect my argument if e.g. there are two large but distinct webs.

  

117

define the parameters of social roles, are in turn constituted by clusters of conventional norms.

Moral Necessity has already been defended widely in both classic and contemporary moral theory. I will only briefly summarize some of these efforts here. It has been argued, for example, that institutions are a precondition for having morally valuable interpersonal relationships (Hardimon 1994; Scheffler 2001); that they support our exercise of autonomy and self-determination (Rousseau 1997 [1762]; Rawls 1971; Sandel 1982; Kymlicka 1989; Blum 1994); that they help “settle” or “determine” natural or moral law (Locke 1960 [1689]; Pallikkathayil 2010); that they decrease the likelihood of making “moral mistakes”; that they decrease the moral cost of discharging our general moral duties; that they help us discharge our duties of beneficence by solving coordination problems, efficiently dividing the moral labor, and encouraging compliance by establishing relationships to which people have deep emotional attachments (Sidgwick 1981 [1874]; Nagel 1979; Held 1984; Pettit and Goodin 1986; Hardin 1996; Simmons 1996); and that such institutions have “adversarial” justifications (Smith 1937 [1776]; Madison 1987 [1788]). Moral Necessity depends on the plausibility of some constellation of the above arguments. Many of these theorists also seem to agree that the extant system that we have is (a) “good enough” to secure many important human goods, and thus a better moral alternative than no institutions at all and yet (b) the system that we have is imperfect in many respects, and may well admit of improvement through deliberation, planning, and activity. (For example, it would be better if men routinely did more housework, and hospital cleaning staff were paid more.) We thus have: (Moral Sufficiency) The actual extant system of institutions that we have (family; law; schooling; emergency services; etc.) does in fact secure a range of important human goods, and is a moral improvement upon having no such system at all. (Systemic Imperfection) The actual extant system of institutions that we have is morally flawed in many respects, and we may be able to improve it.⁴

Taken together, the general approach to convention that I have outlined here, together with the nine theses that I have formulated and motivated in this section, constitute a minimal theory of conventional norms that I hope will be widely ⁴ I’ve noted two ways that conventions can be morally problematic, expressed by Norm Imperfection and Systemic Imperfection. Norm Imperfection concerns individual conventional norms that are morally problematic in themselves, such as anti-miscegenation laws. Systemic Imperfection, on the other hand, applies to the system as a whole. There is a difference between a slur and an insulting sentence made up of words that may not be themselves offensive. Similarly, there is a difference between a morally obnoxious convention and a system of conventions that is open to criticism in virtue of the way its parts work together.

118



endorsed. Let us now turn to the question of whether the resources of such a minimal theory allow us to derive any more substantive claims. I will then return to the question of how this might help secure their moral advantages while avoiding the costs we explored in the previous section.

5. Building up: Accessibility, Regularity and the pro tanto Once we reflect on the definition of convention—social practices that people follow in part because others do—two plausible substantive constraints immediately suggest themselves. It seems that for conventions to arise and be sustained, and as a broadly transcendental condition for their doing so, the conventional norms they generate must conform to the following: (Accessibility) The applicability and success conditions of a conventional norm will not depend on features of a situation that are unknowable to typical participants in a straightforward way. (Regularity) The applicability and success conditions of a particular conventional norm will not depend on features of a situation that are overly unusual or atypical with respect to the convention.

The rationale for Accessibility is that conventions are social practices that people follow at least partly because other people follow them, too. So if a norm is conventional, participants must generally be able to tell when others are conforming to the convention and when they aren’t. Otherwise the convention would dissolve. Perhaps there are moral norms that guide action in a detailed way despite being inaccessible in the above sense, but such norms are not conventional. I will assume that internal mental states, insofar as these differ from outward indicators, are typically not accessible to participants. Conventional moral duties will therefore not normally depend on them. This does not express skepticism about other minds, but only the commonsense view that in many cases our access to others’ motives, intentions, and other internal mental states typically proceeds straightforwardly through outward indicators such as testimony or behavior, however subtle and nuanced. The rationale for Regularity is twofold. First, a regularity of behavior is part of the means by which conventions structure the expectations of participants. Absent a code stipulated in advance (through legislation, say), conventional norms require that people who meet their applicability conditions regularly conform with their success conditions. It is due to this regularity of conformance that participants know what duties attach to various social roles. But features of participants’ situations that are overly unusual have not been encountered sufficiently often to generate conventionalized responses. The applicability and success

  

119

conditions of conventional norms will therefore not normally depend on these features. Second, as noted, many such norms attach to individuals through the social role they occupy. For the most part, the contours of these roles are already set out before typical individual participants occupy them. This is why the occupation of a role can help guide and explain a person’s behavior. There is no limit to the kinds of bizarre circumstances that can beset individuals, but the roles they occupy will typically have already been determined in a way that is insensitive to individuals’ unusual situations. These two reasons taken together explain why the success and applicability conditions won’t depend on features of a situation that are highly unusual with respect to the convention. Accessibility and Regularity together arguably predict and explain a familiar feature of conventional moral duties: they will bind merely pro tanto. Pro tanto moral duties (called by others prima facie duty,⁵ pro tanto reason,⁶ ground of obligation,⁷ or simply a consideration that counts in favor of an action⁸) give us moral reasons for action even though these reasons may be overridden—but not nullified altogether—by other moral considerations. They contribute to deliberation about what to do, but are not dispositive of what a person ought to do allthings-considered. We can thus consider: (Pro tanto) If there are any conventional moral duties, then they only bind pro tanto.

This thesis follows from Accessibility and Regularity, taken together with the fact that what one ought to do all-things-considered may depend on inaccessible features, such as one’s private intentions and motivations, or one’s unusual situation. Conventions cannot take our secret agendas (as per Accessibility) or unique circumstances (as per Regularity) into account, but the domain of ethics is more inclusive, often rendering a more precisely tailored verdict as to what we ought to do all-things-considered.

6. When (and Why) Are Role Obligations Morally Binding? In this section I aim to complete my account of the bindingness of conventional moral duties, and thereby of the typical array of role obligations that guide us. I will examine three more elements of a theory of moral convention, that I will label Global Inheritance, Presumptive Bindingness, and In Situ Holism. These theses are almost certainly less familiar and more contentious than the ones we ⁵ Ross 1930, 19–22. ⁶ Kagan 1989, 17–22. ⁸ Scanlon 1999; Dancy 2004.

⁷ Herman 1993.

120



have discussed. Nevertheless, I believe that when they are properly understood, a compelling case can be made for their acceptance. Most notably, when taken together with the earlier elements of the view, they seem to provide an account of how it is possible that conventional norms could morally bind and guide us. Let us consider them one by one.

Global Inheritance The three substantive constraints described in section 5 can be strongly motivated just by reflection on the nature and essential preconditions of conventional practice itself. The next claim I want to consider does not have this feature. It is closer to a methodological recommendation or Davidson-style “bold conjecture.” It says the following: (Global inheritance) The moral bindingness of individual conventional norms is derived or inherited from the moral good standing of the whole system of such norms, of which they are functional parts.

The question here concerns the primary unit of moral good standing when it comes to convention. What parts of the social web of conventions do we first establish in moral good standing, en route to explaining the good standing of the others? Individual conventional norms? Some internally unified sub-regions of the network, such as particular institutions or social roles? The web as a whole? Global Inheritance maintains that the final answer is correct. If individual conventional norms are binding, then that fact is to be explained by the good standing of the whole web. I do not have a knockdown argument in favor of Global Inheritance to hand. However, we might consider the following by way of preliminary support, and then let the acceptability of this thesis stand or fall with the fruitfulness of its consequences. Consider the following trilemma. Either (a) conventional norms are not morally binding, or (b) their bindingness derives from the moral good standing of the holistic web of which they are a part, just as Global Inheritance describes, or (c) their bindingness can be derived on a case-by-case basis from a small number of foundational principles, independently of the good standing of the whole. The first option is ruled out by Moral Necessity and Moral Sufficiency. Since I am presupposing the truth of the minimal theory for the purposes of this chapter, we can eliminate that option. That only leaves the final, “positive” approach as an alternative to Global Inheritance. Many versions of that approach resemble the norms-as-theorems model that I described in section 3. Individual conventional norms relating to, for example, parenthood, professional roles, or citizenship are considered one by one. In themselves they have a kind of provisional or candidate status, until we see how

  

121

to derive them from something more morally fundamental. The derivation strategy comes in two main flavors. On the first, an individual norm is morally binding because it is a consequence of some small set of moral principles.⁹ On the second, it is morally binding because it stands in the right relation to some social role in good standing, such as parent or doctor. Two aspects of convention raise problems for these strategies. The first is the possibility of Lewis-style “multiple equilibria,” where coordination admits of a range of jointly inconsistent but individually acceptable solutions. The second is expressed by the Interrelatedness thesis above. To see how these motivate Global Inheritance, let’s consider an analogy between the morality of convention and the morality of law. We typically show that individual laws are morally binding by showing how they relate to an institution, namely law, that is in moral good standing (Raz 1979). Why is this? Why not try to justify individual laws one by one by appeal to moral principles, or seek to vindicate legal substructures (tax law, immigration law) and then hold that particular laws are morally binding when they belong to those substructures? The first strategy runs straight up against the problem of multiple equilibria. On the view at issue, showing how an individual law is morally binding is demonstrating how it follows as a consequence from binding moral principles. But normally this would be a bad result, since there are lots of morally satisfactory but mutually exclusive ways of drawing up laws. We could cut the lowest band of tax off at $10,000 and offer $500 of credit elsewhere, or at $10,500 with no credits. Both might be morally fine as laws, and so it would be a bad thing if we could show that either was morally binding all by itself by citing moral principles. Why not evaluate the moral point and purpose of tax law in toto to avoid this problem? The correlate of Interrelatedness shows why. The component parts of tax law will be inextricably bound together with lots of other areas of law—laws about who counts as a citizen, who counts as a legal spouse, what it means to own property, etc. We should expect that the moral good standing of tax law is thus unlikely to be independent of say, the morality of spousal relationships, or property ownership. So, neither the atomic nor the “local holism” approach seems very promising. Unsurprisingly, then, most jurisprudential approaches seek first to vindicate “the moral authority of law,” and then try to account for the moral status of individual laws and sub-specialties by showing how they relate to that. Something similar seems likely to arise in the case of the web of conventional norms. There are lots of individually satisfactory but mutually exclusive webs. It would thus be bad if we could show that any individual conventional norms followed from some morally binding principle. We should only want to say that the individual parts of the system are morally compulsory insofar as they are part of an overall system that has moral authority. Again, we should expect that

⁹ I describe how I am understanding such principles in section 2 above.

122



Interrelatedness rules out the prospect of informative moral reductions of substructures, such as social roles. Because the moral justification of the rights and responsibilities of a parent will turn on whether it is okay to ask certain things of school nurses, co-citizens, school bus drivers, etc., the prospects of an informative account of the former that don’t advert to the latter seem dim.¹⁰

The Nature and Scope of Presumptive Bindingness With the theses that we have in place so far, we can begin to argue that conventional norms are morally binding by default. In particular, I will now aim to motivate the following claim: (Presumptive Bindingness) A particular conventional norm is morally binding unless it violates, all-things-considered, overriding moral considerations in typical cases according to the best moral theory. Conventional norms that fail this test lack moral force and are not even pro tanto morally binding. Thus, they are not conventional moral duties.

There are four points we should consider here about the nature and scope of Presumptive Bindingness. The first is that, in intuitive terms, adding the above thesis to our substantive conventionalist theory presents us with the following picture. Human society is structured by a complex web of interrelated conventional norms. Such a structure is morally necessary for the provision of many important goods, and while imperfect, it is better than no such system at all. We should, no doubt, rebuild large sections of our moral raft, but we can permissibly and effectively do so while standing on the planks that history has left us. The moral legitimacy of the individual conventional norms that act as functional parts of the system enjoy a default good standing inherited from the legitimacy of the system as a whole. But that default good standing can be overturned if it is demonstrated that the norm violates, all-things-considered, overriding moral considerations in typical cases according to the best moral theory. Recall that for the norms-as-theorems model, a conventional norm lacks moral bindingness in the absence of a positive account of how it has acquired it. In my account, in contrast, a conventional norm possesses moral bindingness in the absence of a compelling account of why it lacks it.¹¹ ¹⁰ We can also say something positive about the role of conventions that explains why, like law, they enjoy this form of authority. See Raz 1979 on the way that authoritative institutions (i) put us in reliable touch with right reasons without (ii) asking us to first determine the layout of those reasons. ¹¹ Strictly speaking, what matters is that there is an account, not that we have it. But in practice, Presumptive Bindingness seems to commit us to viewing conventional norms as morally binding unless the burden of proof of showing them overturned has been discharged.

  

123

Second, we should note that Presumptive Bindingness states a sufficient condition for nullifying the moral bindingness of a conventional norm. There may well be other equally sufficient conditions for overturning the default bindingness of such social conventions. For example, it might be sufficient to overturn them if we have a demonstration that they are redundant add-ons that contribute nothing of moral importance and that it would be harmless to excise from our practice. Alternatively, it might be clear that although a given norm does not violate any overriding moral considerations, it could be harmlessly improved, without thereby imperiling anything of moral significance in typical cases. (This is structurally suggestive of something like a notion of supererogation for moral conventions.) Finally, it may also sometimes be the case that we know that our conventions issue us with conflicting directives that cannot all be complied with. I argue in my “Irreconcilable Differences” (Taylor 2013) that such cases of irreducible moral conflict are possible, but must be the exception rather than the rule. There may also be additional forms of sufficient condition that are of interest, but I will not pause to explore the issue further here. Third, Presumptive Bindingness presents evil conventions as negated ab initio, and thus failing to bind even merely pro tanto. Why does it take this form? First, if they bound pro tanto even though routinely overridden all-things-considered, then this would be too demanding on moral agents, who would be doomed to moral failure despite doing as they ought. No plausible moral theory will issue such a view. Second, if they bound pro tanto but were almost invariably overridden, then that would seem to saddle a theory of moral convention with a redundant, useless theoretical wheel. This seems especially true given that standard indicators of the pro tanto, such as moral residue, do not seem present in the case of participants in evil conventions. Fourth, Presumptive Bindingness presents conventional norms as morally binding by default, but this does not entail that they are inescapable, nor that failure to conform to them should invite the same responses in every case. Regarding inescapability, such norms may bind school bus drivers pro tanto in certain cultural circumstances without forcing anyone to take on that role. Regarding failure to conform, turning up 45 minutes earlier to a dinner party might violate a conventional moral duty, but will not deserve the patterns of opprobrium or punishment that more serious forms of moral failure, such as operating on a patient without their consent, might invite.

Why Should We Believe Presumptive Bindingness? Why should we believe Presumptive Bindingness? Given Moral Necessity, Moral Sufficiency, and Global Inheritance, we should view the individual components of our web of conventions as binding by default. If (i) we need some network of

124



conventions or other for strong moral reasons, and (ii) this one is good enough to serve for now, and given that (iii) individual components inherit their good standing from the whole of which they are functioning parts, then we should expect that they possess this status unless there is some suitable moral reason to think badly of them. (If you morally require a car, and you have a car that is morally adequate, and it would be very difficult and risky to secure a completely different car, then the functioning parts of your current car are things that you have good reason to think well of from the moral point of view, even if there is no self-standing story that takes you from moral theory to a demand that you own just that part.) But Imperfection tells us that we will sometimes have such reason. So we need a method of weeding out the evil conventions. We know we can’t do anything that would imperil the system itself or anything else of non-negotiable moral value. So what should we look at? It seems very natural to answer that we should look at the impact of the functioning of that part of the system in the actual world. But since the relevant part of the system is a conventional norm, it functions to establish and sustain a social regularity, defined by its applicability and success conditions. (These kinds of agents are responsible for bringing about these kinds of outcomes in these kinds of ways.) By Accessibility and Regularity, we know that these conditions can’t involve conditions that are overly unusual, atypical, complex, or unforeseeable. So to determine whether the component conventional norm is bad enough to be stripped of its default good standing, we should see whether it involves a serious enough moral violation in its typical operation in the actual world. Thus, for example, we will not evaluate the bindingness of actual conventional norms by seeing whether they would be unacceptable in science fiction or other kinds of unrealistic scenarios. Even if our conventional parental norms toward children were unacceptable in worlds where children resulted from “people seeds” drifting through windows (Thomson 1971), this would have little bearing on the moral evaluation of these norms in the actual world, where children do not grow in that manner. This is just what Presumptive Bindingness asserts. Quite apart from this functional approach to the components of morally necessary systems, we can justify Presumptive Bindingness by its fruits. As I argue in section 7, it helps us see why moral agents are entitled to the epistemic and practical benefits of convention.

In Situ Holism What resources should we appeal to in evaluating whether the default good standing of conventional norms is overturned? Where should we look to determine whether they bind by default, or lack moral force? I suggest the following:

  

125

(In Situ Holism) The moral bindingness of a particular conventional norm is determined by the overall impact of its observance on those who participate in, or are otherwise affected by, the actual interrelated web of institutions of which that norm is a part.

Although In Situ Holism directs us to look to a norm’s overall impact in the actual world, this need not presuppose a consequentialist moral theory. Thus, the overall impact of a particular norm may include: (i) whether it is overly demanding for typical participants; (ii) whether it excludes classes of people from the benefits of social cooperation; (iii) whether it is exploitative; (iv) whether it gives improper weight to typical participants’ autonomy; (v) whether it harms typical participants; or (vi) whether it is inconsistent with our general moral duties or foundational moral values. Thus, In Situ Holism is concerned about impact only in the minimal sense that any plausible moral theory is concerned about it.¹² Why should we accept In Situ Holism? Moral Necessity tells us that we need some web of conventional norms. Moral Sufficiency tells us that the extant system is better than having no system at all. Global Inheritance tells us that the moral bindingness of particular conventional norms is explained by the overall legitimacy of the system they belong to. Presumptive Bindinginess tells us that they will bind us unless they violate overriding moral considerations in typical cases in the actual world. So, it is quite natural to think that in evaluating such individual norms we should look to those actually affected by its typical observance in the overall system. Consider what the alternatives would be. Particular conventional norms might require positive vindication on a case-by-case basis. But that would be to pursue the positive atomist project, which Global Inheritance already counts against. If these norms aren’t morally binding, then nothing would determine their bindingness. But that plausibly conflicts with Moral Necessity and Sufficiency, on the assumption that a morally essential whole cannot be composed solely of morally neutral parts. So, it seems that the primary rival to In Situ Holism would be a view on which some kind of local, limited holism fixed the moral bindingness of conventional norms. On my view, the conventional web as a whole is the primary locus of moral evaluation, and then individual conventional norms bind by default because of the good standing of that. But an alternative approach could be tripartite in form. The moral necessity of the overall network could imbue social roles and institutions with a default good standing, and then individual conventional norms could be justified insofar as they were component parts of

¹² In Situ Holism directs us to the actual world, but considerations of, e.g., risk will plausibly demand that we consider things that could easily happen, whether or not they do. I am presupposing that an actualist theory can finesse this worry by appeal to actual patterns of objective chances, evidential probabilities, etc.

126



social units that are non-excluded. The moral bindingness of such norms would then presumably be determined by the moral status of the social unit of which it is a part.¹³ The most common version of this view in the literature selects social roles as the units in question. These provide an evaluative unit larger than the individual conventional norms, but smaller than the web as a whole. For example, to see whether a particular professional norm morally binds doctors, we look to the moral point and purpose of doctoring. Theorists employing this strategy may of course have different views about the moral point and purpose of particular social roles. This general strategy is unsatisfactory for two reasons. First, particular social roles are mostly “mixed bags” that lack one readily identifiable moral point or purpose. This can happen when a single role has multiple, potentially conflicting motivations and values (as when a university professor plays the role of educator, researcher, student advisor, employment referee, departmental and university colegislator and co-administrator, etc.). It is also unclear that there is any such purpose attaching to some morally important roles (sibling, spouse, sangoma). In such cases, it seems unpromising to seek to derive the legitimacy of individual conventional norms from the distinctive point or purpose of the role involved. Second, the moral point or purpose of a particular social role is often insufficient to determine the moral bindingness of the obligations attached to it. Morally binding role obligations may attach to bad social roles. Consider the role of guard at an internment camp providing the slave-labor of innocent prisoners. Let us assume that the best moral theory says that the overall point and purpose of this role is morally obnoxious; it is an overall evil role in the service of an overall evil institution. Nevertheless, morally binding conventional norms may attach to this social role, such as the norm to alert a medic if a prisoner falls ill.¹⁴ The fact that the guard has a duty to alert a medic in such circumstances is a conventionally specified way of discharging a general duty of beneficence. Had he another occupation (such as cook) in the same camp, he may have been required to discharge his duty of beneficence in another way.¹⁵ Therefore, alerting a medic in this case may be a conventional moral duty, despite the overwhelmingly and irredeemably evil nature of the role to which it is attached. Conversely, there may be illegitimate role obligations that attach to morally good social roles. For ¹³ I here consider a local holist approach that endorses Global Inheritance and Presumptive Bindingness. As a matter of fact, most social role-based accounts do not proceed in this fashion, but instead seem to endorse something closer to the norms-as-theorems approach. On that view, social roles are only justified if some positive account of their moral good standing can be derived from moral principles or values. ¹⁴ This norm would be conventional in my sense even if it relates to a general duty of beneficence. As per my discussion in section 3, the norm is conventional because an institution is at least part of the means by which the duty devolves to the guard. It also plays a justificatory role in specifying what exactly the guard should do. ¹⁵ Recall that the ability of institutions to “divide the moral labor” in this way is one reason for their moral necessity.

  

127

example, the demand that wives take on primary responsibility for housework, cooking, and childcare might be unfair and overly demanding. It is therefore better to view the total web as the primary unit of moral evaluation, and then excise individual conventional norms—individuated by their Applicability and Success Conditions—from that. To be clear, although I take the above considerations to motivate In Situ Holism, I do not intend them as a cast iron establishment of its truth. Rather, as with Presumptive Bindingness, we can judge In Situ Holism on its theoretical explanatory benefits. I will turn to some of those now.

7. Nice Things Regained Let’s examine how the view I’ve recommended helps us remove the obstacles to enjoying the epistemic and moral-theoretic benefits described in section 3. I will consider these one by one, and in reverse order, but the general motivation for the view is intended to be integrated and cumulative.

Norms-as-Theorems Model The norms-as-theorems approach to role obligations has two notable features. First, the issues are typically discussed in a locally holistic manner. We describe the contours of a social role, and then look to its moral point or purpose in arguing that their associated conventional norms are morally binding. My approach is in one sense more atomistic than this, but in another sense more global. On my view the primary locus of moral evaluation is typically individual conventional norms, not roles.¹⁶ In evaluating whether a particular conventional norm is morally binding, however, I suggested that we should view it as (i) morally binding by default unless (ii) it violates overriding all-things-considered moral considerations in typical cases in the actual world. This has two consequences. First, the burden of proof is shifted. Conventional norms are no longer viewed as merely provisional in the absence of a vindicating explanation of their legitimacy, as the norms-as-theorems model has it. Rather, we should view them as binding unless there is an explanation of why they are non-binding. We thus avoid the kind of Williams-style loss of moral knowledge that the norms-as-theorems model

¹⁶ Because these norms are individuated with respect to their application and success conditions, they are quite finely distinguished. Thus the convention that motorists drive on the right in the US, and the convention that Honda owners drive on the right in the US, will count as two different conventions for the purposes of the evaluative procedure described by Presumptive Bindingness and In Situ Holism.

128



threatened to inflict. We can’t move from our inability as philosophers to morally vindicate the demands of a given norm to an enforced agnosticism about whether the norm actually binds us. Secondly, we are free to look beyond the point or purpose of a role in evaluating whether any of its attached rights and duties are indeed negated ab initio. Instead, we are to look in a holistic way at the overall impact of a norm’s observance on those who are affected by the actual interrelated web of institutions of which the norm is a part.

Explanation and Action-Guidingness This two-stage approach—individual conventional norms are binding unless they violate overriding moral considerations in typical cases in the actual world—helps us resolve the tension between explanation and action-guidingness. That tension, remember, was that as moral theorists we are apparently committed to two different desiderata for moral theory that will foreseeably often conflict. On my view, action-guidingness is primarily delivered by conventional norms that bind us by default. These norms are nuanced and contextually sensitive, just as actionguiding requires. But because conventions obey Accessibility, they are also guaranteed to take a form where participants can know who the convention binds and what it directs them to do in typical cases, and Regularity ensures that most cases are typical. When a convention is non-binding, there will always be an account that explains why this is. Often this account might involve citing a general moral consideration that the convention violates—indeed, for public policy reasons this will often be morally important. These statements of value will foreseeably be formulated at the level of generality and fundamentality that our interest in explanation demands, and it will often be illuminating to see how they apply in striking down morally bad conventions.

Formulating Principles However, what is required to strike down a conventional norm is only that it violates overriding moral considerations. There is no requirement that we must be able to formulate a wholly general, exceptionless, and quibble-proof principle from which these considerations flow. In this sense, the rejection procedure for conventional norms is wholly compatible with moral particularism (Dancy 2004; Hooker and Little 2000). It may be that we can simply see that no adequate moral theory could allow such a convention, and give some reasons why this is the case, without those reasons being codifiable without loss in some much more abstract and general way.

  

129

Moreover, because we evaluate conventional norms in situ, with an eye to their overall impact on those they affect in real-world typical cases, many of the weird and wonderful counterfactual and idealizing sources of counterexamples will not be available to us, in ways that would threaten to complicate or gerrymander moral principles. The twin demands of action-guidingness and explanation are better served by employing an “8∃ rather than a ∃8” structure in our explanation. For any rejected norm there is some story or other about why it is struck down, but there is no requirement to give a single account that covers every case of rejection. Our inability to formulate principles that are extensionally adequate, explanatory, and command widespread acceptance thus raises less of a problem. If it turns out that such a set of principles is forthcoming, then my approach can benefit from explanatory appeal to them. But in their absence, action-guiding normative direction is still available.

Fundamental Conflict Unsurprisingly, my theory does not remove nor resolve all sources of moral conflict. However, it does allow us to sweeten the bitter pill of fundamental clashes of value. First, the burden of proof is shifted onto those who would appeal to competing values in seeking to reject some apparently binding norm. One cannot merely gesture at how hard it is to provide a vindicating explanation via a suitably plausible and resilient moral principle. Secondly, the means by which conventional norms are struck down are notably restrained. In Situ Holism counsels us to look holistically at the real-world effects on actual practices and participants. This will serve to exclude far-fetched hypothetical or counterfactual analogies, weird counterexamples that trade on cases that we never encounter and don’t care about, and all the rest of the disreputable toolkit of the working analytic ethicist. Of course, that will not resolve Singer-style worries about whether we should spoil our children on their birthdays in a world where many do not have enough to eat. But it would be a bad feature of an account that it did rule out such real-world and pressing moral issues by insisting on a certain choice of methodology.

Evil Conventions The fact that some conventional norms are morally bad prevents any simple identification of them with moral duties. My account deals with this evident fact by (i) having individual component conventional norms bind morally by default but (ii) strip all moral force from norms that are reprehensible. Because my account evaluates conventions in a finer-grained manner than role-based

130



accounts—distinguishing, for example, between the convention that establishes camp guards and the convention that Guard A feeds the prisoners as prescribed— I can explain why even social roles that are overall evil and should be abolished may have some associated conventional moral duties that bind by default. The former convention that establishes labor camps plausibly violates, all-thingsconsidered, overriding moral considerations according to any adequate moral theory, but the latter directing Guard A to feed prisoners does not. Any plausible moral theory should both prohibit camp guards being established, and instruct a given guard to feed the prisoners as instructed even when he is feeling too lazy to do so.¹⁷ Because the account is compatible with both principle-based and particularist explanations of the moral illegitimacy of certain conventional norms, it also has the following advantage. We can explain why the victims of morally obnoxious conventional norms often possess a greater range of moral knowledge than those who support or facilitate its existence. The victims are often in a position to cite or recognize reasons why the norm should be negated ab initio, and that is sufficient to undermine its force. It isn’t necessary that they should be able to formulate some bulletproof explanatory generalization about why that is the case, and defend it in the light of far-fetched analogies and weird counterfactual circumstances.

8. Conclusion I have appealed to a substantive theory of convention to argue that the demands of social roles morally bind us by default. If this is correct, then two significant advantages ensue. First, we can explain why our social roles provide a reliable source of moral guidance. Second, we have a method for assessing those rolebased demands with a view to reshaping them in light of the good.¹⁸

References Aquinas, St Thomas. 1989. Summa Theologiae: A Concise Translation. Timothy McDermott (ed.). London: Eyre and Spottiswoode. Blum, Lawrence A. 1994. Moral Perception and Particularity. New York: Cambridge University Press.

¹⁷ See Aquinas’ discussion on antecedent and consequent will at Summa Theologiae, Ia, Q. 19, A. 6. ¹⁸ This chapter has only concerned itself with the epistemic and practical benefits of moral guidance that is provided by extant conventional norms. However, changes in historical and cultural circumstances, and the moral obligation to remedy imperfections in our extant web of conventions, plausibly require that we implement and support new kinds of conventions. The material in this chapter is not intended to address this general issue. I consider it in depth in a sister paper, “Taming Wickedness: Towards an Implementation Framework for Medical Ethics” (Taylor 2022).

  

131

Burge, Tyler. 1975. ‘On Knowledge and Convention.’ The Philosophical Review 84 (2): 249–255. Dancy, Jonathan. 2004. Ethics Without Principles. Oxford: Clarendon Press. Hardimon, Michael. 1994. ‘Role Obligations.’ The Journal of Philosophy 91 (7): 333–363. Hardin, Russell. 1996. ‘Institutional Morality.’ In The Theory of Institutional Design. R. E. Goodin (ed.), New York: Cambridge University Press. Held, Virginia. 1984. Rights and Goods: Justifying Social Action. New York: Macmillian. Herman, Barbara. 1993. The Practice of Moral Judgment. Cambridge, Mass.: Harvard University Press. Hooker, Brad and Little, Margaret. 2000. Moral Particularism. Oxford: Clarendon Press. Kagan, Shelley. 1989. The Limits of Morality. Oxford: Oxford University Press. Kymlicka, Will. 1989. Liberalism, Community, and Culture. Oxford: Oxford University Press. Lewis, David. 1969. Convention: A Philosophical Study. Cambridge, Mass.: Harvard University Press. Locke, John. 1960 [1689]. Two Treatises of Government. Cambridge, Mass.: Harvard University Press. Madison, James, Hamilton, Alexander, and Jay, John. 1987 [1788]. The Federalist no. 10 in The Federalist Papers. Harmondsworth: Penguin. Marmor, Andrei. 1996. ‘On Convention.’ Synthese 107 (3): 349–371. Miller, Seumas. 2001. Social Action: A Teleological Account. Cambridge: Cambridge University Press. Nagel, Thomas. 1979. Mortal Questions. Cambridge: Cambridge University Press. Pallikkathayil, Japa. 2010. ‘Deriving Morality from Politics: Rethinking the Formula of Humanity.’ Ethics 121 (1): 116–147. Pettit, Philip and Goodin, Robert. 1986. ‘The Possibility of Special Duties.’ Canadian Journal of Philosophy 16 (4): 651–676. Rawls, John. 1971. A Theory of Justice. Cambridge, Mass.: Harvard University Press. Raz, Joseph. 1979. The Authority of Law. Oxford: Oxford University Press. Ross, W. D. 1930. The Right and the Good. Oxford: Clarendon Press. Rousseau, Jean-Jacques. 1997 [1762]. The Social Contract and Other Later Political Writings. Victor Gourevitch (trans.), New York: Cambridge University Press. Sandel, Michael. 1982. Liberalism and the Limits of Justice. Cambridge, New York: Cambridge University Press. Scanlon, T. M. 1999. What We Owe to Each Other. Cambridge, Mass.: Harvard University Press.

132



Scheffler, Samuel. 2001. Boundaries and Allegiances: Problems of Justice and Responsibility in Liberal Thought. Oxford: Oxford University Press. Sidgwick, Henry. 1981 [1874]. The Methods of Ethics Book 3 (7th edn.). Cambridge, Mass.: Hackett Publishing Co. Simmons, A. J. 1996. ‘External Justifications and Institutional Roles.’ The Journal of Philosophy 93 (1): 28–36. Smith, Adam. 1937 [1776]. An Inquiry into the Nature and Causes of the Wealth of Nations. New York: The Modern Library. Taylor, Erin. 2013. ‘Irreconcilable Differences.’ American Philosophical Quarterly 50 (2): 181–192. Taylor, Erin. 2022. ‘Taming Wickedness: Towards an Implementation Framework for Medical Ethics.’ Health Care Analysis. Advance online publication. https://doi.org/ 10.1007/s10728-022-00445-5 Thomson, Judith Jarvis. 1971. ‘A Defence of Abortion.’ Philosophy and Public Affairs 1 (1): 47–66. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge, Mass.: Harvard University Press. Williamson, Timothy. 2002. Knowledge and Its Limits. Oxford: Oxford University Press.

7 The Part We Play Social Group Membership as a Role Tracy Isaacs

1. Introduction This chapter explores the nature of social group membership understood as a role. Social group categories are classifications such as gender, race, class, sexuality, etc., that we use to identify social groups to which individuals belong. I understand roles as parts people play within a given collective context. The context also defines the roles. Role occupants have powers, obligations, and responsibilities qua role occupant. My question is: does considering social group membership as a role enable us to deepen our ethical understanding of oppressive social and political contexts and of what is morally required of us when we are living in such contexts? To the extent that roles help to define a range of powers, obligations, and responsibilities, they help us determine what we ought to do in a specific context. I argue that a role analysis of social group membership in oppressive social contexts can help people living within them to gain a more determinate understanding of their obligations with respect to the collective goal of eradicating injustice. I embed my discussion within a broader understanding of the ethical dimensions of collective contexts. Whatever else they are, roles are a function of collective contexts. Roles generate presumed obligations that govern what their occupants supposedly ought to do qua occupants of those roles. I qualify these as presumed because sometimes, for example in collective contexts of oppression that will be the focus of this chapter, the presumed obligations typically associated with a role may not align with what occupants of those roles actually ought to do. I understand oppression as structural injustice that affects people in virtue of their membership in social groups. Sex, gender, race, ethnicity, religion, disability, sexual orientation, economic standing, and class are among (though they may not exhaust) the dimensions along which this sort of unjust structural disadvantage might take place. Someone is or is perceived to be a member of these social groups on the basis of some feature they have or are perceived to have that is considered common to all members of the group. When the disadvantages that befall

Tracy Isaacs, The Part We Play: Social Group Membership as a Role In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Tracy Isaacs 2023. DOI: 10.1093/oso/9780192843562.003.0007

134



individuals as a result of their group membership are unearned and undeserved, built into the very structures of our institutions and the way they operate, it creates a social and political context of structural injustice that constitutes oppression. This type of injustice requires a collective analysis because oppression operates at the collective level of structures and institutions: it is in virtue of a collective feature—their membership in a social group—that individuals are negatively affected by oppression, not in virtue of their individual choices or behaviours. For example, Canadians of African or Caribbean descent have given accounts of ‘driving while Black’ (Loewen 2020) or ‘shopping while Black’ (Bousquet 2020), noting incidents of being pulled over for ‘random’ checks several times a month (more if they are driving what would be considered a ‘nice’ vehicle) or subject to security checks while shopping (Loewen 2020 and Bousquet 2020). In an oppressive social context, recognized and acknowledged social roles do not confer only structural disadvantage. They may also confer structural privilege, wherein the very structures of society operate to provide an inequitable distribution of opportunities and advantages to some over others, in virtue of social group membership. Being in a position of structural privilege grants a person more power and agency. However, as we shall see later, it also makes awareness of injustice more difficult to grasp, thus generating moral ignorance. This epistemological difficulty creates a moral risk of behaving wrongly, and in ways that are potentially inexcusable and therefore blameworthy. It enables individuals to take part, unknowingly, in widespread wrongful social practices that perpetuate the oppressive social structures and thus cause harm. Whereas someone facing structural disadvantage might, for example, experience ‘the glass ceiling’, someone possessing structural privilege might experience ‘the glass escalator’ (Williams 1992; Budig 2002). I am going to argue that in an oppressive social context, this sort of moral risk can create moral opportunities, that is, opportunities to act in morally good ways that promote social justice rather than perpetuate social injustice. In unjust social hierarchies, the presumed obligations associated with a social role may, pejoratively and objectionably, appear to be all about ‘knowing one’s place’ within that hierarchy and acting in ways widely considered appropriate to it. But we can replace this negative connotation of ‘knowing one’s place’ with the more hopeful idea of playing one’s part in a collective social justice effort. This revised understanding of what it means to occupy a role that is defined by social group category membership will challenge the presumed obligations that support an unjust status quo. Social group categories, particularly but not exclusively those of privilege, yield a range of obligations that create important moral opportunities for people specific to their role. If we move to a higher level of abstraction—away from specific social group categories and the roles they may suggest—we might understand occupying social roles of privilege and occupying social roles of disadvantage as themselves social

   

135

roles of a more general sort. When we take this view, we may use roles as a mechanism for understanding what part variously situated agents can and more importantly should play in addressing structural and systemic social injustice. In what follows, I explore the potential of this approach, and argue that it adds a powerful and hopeful dimension to understanding the actual moral requirements of members of different social groups in oppressive social contexts.¹ Section 2 considers what roles are, paying special attention to explaining that in addition to generating duties and obligations, they also generate power and agency. Section 3 takes up roles, responsibility, and social group membership, with specific attention to oppressive social contexts. This part includes a discussion of the controversial concept of ‘knowing our place’ in social hierarchies. Epistemological considerations borrowed from standpoint theory support the claim that different social roles carry with them different moral insights. Building on that idea, in Section 4 I show how to move from ‘knowing our place’ to ‘doing our part’ by considering how the collective context of oppression yields morally significant descriptions of seemingly benign individual actions. Finally, Section 5 combines these ideas to give an account of the way awareness of social roles, their power, and the differential knowledge they yield can generate moral requirements that can help shift oppressive social contexts into more just social arrangements. I first turn to a brief consideration of what a role is.

2. Roles and Their Features What Is a Role? I understand a role as a type of part that a person plays within a collective context. Roles are defined by the context in which they exist. Sometimes that context has structures in place that enable roles to be very well defined, as in an organization, where roles will have specific descriptions of responsibilities and obligations attached to them. Other times the context is looser, less structured, where occupants of roles need to learn it as they go; the role itself may evolve over time, such as roles within a family (e.g. the mother), within a group of friends (e.g. the organizer), or within a social setting (e.g. designated driver). But whether the role is well defined or more amorphous, it would not exist without the collective context that defines it. There would be no designated driver if there were no ¹ Zheng (2018) explores the question of the individual’s role in ‘changing the system’, where she too is concerned with structures of injustice and offers an insightful role-related analysis that focuses on ‘social’ roles such as mother, teacher, student, colleague, employer, citizen, etc. as ‘the site where structure meets agency’ (869). My focus on social group membership as a role takes this issue up at a different level, arguing that membership in social groups may itself be understood as a role and, as such, can help people to navigate their obligations for changing oppressive social structures.

136



laws against driving under the influence and no group outings to an event where most attendees drink alcohol. In that sense, roles are fundamentally products of a collective context. Whether the role is that of mother, father, sibling, babysitter, designated driver, brick layer, CEO, cashier, line chef, physiotherapist, doctor, or engineer—the roles we occupy have a normative dimension to them that constrains, delineates, and guides our choices. They generate obligations of a sort—whether presumed or actual—that both justify attributions of responsibility and make certain actions possible. For example, in her role as mother, an individual has obligations, choices, and authority with respect to her children—say, over where they will attend school or how much screen-time they will be allowed each day—that a babysitter does not have in the role of babysitter.² This feature of roles distinguishes them from other sorts of descriptions of people, such as ‘an introvert’ or ‘a blonde’. These differ from ‘a mother’ and ‘a babysitter’ in that they are simply descriptions (Jeske, this volume). They do not identify obligations or powers and do not place normative constraints on choices. The way roles constrain, delineate, guide, and empower is a key feature of them that has important implications, to be outlined in due course, for the analysis of social group membership as a role. Michael O. Hardimon (1994) has helped shape the discussion of role obligations with an influential paper in which he claims that roles are defined within institutions. Hardimon is right that some roles are defined within institutions. But institutions are just one type of collective context. For the present analysis, the pertinent point is that roles function within a collective context. The collective nature of roles embeds them and their normative requirements within structures. These structures provide organization of a sort, but institutional structures do not exhaust the types of collective contexts in which roles might function. Role ethics seeks to understand and explain the normative ethical features of roles. As such, role ethics assumes that besides the natural duties (Jeske 2001, 19–23) that all moral agents owe all other moral agents, there is a range of more specific, less generalizable duties that are tied to the roles we might occupy. As noted above, in professional contexts people enter into roles that require them to do specific things. This is not to say that every role embodies objective duties or obligations that ought to be upheld, regardless of what that role might be. There are countless instances in which the larger social context that defines a role is itself the product of an unjust social arrangement. Indeed, this is the case in the oppressive social contexts to which this chapter will be attending, where what is presumed to be required of a person in a role may differ from what is actually ² These are sometimes called ‘social roles’ (Zheng 2018). I avoid using the word ‘social’ here as a qualifier of ‘role’ so as to fend off possible confusion between my idea of social group membership as roles with the idea of ‘social’ roles such as mother, father, colleague, citizen. My emphasis is on social groups, and that makes my analysis significantly different from an analysis that focuses on what are often called ‘social roles’, even if all roles gain important features from a social context.

   

137

required. In those cases, we will see that the difference between ‘knowing one’s place’ and ‘playing one’s part’ is significant in redirecting role occupants towards the ways in which the power and authority of their roles might be used to address rather than perpetuate social injustice. A clearer understanding of this distinction first requires further attention to the idea of power and authority as features of roles. I turn to that now.

Roles, Power, and Authority One feature of roles that doesn’t get a lot of attention is the way in which, as they define a range of obligations, they also define the scope of a person’s authority and power. We might understand agents’ authority and power as the ability to have influence over outcomes or to exercise their agency over a wider range of areas because of their role. In a professional setting, when someone is in a managerial role, for example, their power and authority extend over a larger portfolio than their subordinates. We might contrast the CEO with the worker in the mailroom. The CEO has more influence over organizational decisions and actions, as well as control over subordinates, because the organizational structures authorize her or him in ways that those in lower ranks in the organizational hierarchy are not authorized. I have argued elsewhere (Isaacs 2011, chapter 5) that in organizational contexts, it is most frequently the case that people who have more power and authority within the organization—the senior executive team members—have a correspondingly larger share of responsibility for organizational functioning, including an organization’s moral functioning. This is not to say that their actions are the actions of the organization. But it is to say that they must be held to high standards of individual responsibility for their actions and omissions within that context. One reason for this higher standard of responsibility is that their very position enables them to act in ways that have a great impact on organizational actions. Moreover, the organizational machinery and structures give them more power in their roles. And these factors create opportunities to exercise their agency that they have only in virtue of their roles. It makes sense, therefore, that they are to be held to a higher level of responsibility for their actions within the organization than the people who are less powerful. In Section 3 the relationship between roles and responsibility is discussed in more detail. The point here is to motivate the idea that roles can generate power and authority, and that in turn suggests a higher level of agency and responsibility. This idea has great potential when we consider what may be required specifically of people in positions of privilege in an oppressive social context. In Section 3, I explain the way role-based power dynamics operate in such contexts very similarly to the way they do in the organizational contexts just described.

138



Before we get there, I would like to address the question of whether the normative dimension of roles suggests that they must be voluntarily chosen.

Must Roles be Voluntarily Assumed? If roles carry with them duties and responsibilities, some wonder whether they need to be voluntarily entered into and not just foisted upon people. We think of professional roles in this way, where people accept employment with a sense of what they are taking on and a willingness, demonstrated by their choice, to discharge the duties associated with that role. Perhaps that is an ideal scenario, but just because we would rather have choice does not mean that role occupation is always voluntary. For example, consider roles within a nuclear family. We may most frequently consider the role of mother to be chosen (though of course it is not always so), but a person is born into the role of daughter or sister. And though the duties and responsibilities of those roles might not be fully clear or consistent across all family units, there generally are duties and responsibilities associated with those roles. And while the demandingness of those duties might sometimes be conditional upon others doing their part as well, family roles are not strictly voluntary. It would be odd not to consider them roles of a sort, because they actually seem quite definitive in terms of someone’s obligations within a family. Most of us believe we ought to do more for our own family members than for the family members of our next-door neighbour. Moral philosophers often capture this under the umbrella of special obligations—that is, special obligations might be one way that the idea of roles and role obligations plays out. But whether they are or are not special obligations, there certainly do appear to be role-defined moral requirements that sometimes attach to roles that are not chosen. If we allow that not all roles are voluntarily undertaken, then that leaves open the possibility that membership in certain social groups (specifically those implicated in hierarchies of power, privilege, and oppression, such as those derived from categories of gender, race, class, or ability), which are also not chosen, might generate roles. If we understand roles in the way I have articulated, as having normative dimensions with respect to responsibility and obligations, then I suggest that understanding social group membership as a role provides a fruitful analytical frame for thinking about social group membership and its associated moral requirements in contexts of oppression. Thus, henceforth I assume that we can occupy roles that we do not volunteer to occupy, and that these roles may indeed generate obligations specific to their occupants. Up to now, the discussion has focused on features of roles, establishing that they gain these features from the collective contexts that define them. We have seen that roles generate obligations, whether actual or presumed, that their occupants are expected to fulfil. They also function to expand or restrict the

   

139

ways in which their occupants might exercise their agency and their range of powers and authority within the role and the realm or context in which it exists. This feature in turn can generate larger or smaller shares of responsibility within the relevant realms. And finally, roles may be voluntarily or involuntarily occupied. I now turn to a closer consideration of roles and responsibility, particularly in collective contexts of oppression, in order to bring together the ideas of social group membership, roles, and responsibility.

3. Roles and Responsibility Individual Responsibility in Collective Contexts The contextual features of roles outlined in Section 2 make them apt focal points for thinking about individual responsibility in collective contexts, which is where Section 3 begins. Often when we are faced with huge social or global issues that seem beyond the reach of individuals to fix, we turn our attention to collaborative or coordinated efforts instead. For example, climate change and global warming require a global solution. Similarly, the COVID-19 pandemic necessitated the coordination of local, national, and international responses, both at the outset and later, when vaccines became available to help bring it to an end. Social injustice and oppression are further examples of challenges whose solutions require working together. These situations, while requiring collective effort and action, also require individuals to ‘do their part’ if together we are to be successful in achieving the collective goals. For example, we have seen how people who resist getting vaccinated hinder the success of the collective project of ending the COVID-19 pandemic. I have argued elsewhere for the view that there are collective agents who are capable of collective actions, and I have also maintained that within collective contexts individuals also have obligations that are based on their place within the collective (Isaacs 2011). Individuals often see large problems as insurmountable when they consider how seemingly meaningless their own actions as individuals appear to be. This individualistic view of their own actions can lead individuals who would like to contribute to change to feel hopeless about their ability to make a difference. Thus, they do not know where to start. It can also provide less motivated individuals with a line of reasoning that justifies inaction. But this way of thinking overlooks that individuals can make a difference when they act in concert with or alongside others. Their actions can do harm and their actions can help. It is wrong to consider individual acts to be meaningless and without impact; such acts constitute important contributions when understood as part of the larger picture, together with the actions of many others. The remainder of this chapter explains how this idea applies in social contexts of oppression, using the idea of social group membership as role occupancy to

140



bridge the gap between the apparent futility of individual action in the face of structural injustice and effective social change. I suggest that if people living in oppressive social contexts understand their ‘place’ as possibly defined by their role within oppression, then they might understand too how to approach questions such as ‘what is my part in maintaining unjust social structures?’ and ‘what can I do as an individual, together with others, to advance a shared goal of creating social change?’ In short, considering social group membership as role occupancy helps to clarify a confusing and challenging moral landscape for moral agents who have genuine moral concern for the part they play and what they can do to make things better. It offers hope that their actions can make a difference. In the next section I discuss social group membership and some features of it, including some complicating epistemological features, as the next step in showing that a role analysis of it can do this analytical work.

Social Group Membership, Moral Knowledge, Moral Risk, and Moral Opportunity Recall that we are speaking of membership in social groups, which are a function of social group categories. Social group categories may include but are not limited to gender, race, class, sexuality, ethnicity, disability, and religion. Each of these categories yields within it a range of social groups. For example, women, men, and non-binary people are among the social groups that gender as a social group category yields; Black and White are among the social groups that race as a social group category yields; Protestant, Catholic, and Muslim are among the social groups that religion as a social group category yields; similarly concerning disabled and non-disabled with respect to ability; gay, bisexual, and straight among others with respect to sexuality; poor, middle-class, and rich with respect to class. In oppressive social contexts, unjust social arrangements result from systemic privilege and disadvantage that affects individuals not because of their actions or accomplishments as individuals but because of their social group membership and the location of their social group in social and political hierarchies. Social group membership in such contexts can create some epistemological barriers to understanding one’s part in injustice. Cheshire Calhoun has captured this in her discussion of moral ignorance and moral risk (Calhoun 1989). She maintains that in oppressive social contexts, widespread participation in wrongful social practices can put groups of people at moral risk because they participate in wrongdoing without realizing that it is wrong. Members of privileged social groups are especially vulnerable to this type of moral ignorance, and hence moral risk (Calhoun 1989, 389, 396). People are hesitant to acknowledge even to themselves that their social position is or even might be unearned. Indeed, people with structural privilege along any dimension are often quick to deny it,

   

141

not wanting to accept that they had a head start in virtue of the social group category or categories into which they were born. This is not to say that everyone equally benefits from a particular axis of privilege. But it is to say that there are some social groups for which to be a member is to have an advantage over members of other groups because of benefits associated with or conferred upon people who are (or even are perceived to be) members of that group. Standpoint epistemologists have maintained that social location can contribute to specialized knowledge, or at least make some knowledge more accessible to some people than others (Harding 1992; Wylie 2003). One of the reasons early feminists called for more ‘women’s voices’ or claimed that ‘women’s moral experience’ was excluded from traditional moral theories, and therefore that the theories fell short, was their belief that different perspectives and experiences yield different questions and facilitate different insights. Marilyn Frye, in her discussion of oppression, notes that Black people have to be conversant with the ‘ways of ’ the dominant social group (White people) as well as the more unique dimensions of their own experiences of navigating the world (Frye 1983, 1–16). The meaning of ‘Black Lives Matter’ grows out of this experience. When someone who is not Black responds with ‘all lives matter,’ it is a way of dismissing a legitimate claim of exclusion due to racist social structures. It should not need to be said that Black Lives Matter, because of course if all lives matter then by extension so do Black lives. However, the experience of Black Americans and Canadians shows that their lives appear to count for less in much the same way the experience of Indigenous Canadians indicates that they do not share equal social status with other Canadians. This structural disadvantage is not as obvious to those who do not experience it, thus we may claim that the occupants of these disadvantaged social locations have insights that others lack. Not only do others lack these insights, but they are also very difficult to convey to someone whose experience of the world is more seamless. Much has been written in recent years about micro-aggressions—seemingly small acts of discrimination or dismissal that are so apparently insignificant that it is difficult to say what’s wrong with them (Freeman and Weekes Schroer 2020; Fleras 2016; Rini 2021). But their effect is cumulative. If an Indigenous woman is interrupted once in a meeting then perhaps it is just rude. But if she is repeatedly interrupted, repeatedly not called on when she has something to say, repeatedly experiences no uptake on her contributions when she has an opportunity to speak, then this adds up. Some feminists have likened this phenomenon to ‘a ton of feathers’ (Davis, Joyce, and Murch 1991). Each feather may be lightweight and not hurt, but as the old riddle goes, a ton of feathers weighs exactly the same as a ton of bricks. It will still crush a person. Given that social location can affect our ability to grasp (or more easily grasp) certain moral truths, it makes sense to understand social group membership as placing people in different positions of moral awareness and moral knowledge.

142



Later this idea will form the basis of why membership in social groups may be usefully understood as a type of role. If positions of privilege carry with them moral risk because they often involve people in moral ignorance concerning their part in widespread wrongful social practice, they may also carry with them unique forms of moral opportunity. A ‘moral opportunity’ is an opportunity to fulfil a moral obligation or to play a part in a coordinated effort towards a morally worthy goal such as the elimination of oppressive social structures. This sort of opportunity can arise differently for different people, depending on their social group membership. The next sections will develop more fully the idea that people who hold social privilege have obligations that may differ from what is required of people who do not share that privilege. Thinking of social group membership as a type of role helps to explain how members of different social groups face different moral risks and have different moral opportunities and moral obligation qua their social group membership. If membership in social groups is a type of role, these obligations are role obligations. They exist because being a member of a social group with privilege (or by contrast social disadvantage) is best understood as a role that extends (or by contrast restricts) the range of agency and power of those who occupy it. Thus, members of social groups that experience systemic privilege are subject to particular sorts of moral risk in virtue of their social group membership and the obstacles it may generate to moral insights about the oppressive context itself. And the moral opportunities available to them to contribute to change are, because of their privilege, different from the moral opportunities available to members of less privileged social groups. This analysis helps to explain why the answer to ‘what can I do?’ must depend at least in part on a person’s social group membership because, understood as a role, social group membership generates obligations particular to the role. Earlier (Section 2) we saw that one important dimension of roles is that they enable or restrict the role occupants’ range of power and authority. Social group membership in contexts of oppression operates in the same way, offering us a further reason to favour a role analysis of it. Before moving on to Section 4, which outlines in more detail what we can expect to gain from a role analysis, I want to highlight the significance of power relationships between social groups in oppressive contexts. I turn to that now.

Social Group Membership, Power, and ‘Place’ Class oppression, race oppression, gender oppression, oppression on the basis of sexual orientation, religion, ethnicity, ability, economic status—these are all dimensions along which unjust social structures have normalized distributions

   

143

of power that define a person’s ‘place’. If power relationships can define roles in ways that are unjust, then people who hold more power may be morally required to play an active part in creating a shift to more just social and political arrangements. This obligation may arise in part from their ability to make a difference, given the power their privilege affords them, and may gain still further traction from the way in which they have benefited and continue to benefit from unjust social arrangements.³ One of the most fundamental elements in unjust social arrangements is the unequal distribution of power. Social and political inequality are structural, and unjust social arrangements empower some, creating more opportunities for them to exercise their agency than others. They also disadvantage some, creating fewer opportunities for them to exercise their agency, and even narrowing the range of their life choices in material ways by placing limits on what benefits and social or political arenas they can reasonably expect to have access to as they pursue their life plans. When social group membership creates default privileged and disadvantaged positions, it imposes implicitly and explicitly understood limits on people in virtue of their membership in certain social groups. To the extent that these limits assign people to different ‘places’ in a social hierarchy, and consistent with a role analysis, we might understand them as assigning unequal roles. This is not to say that all inequality or even all unequal power relationships are unjust. There may be circumstances in which someone has earned their authority or wherein it is necessary that someone be in charge and lead. For example, when a ship is in a storm, you want a captain making key decisions without being second-guessed. Not everyone can be doing their own thing because you need a coordinated effort. Similarly, a restaurant needs a hierarchy in the kitchen—the ‘chef ’s line’—because again, only a coordinated effort will do and someone needs to be in charge of it. In oppressive social contexts, however, these inequalities of the distribution of power are embedded in the social structures, working to the undeserved detriment of some and the unearned benefit of others. This is the nature of oppressive social structures, imposing glass ceilings and creating glass elevators. The structures and processes themselves empower and disadvantage, yield realities about who has a voice and who does not, and generate great opportunities for some, stunted opportunities for others. The normalization of structural benefit is often called ‘invisible privilege’ because of the extent to which it remains unrecognized as privilege. While members of privileged social groups may experience invisible privilege, members of disadvantaged social groups are expected to ‘know their place’. ‘Knowing one’s place’ has a pejorative meaning in social hierarchies of domination and subordination, and rightly so. It is very much associated with members of more ³ See Kisolo-Ssonko (2019) for a discussion of the perils, particularly for members of disadvantaged social groups, of considering social group membership as a role that generates role obligations.

144



subordinate, ‘lower’ positions on the social ladder being cognizant of ‘their place’, not being ‘upstarts’ who try to do things that they are not, because of the social norms and conventions, ‘entitled’ to do. In these contexts, membership in a subordinate social group functions as a role that generates normative social expectations for people to conduct themselves in ways considered appropriate to the social position they are perceived to occupy. These expectations may present themselves as presumed obligations that are not necessarily grounded in objective moral reasons. Nevertheless, they exert both internal and external, as well as implicit and explicit, social pressure on people to conform to them. Emmet Till, for example, was lynched for not ‘knowing his place’ when he came from Chicago to visit his cousins in the South in 1955. In offering this analysis, I am not suggesting that ‘the place’ that an unjust social hierarchy assigns is by any means just, justified, or right. Nor am I suggesting that the presumed obligations associated with these roles are actual requirements. On the contrary, the very facts of oppression make it the case that the social hierarchy is an unjust social arrangement that embodies and is built on structures of discrimination that keep people in disadvantaged social positions that create undeserved hardship while providing unearned entitlements and social and political power for others. The inequities are products of wrongful social practices. And the obligations not to transcend one’s ‘place’ are unjustly imposed. Despite the pernicious features of the dynamics of power and ‘social place’ that attach to social group membership under oppression, understanding social group membership as a role can deepen our understanding of actual moral obligations in contexts of oppression. This idea has great potential to help people gain a more determinate understanding of what they may contribute to a collective social justice goal. The presumed obligations people have in their respective ‘places’— be those privileged or disadvantaged—in unjust social arrangements may be replaceable by a different set of obligations, also connected to social position, understood in terms of roles and the associated power (or absence of power) different social positions afford their occupants. Section 4 takes a closer look at the potential of that approach.

4. Social Group Membership as a Role, Revised Assume that we can understand social group membership as a role. Roles generate presumed or actual obligations. What is actually required in a given role may differ from what is generally thought to be required. Moreover, social group membership positions people differently with respect to both their access to certain kinds of moral knowledge and the scope of their agency. Membership in different social groups, then, as roles, can yield different obligations, and these might depend in part on available moral knowledge and available opportunities

   

145

for agency under oppression. It is this idea as it applies to shifting out of oppression to a more just social context that I explore for the remainder of the chapter. Consider that a just society is objectively better than an unjust society. Moreover, citizens have an obligation to work for a more just society. Many people seek not only to uncover injustice but to work towards eradicating it. Taken together, these ideas offer a starting point for a different way of thinking about roles and their potential for ameliorating rather than perpetuating injustice and for seizing moral opportunities. The social justice objective of these obligations, insofar as they are obligations to eradicate an unjust social order, distinguishes them from mere ‘noblesse oblige’, wherein those of high social ranks are expected to act honourably and generously towards those less fortunate.⁴ While positions of privilege obligate in particular ways, we have seen that the epistemological conditions of normalization of wrongful social practice mean that members of privileged social groups are less likely, not more likely, to be aware of their place in sustaining unjust structures. Similarly, this makes it difficult for them to grasp what is required of them to address such injustice. While it is true that I am proposing that their positions require that they use their power and agency for good, the specific ‘good’ in view challenges the very social structures that afford them much of that social and political power. I now turn to an exploration of how understanding social group membership as a role can create moral opportunities in oppressive social and political contexts.

Revisioning Roles That Perpetuate Injustice as Roles That Ameliorate Injustice Recognizing that the idea of social roles carrying with them obligations can seem abhorrent because of its connection to ‘knowing one’s place’ in a social hierarchy where presumed obligations and the pressure on people to uphold them support the unjust status quo, I suggest that there is a more positive reading that we might impose on it. I have in mind that people in privileged social positions have obligations that they may not be realizing, and that recognition of the power afforded to them in virtue of their roles might actually indicate a further level of moral engagement in which they may involve themselves. In order to understand how this might work, we need to recognize how widespread ignorance functions in oppressive contexts. When injustice is structural, as in oppression, wrongful practices appear normal and their wrongness is not widely known. Ignorance abounds. As noted above, social and political location ⁴ Thanks to Alex Barber for asking how my suggestion differs from ‘noblesse oblige’ and ‘white man’s burden’.

146



can impede moral awareness. Therefore, people can take part in widespread wrongful practices without considering it wrong because the practices have been normalized to the point where people are ignorant that they are participating in wrongdoing. As an example, the human consumption and use of animal products is an instance of widespread and unrecognized participation in wrongdoing. The use and abuse of billions of animals per year for human nourishment and gustatory pleasure is so normalized as to be difficult to avoid. Though there are pockets of people who are vegan or vegetarian or otherwise ethically conscious in their approach to eating, they are typically considered to be on the fringe of normal eating in most parts of the world (which is not to say that there aren’t any cultures in which vegetarianism is the norm, such as Buddhist communities). I provide this as an example of a moral issue that is not widely acknowledged as such in ‘the mainstream’.⁵ A similar sort of widespread moral ignorance is at play with respect to racial discrimination against people of colour—Black, Indigenous, Asian—in a number of parts of the world, including the United States, Canada, Europe, and Australia. Even if we set aside racist individuals, data show that structural discrimination and disadvantage abound in these societies. If it is true that White people, even well-meaning White people, are positioned such that it is difficult for them to recognize structural racism when they are not on the receiving end of it, then the White majority will not be easily aware that they are taking part in social practices that sustain racist social structures. And yet many of these people may indeed welcome the possibility of change and would willingly take action if they felt hopeful that their efforts would contribute to social equality.⁶ In order to achieve that shift in perspective, from unknowing participant to engaged agent for change, it is necessary to create moral awareness in people who are not as easily disposed to recognize their part in maintaining oppressive social structures. This kind of shift is not so easily achieved, but it is not impossible. Moral progress is possible and does happen. Consider for example the shift in attitude concerning same sex marriage, now legal in Canada, the United States, Mexico, Australia, New Zealand, most European countries, South Africa, the UK, and several South American countries. The moral challenge for those who have moral information that others lack with respect to issues where widespread oppressive attitudes thrive in the mainstream, is to draw attention to the ways in which social structures discriminate. In so doing, it becomes possible to identify the ways that

⁵ We know it’s not widely acknowledged because if it were, then animal products wouldn’t figure so prominently on restaurant menus and family tables. ⁶ A relevant example is the current political context (Summer/Fall 2020), in which Black Lives Matter has become mainstream, and many White people are keenly aware of their privilege and truly seeking ways to be effective agents for change. The question ‘how can I be a good ally?’ comes up a lot. It’s as if a new level of understanding and insight has been achieved and people want to find a way forward, and understand the part they may play in that.

   

147

the everyday actions and attitudes of well-meaning and even morally engaged people contribute to upholding them, perhaps in whom they take seriously when hiring, how they respond to professionals who are not ‘what they expect’ (e.g. women or people of colour), or even what sorts of jokes they laugh at (or tell). Patterns of discrimination, when normalized in the ways described earlier, are difficult to uncover and eradicate. But, as noted, it is not impossible.

The Tension of Moral Burden and Moral Insight The epistemological situation, however, creates a tension. Considering roles related to social group membership at the most abstract level—that is, social roles that carry privilege and social roles that carry disadvantage—we see that the people least likely to recognize their part in wrongful social practice are the people who most need to gain awareness. On the flip side, the people most likely to recognize injustice, probably because they actually experience the disadvantages of structural injustice, are those who already bear the burden of structural disadvantage. If the social injustice and their epistemological position mean their role obligates them to be agents of social change, this outcome appears not only to increase but also to moralize their burden. They become morally obligated to take part in actions that address the very structural injustice that consistently puts obstacles in their path. Some might consider this to be an unfair consequence that should be avoided. However, if they have the insight required for progress, this added moral burden might be unavoidable. This issue comes up in the literature most frequently in the context of ‘the obligation to resist oppression’. Carol Hay asks ‘whether it is permissible to sometimes sit by and let oneself be oppressed’ (2011, 35). This formulation of oppression elides its collective nature and thereby represents it as something that befalls individuals and can therefore be overcome through individual acts of resistance. Though particular incidents can and do contribute to the collective context of oppression, these particular incidents are not strictly speaking what oppresses. Individuals are oppressed in a social context wherein, regardless of anything at all about them as individuals, they are subject to actions that discriminate, harass, exploit, etc. for no reason other than that they are members of an oppressed social group (or at least are read as such). Given the group character of oppression, resisting it need not be a matter of speaking up every single time an incident that contributes to, perpetuates, or is a consequence of a group’s oppression takes place. Because oppression is a collective condition, a person cannot ‘liberate themselves’ such that they are no longer oppressed without taking part in the larger process of collective action towards a shared goal of confronting and eradicating social injustice. To say that this shift must take place at the collective level is to say it must involve resistance and change to

148



the structures and practices that support and perpetuate the injustices associated with oppression. Here is where a role analysis might perform a useful function. Where the goal is a transition from oppressive structures to a more just society, we may ask what the respective roles of members of oppressed social groups and members of privileged social groups are. Here, it is not the presumed obligations of ‘knowing one’s place’ that are in question. Rather, a role analysis can help give shape to individual obligations within a collective undertaking for change that the different social positions yield. Section 5 offers an answer to that question, taking seriously the differences that social group membership imposes with respect to access to moral knowledge and access to power.

5. Roles and Shifting the Context In the spring of 2020, after the wide release of a video that captured Minneapolis police officer Derek Chauvin murdering George Floyd, a renewed discussion ensued about what it means for White people to be good allies in social justice causes concerning Black, Indigenous, and People of Colour (BIPOC). This question arises when members of privileged and dominant social groups who have an interest in being a part of social change recognize that they do not always know what is required of them to be effective. Men often ask how they can be good allies for women within the context of feminism. Non-disabled people often ask how they can be good allies for disabled people within the context of activism for accessibility. Straight folks often ask how they can be good allies for members of the LGBTQ+ community. This question of ally-ship is relatively recent and important because it indicates a shift in perspective. In the past, before this issue of how to be a good ally arose, social justice advocates who occupied privileged social and political positions often assumed that they had a good understanding of the situation and knew what was required to fix it. Much like the missionaries of the past, they considered themselves to know best what members of oppressed social groups might need or seek in order to live their lives well. But to make this assumption is to do what, as a privileged person in society, one has always done, namely, follow the agenda as one has set it. Moreover, it frequently involves ignoring how one’s privileged position might in fact implicate one in the very injustice one is attempting to resolve. Second wave feminists, for example, came under criticism for assuming that they were speaking on behalf of all women. Black feminist critics and others challenged them, paving the way for the importance of recognizing a more explicitly intersectional feminism (Hull, Scott, and Smith 1982; Lugones and Spelman 1983; Crenshaw 1991). This is not to say that privilege does not also sometimes yield specialized knowledge. Educational privilege, for example, clearly gives some access to knowledge that others lack. The

   

149

question of how to be a good ally, however, is a question of how to grant that in contexts of oppression, people who occupy more marginalized social positions may have insights that deserve, but have not been given (for reasons having to do with the conditions of oppression themselves, in which only certain voices are considered credible, authoritative, or plausible sources of knowledge), due consideration. The issue for those with structural privilege is not simply lack of recognition of their own role in eradicating social injustice. It is also an inability or difficulty seeing their own structural privilege (especially as it operates alongside structural disadvantage along other dimensions, such as gender) and how it is a part of the structural injustice that needs to be eradicated. To illustrate, it is not that White women have not endured sexism. Rather, their assumption that their experience is every oppressed woman’s experience erases difference and ignores the importance of an intersectional analysis (Crenshaw 1991). Intersectional analyses recognize that there are intersecting lines of oppression, and that different intersections can yield distinct forms of and experiences of oppression. A Rwandan woman selling fruit in the rural hills outside of Kigali does not have the same experience as a White woman practising law on the upper west side of Manhattan, even if the lawyer in Manhattan has faced a glass ceiling in her profession. An Indian woman practising family medicine in Old Delhi has a different experience than a White Canadian woman working at a fast-food counter. An Indigenous Canadian woman living in the northern territory of Nunavut does not have the same experience as a White elementary school teacher in a Vancouver suburb. Differences in geography, cultural context, economic standing, education, employment opportunity, race, class, sexuality, gender, ability, language, age, and any number of other possible social categories can combine to create different experiences and yield unique insights. I offer the question of how to be a good ally as a question about obligations within a role. The very question recognizes that not everyone is similarly positioned. People ought to avoid ‘speaking for’ others while recognizing at the same time that, when they are members of dominant social and political groups they have power, authority, and credibility that members of more marginalized and disadvantaged social groups do not have in equal measure (Alcoff 1991–1992). As such, the role requires a balance of recognizing one’s power and the part one must play if change is ever to occur, on the one hand, and paying attention to the marginalized and often voiceless individuals who do not have their fair share of social and political power. At the same time, we want to avoid a totalizing analysis of social and political powerlessness. Inequality, even unjust inequality, does not completely erase or take away agency even though it compromises and constrains it. If it did, then grass roots social movements would never gain traction, members of privileged social groups would need to do all the work, and unjust social structures would

150



change only through their efforts. Similarly, we want to avoid an analysis that suggests that members of privileged social groups can only act badly and in moral ignorance. That would be a demoralizing state of affairs that fails to adequately acknowledge real possibilities for moral agency. Members of privileged and of disadvantaged social groups both have parts to play in addressing social injustice. Invoking social group membership as a role is a good way to avoid a totalizing analysis of powerlessness because it enables us to recognize that different people have different parts to play. We have already talked about the pejorative idea of ‘knowing one’s place’ in a social and political hierarchy. I now want to turn to something more positive: ‘the part we play’. In my work on collective responsibility, I have argued that some of our contributions must be understood with reference to the collective context in which they occur (Isaacs 2011, chapter 2). My argument is that for some of our actions, it is only with reference to the collective outcome to which they contribute that we can gain an understanding of their moral dimensions. Following Donald Davidson’s and Joel Feinberg’s theories of action in which actions can have various ‘levels of description’ (Davidson 1980; Feinberg 1965), we might say that some ways of describing a person’s action contain moral content that other ways of describing it lack. For example, if I punch someone, we might also describe my action in a less morally loaded way as ‘I moved my arm.’ But though it is true that my moving my arm (in this case) is identical to my punching the person, the former does not suggest the moral contours of the act that the latter does. This same analysis can apply to descriptions that cross the individual– collective divide. Consider, for example, if someone murders another person. If we accept Bernard Williams’s (Williams 1985) distinction between thick and thin moral concepts (and we should, because it’s useful), we can understand ‘murder’ as a thick moral concept, meaning it has the idea of moral wrongness built into it. To identify a killing as a murder is to identify it as wrong. Killing can be accidental and not morally wrong, but murder cannot. When we clarify that a particular act of killing is indeed an act of murder, the description we use to identify the action provides information of a morally relevant nature. Now we might take this analysis one step further if we consider that this murder takes place within the context of and as part of a collective action: genocide. Murder and genocide are both wrong, but in moral terms they are qualitatively different. Describing a particular act of murder as ‘a contribution to genocide’ adds to it a moral element that cannot be captured independently of the collective. Genocide is a collective act, and can only be understood as such (Isaacs 2011). Contributions to genocide, when identified specifically as such contributions, are morally distinct from murders that do not take place in the context of genocide, even if instances of both types of actions are wrong.

   

151

This (admittedly brief and rough) action theoretic analysis can help us get a sense of the part different people might play in shifting ignorance to knowledge of oppressive social structures. These structures are difficult to change because, as we have seen, epistemologically the extent or even the very existence of them is not recognized by all who participate in them. Not all who have power recognize it as such. Many people, whether of privilege or disadvantage, consider their way of life and the opportunities that present themselves as the normal course of events. Oppressive social contexts normalize injustice. People who do not have to confront the barriers of this injustice in their own day-to-day lives, that is, those members of privileged social groups who benefit, are especially unlikely to recognize it. When it is normalized, people do not consider act descriptions of their individual actions that identify them as ‘contributions to unjust social arrangements’ or ‘contributions to oppression’ or even more simply as ‘acts of harm’. When actions are so widespread and so many people are engaging in them, it is easy to overlook the cumulative impact because each individual contribution appears innocuous on its own. Many of us may, as children, have been taught to consider what the world would be like if ‘everyone did that’ when it comes to acts such as lying, cheating, or littering. But we have not been taught to consider the impact of the things that everyone or almost everyone (or lots of people) actually do as part of their everyday life, such as eat animal products produced on factory farms, engage in hiring practices that favour White men, engage in acts of cultural appropriation when dressing for costume parties, or support sports teams whose names trade on harmful and even painful stereotypes, erase past colonialist histories, and persist despite objections from Indigenous groups, whose members are already marginalized because of their Indigenous identities. These sorts of actions have for a long time taken place in the absence of widespread recognition of their wrongness and associated harms. When we consider possible roles that social group membership might generate in an oppressive social context, we can see that epistemological obstacles that create moral risk for privileged moral agents can also create moral opportunities. The moral risks in question are the unrecognized contribution individual actions make to harmful collective outcomes. It’s easy not to recognize these risks and yet they are real—they put people at risk of the wrongdoing of contributing to unjust outcomes, like oppression, without recognizing that they are so contributing. If the people ‘in the know’ are mostly members of marginalized social groups, then the social structure needs to shift in such a way that their voices are not silenced and the material realities of their experiences have credibility. Members of what Calhoun refers to as ‘knowledge-acquiring subgroups’ (Calhoun 1989, 396) face a paradoxical circumstance in oppressive social contexts because quite frequently they are regarded with scepticism and not authority.

152



We must not overlook power differentials. As noted, we might place an undue burden on these individuals when their epistemological advantage positions them in such a way that they are required to do the work of educating. People are often not open to being educated, especially by those who have lived on the margins. It is not comfortable to be ‘called out’ and indeed ‘call-out culture’ or ‘cancel-culture’ are among the newer pejoratives for ‘political correctness’. People who are used to having a voice and who as a matter of course are regarded as credible and worthy without having to fight to be heard do not easily let go of that privilege to make way for others who are pointing out unjust inequality. The beneficiaries of this inequality may not welcome the message that they need to share the space. But if they have the most social and political power in the present oppressive context, they need to play a part in any successful effort to change. Their role as members of privileged social groups creates a moral obligation on their part to do so. But members of more disadvantaged social groups also have obligations, even if these place burdens. Given that they have insights that others lack, then it would seem as if they have obligations to make those insights visible to others. That does not mean taking every opportunity to address every single injustice—that would be too demanding. Returning to the example of the obligation to resist, we might recognize that it clearly places a burden. But is that burden ‘undue’? The thing to be aware of here is that we cannot have it all ways. Morality sometimes does create demands—that is part of what it means for morality to have a normative dimension. It makes demands on us. It sometimes feels as if those demands are unfair, as when bad moral luck places us in circumstances that some others do not have to face. We can find ourselves in circumstances where we are put to the moral test in ways that other moral agents are not (Nagel 1979, 25–38; Williams 1981, 20–39). As noted, in circumstances of oppression, members of different social groups are positioned differently and therefore have different roles within the project of addressing the injustice. It is not that only those in roles of privilege or only those in roles of disadvantage have obligations. Recognizing that members of disadvantaged social groups do not alone bear a moral burden in these contexts can help alleviate the moral burden on those who are already burdened by structural disadvantage. Recall that oppression does not befall individuals as individuals. It requires a collective solution, a goal towards which individuals, through collective effort, may strive. Adding a role analysis to this picture allows that differently positioned individuals may have different obligations towards the collective goal because different possibilities and moral opportunities are available to them. It would be as unhelpful and even offensive to say that people in marginalized social groups have no agency as to say that they bear the full moral burden of addressing social injustice. We want to allow that this is not a totalizing scenario that robs people completely of their agency even while acknowledging that it puts many obstacles in their path. People who hold more privileged positions must use their positions

   

153

in the service of the collective goals against oppression. Understanding how they may do that will require collaborative effort with those who are differently placed.

6. Conclusion The question at the heart of this ethical enquiry has been whether and how understanding social group membership in terms of roles might help us find a way forward to more just social arrangements in oppressive social contexts. At first this possibility might seem fraught, more likely to create unjustified further burdens on people who already experience the disadvantages of structural injustice on the basis of race, ethnicity, sexuality, gender, class, ability, poverty, and the intersections of these and other axes of oppression. But as we recognize that people’s different social and political locations constitute different epistemological standpoints, then it would be unrealistic to think that change can happen without some action on the part of those who have the most insight about the unjust social arrangements. It would be equally unrealistic to think that change can happen without action on the part of the people who benefit, experiencing privilege within the social arrangements. Throughout the chapter, I have drawn from the nature of roles within organizational structures. In those contexts, people occupying roles with more authority and power have moral obligations and responsibilities that those in positions of less authority and power do not have. Being in a position of power carries with it significant moral risk. This moral risk is quite severe in a context of widespread wrongful social practice, where one’s own moral ignorance means one routinely engages in behaviour that is morally wrong. This analysis enables us to regard being in a social group that is privileged and being in a social group that is disadvantaged by structural injustice as roles. These different roles obligate people differently with respect to transforming the social arrangements. When is this relevant? It is relevant when members of privileged groups start to ask how they might be good allies, what they can do to contribute positively to change, and how they might educate themselves about the social and political realities that others face. In addition, it is relevant when members of disadvantaged social groups make considered judgements about when they should exercise their obligation to resist and be vocal about the moral insights that they have and that might help transform from an oppressive to a just social and political context.⁷ ⁷ Many thanks to Sean Cordell and Alex Barber for editing this volume. I am grateful that they invited me to be a part of it and to the Role Ethics Network that led to it. I also very much appreciate their helpful feedback and edits on earlier versions of this chapter. Thanks also to participants in the Role Ethics Network workshops I attended in Edinburgh (2017) and London (2017) for their helpful feedback on earlier versions of this chapter. Special thanks to Diane Jeske for invaluable feedback on multiple versions.

154



References Alcoff, Linda. 1991–1992. ‘The Problem of Speaking for Others’. Cultural Critique 20: 5–32. Bousquet, Tim. 2020. ‘Shopping while Black: Santina Rao’s Experience at Wal-Mart’. Halifax Examiner, January 20, 2020. . Budig, Michelle J. 2002. ‘Male Advantage and the Gender Composition of Jobs: Who Rides the Glass Escalator?’ Social Problems 49 (2): 258–277. Calhoun, Cheshire. 1989. ‘Responsibility and Reproach’. Ethics 99 (2): 389–406. Crenshaw, Kimberlé. 1991. ‘Mapping the Margins: Intersectionality, Identity Politics, and Violence against Women of Color’. Stanford Law Review 43 (6): 1241–1299. Davidson, Donald. 1980. ‘Actions, Reasons, Causes’. In Essays on Actions and Events. New York: Clarendon Press, pp. 3–20. Davis, C., Joyce, S., and Murch, K. 1991. The Chilly Climate for Women in Colleges and Universities [Videorecording]/Executive Producers: Western’s Caucus on Women’s Issues and The President’s Standing Committee for Employment Equity, U.W.O.; produced by Kem Murch Productions; written and directed by Kem Murch. London, Ont: Distributed by University of Western Ontario, Dept. of Equity Services. Feinberg, Joel. 1965. ‘Action and Responsibility’. In Philosophy in America. Max Black (ed.). Ithaca, NY: Cornell University Press, pp. 134–160. Fleras, Augie. 2016. ‘Theorizing Micro-aggressions as Racism 3.0: Shifting the Discourse’. Canadian Ethnic Studies 48 (2): 1–19. Freeman, Lauren, and Schroer, Jeanine Weekes. 2020. Microaggressions and Philosophy. New York: Routledge. Frye, Marilyn. 1983. ‘Oppression’. In The Politics of Reality. New York: Crossing Press, pp. 1–16. Hardimon, Michael. 1994. ‘Role Obligations’. Journal of Philosophy 91 (7): 333–363. Harding, Sandra. 1992. ‘Rethinking Standpoint Epistemology: What is “Strong Objectivity”?’ The Centennial Review 36 (3): 437–470. Hay, Carol. 2011. ‘The Obligation to Resist Oppression’. Journal of Social Philosophy 42 (1): 21–45. Hull, Akasha (Gloria T.), Scott, Patricia Bell, and Smith, Barbara (eds.). 1982. All the Women Are White, All the Blacks Are Men: But Some of Us Are Brave. Old Westbury New York: The Feminist Press. Isaacs, Tracy. 2011. Moral Responsibility in Collective Contexts. New York: Oxford University Press. Jeske, Diane. 2001. ‘Special Relationships and the Problem of Political Obligations’. Social Theory and Practice 27 (1): 19–40. Jeske, Diane. This volume. ‘Are Obligations of Friendship Role Obligations?’

   

155

Kisolo-Ssonko, Joseph. 2019. ‘Race and the Responsibility to Abide by the Norms of Unchosen and Unjust Social Roles’. The Monist 102: 172–186. Loewen, Claire. 2020. ‘With Rolling Protest, Black Montrealers Denounce the Challenge of “Driving while Black”’. CBC News, July 6, 2020. . Lugones, Maria, and Spelman, Elizabeth V. 1983. ‘Have We Got a Theory for You! Feminist Theory, Cultural Imperialism, and the Demand for “the Woman’s Voice” ’. Women’s Studies International Forum 1 (6): 573–581. Nagel, Thomas. 1979. ‘Moral Luck’. In Mortal Questions. Cambridge: Cambridge University Press, pp. 25–38. Rini, Regina. 2021. The Ethics of Microaggression. New York: Routledge. Williams, Bernard. 1981. ‘Moral Luck’. In Moral Luck. Cambridge: Cambridge University Press, pp. 20–39. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press. Williams, Christine. 1992. ‘The Glass Escalator: Hidden Advantages for Men in the “Female Professions” ’. Social Problems 39 (3): 253–267. Wylie, Alison. 2003. ‘Why Standpoint Matters’. In Science and Other Cultures. Robert Figuera and Sandra Harding (eds.). New York: Routledge, pp. 26–48. Zheng, Robin. 2018. ‘What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice’. Ethical Theory and Moral Practice 1: 869–885.

8 Explaining Role-Based Reasons Reid Blackman

1. Introduction It is difficult to overstate the ubiquity and pervasiveness of social roles. But I will try. If you are reading this, you likely spend much of your time talking to professors, students, administrative assistants, librarians, deans, members of IT support teams, and so on. Some of these people are also your friends. On your way home, you may stop at a store where you’ll ask the clerk for help and then chat with a fellow citizen in the checkout line about the latest political scandal while you wait for the cashier to ring up your purchases. You might pause longer than usual at the stop sign when you see a police officer stationed nearby, and when you get home, you might be greeted by your spouse and children, and later you might call your parents. There are surely exceptions, but for the most part it is difficult to interact with someone in which the relation between you and that person cannot accurately (even if not fully) be described as one that obtains between occupants of social roles. If things had been otherwise, or rather, had you been otherwise, you would have reasonably done other things, and that is because professors, parents, and spouses have normative reasons that non-professors, non-parents, and non-spouses lack. The obvious explanation for this is that the former occupy social roles that the latter do not; occupying a role can make a difference to the stock of an agent’s reasons. But how does it do this? We begin to get a grip on the options for answering this question by bringing together two areas of philosophical enquiry that are too infrequently brought together: normative reasons and social ontology. In theories of normative reasons for action (meaning, here, reasons that stand in a justificatory relation to action) there is a well-known distinction between internalism and externalism. If one is an internalist about normative reasons then one will take all of an agent’s normative reasons qua role occupant to depend, at least in part, on the motivational constitution of the agent.¹ For instance, an ¹ This conception of internalism is meant to echo the characterization by Finlay and Schroeder (2017). Internalists are standardly found in broadly Humean and broadly Kantian camps. The former include Williams (1981; 1995), Smith (1994), Schroeder (2007), Joyce (2001), Brandt (1979), Railton (1986), Manne (2014) (though Manne is better characterized as a Strawsonian than a Humean), and the latter include Darwall (1983), Korsgaard (1986), Velleman (1996), and Markovits (2011).

Reid Blackman, Explaining Role-Based Reasons In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Reid Blackman 2023. DOI: 10.1093/oso/9780192843562.003.0008

 - 

157

agent who is a parent will have a normative reason to do what (good) parents do on the condition that doing so promotes the satisfaction of at least one of that agent’s desires. If one is an externalist then one endorses no such requirement; an agent who is a parent may have a normative reason to do what (good) parents do independently of whether doing so promotes the satisfaction of that agent’s desires.² When it comes to the ontology of social roles one may endorse a normative or non-normative account. According to the normative account a role just is a set of normative relations, such as obligations, rights, and normative reasons.³ To be a doctor, for instance, just is to have an obligation to care for the welfare of one’s patients, a normative reason to run a diagnostic test, and so on. If this account is correct then agents have normative reasons just because they are role occupants. Being in the role makes it the case that one instantiates the normative relations that constitute the role. According to a non-normative account of roles there will have to be more to the explanation for why agents qua role occupants have their normative reasons than simply that they are role occupants and roles are constituted by normative reasons. That is because, whatever roles are on the nonnormative account, being in a role would not sufficiently explain why the agent has her respective reasons. From these distinctions in metaethics and social ontology we can see four paths forward in answering the question, ‘How does occupying roles affect the stock of an agent’s reasons?’ (Table 8.1). These paths are not equally plausible; the position marked by ‘A’ is particularly unpromising. That is because it is highly unlikely that the entry and exit conditions for roles always depend on the motivational constitutions of agents: citizens are born, some parenthoods are the result of accident, and lawyers are disbarred. But if one can meet the entry conditions for a role independently of one’s motivational constitution and roles are constituted by normative reasons, then one could have reasons independently of one’s motivational constitution, a view that is incompatible with internalism. Internalists will thus need to adopt B. Table 8.1. Four positions towards the externalism/internalism debate in metaethics and the normative/non-normative debate in social ontology.

Internalism Externalism

Normative Account of Roles

Non-normative Account of Roles

A C

B D

² Externalists include Enoch (2011), Dancy (2000), and Parfit (2011). ³ In social ontology, see Searle (1995, 2010), Thomasson (2003), Khalidi (2015), Cole (2015), and Ritchie (2015). In normative ethics, see Hardimon (1994), Swanton (2007), Sciaraffa (2011), Cordell (2011), and Dare (2016).

158



In light of ruling out A there are two debates to be had, one between internalists and externalists and one between two kinds of externalists. As for the former, in order for an argument to be non-question begging against the internalist, an externalist will have to assume a non-normative account of roles; that debate is between B and D. As for the latter, and putting internalism to the side, we have a debate between C and D. I have discussed the debate between B and D at length elsewhere (Blackman 2017) and will, for the sake of this chapter, assume that B is implausible. Here I advance the debate between C and D. More specifically, I spell out the details of possible positions in D and then articulate and defend one such position. So far as I know, this position has not been articulated, let alone defended, by anyone. I will not, however, conclude that we ought to endorse D. That is because, as we have seen, there is a debate between C and D, and we cannot settle that debate without thinking in great detail about the costs and benefits of normative and nonnormative accounts of roles. Getting a clearer view of D along with its benefits, however, will put us in a good position to have that debate. In the next section I provide a sketch of a non-normative account of roles and explain an aspect of it that is particularly attractive. I then articulate the logical landscape of views on role-based reasons—that is, the set of reasons agents have, at least in part, by virtue of being occupants of roles—that are both externalist and assume a non-normative account of roles. Subsequent sections develop and defend one of those views (i.e. views that follow pathway D) as the best among them, and call it Role Pluralism. Before proceeding, a quick note is in order. One reason why something like the view I am about to defend has not been taken seriously is the result of a hastily made objection that quickly arises when one talks about the normativity of roles: Nazis, mob bosses, and assassins. Surely, the thought goes, social roles by themselves cannot serve as any kind of ground for normative reasons, since then we’d have to allow that Nazis have normative reasons to do nasty Nazi-things, and that is something we cannot allow. Please rest assured that I will discuss this objection.

2. A Non-Normative Account of Roles and Two Approaches to Role-Based Reasons Any option in D depends on a non-normative account of roles. That means, whatever else we say about roles, we cannot say that they just are a set of normative relations, such as sets of obligations, rights, and normative reasons. One essential data point for constructing a theory of roles is that occupants of roles can be good or bad occupants of their respective roles. There are not only professors, parents, and citizens, but also good professors, bad parents, and mediocre citizens. An account of roles must be able to articulate what makes it

 - 

159

the case that occupants are evaluated in the way they are, that is, what explains why these are the good ones and those are the bad ones. We are helped here by thinking of two parallel distinctions, one found in ethics and the other in philosophy of biology. The former is between good of a kind and instrumental value, and the latter is between the function of a biological trait versus a lucky accident that results from that trait. When something has instrumental value or is a lucky accident, it is good for bringing about some end. Hammers are good at driving nails into wood, though a steel-toed boot can do just as well; they can be of equal instrumental value. And it is a lucky accident that noses can hold up glasses but the function of the nose is to aid in the respiratory system. When something is good of a kind, on the other hand, it is good by virtue of its realizing its function or end. The hammer is good of its respective kind because it can drive nails into wood; that is the function of a hammer. But the steel-toed boot is not good of its respective kind because it can drive nails into wood; that is not the function of a boot. A thing is good of its kind when it realizes the function or end that is constitutive of its kind and the function or end is explanatorily prior to the good-making features of the object. It is because this thing has the function of driving nails into wood that its having a sturdy handle, and not its being blue, is a good-making feature of the hammer. The function of a heart is to pump blood, which explains why its contracting in such-and-such a way, but not its making a thumping sound, is a good-making feature of a heart. That it makes a thumping sound is a lucky accident, however, since it aids doctors in diagnosing illness and defect. A fruitful approach to a non-normative account of roles is to take the structure of the property ‘good of a kind’ and apply it to roles. On such a view, roles have functions or ends that are explanatorily prior to the good-making features of its occupants. For example, the function or end of a parent qua parent is the welfare of her children, which explains why ensuring their education is a good-making feature of a parent. Notice that, just as we do not need to appeal to obligations, rights, and normative reasons to explain what a good hammer or heart is, nor do we need to appeal to them to explain what a good occupant of a role is. Put differently, while this account of roles gives us the resources to ‘grade’ occupants of various roles, nothing in the ontological account of roles indicates that they have an obligation or normative reasons to do well on that test. A quick note on terminology. Not all artifacts have functions, nor do all biological traits. When I talk about artifacts or biological traits as having occupants that are good of their respective kinds, I mean the subset of artifacts and biological traits/organisms that can be evaluated in this way. Similarly, not all social roles have functions/ends, but when they do, occupants can be good of their respective kind. Hereafter I use ‘role’ to designate the subset of social roles that do have functions/ends and (hence) occupants that are good of their respective kind.

160



One attractive feature of this account is that it provides the conceptual underpinnings that make possible the debate between internalists and externalists. As we saw above, internalists cannot accept a normative account of roles and to the extent that that position is plausible, we will need an account of roles that does not rule it out. But the more attractive feature of this non-normative account is that it is commensurate with a general account of the structure of the property ‘good of a kind’. According to the general account, the property ‘good of a kind’ just is the property of realizing the function or end of one’s respective kind where the function or end is explanatorily prior to the good/bad-making features of the occupants of the kind. That account of the property allows us to say that good hammers, good hearts, and good doctors are all good occupants of their respective kinds and that the property ‘good of a kind’ is uniform across domains. Similarly, it allows us to say that good hammers, good hearts, and good doctors are each good of their respective kind without equivocating on ‘good of a kind’. It is not that the way in which hammers are good of their respective kind is different from the way in which hearts and doctors are good of their respective kinds.⁴ An externalist who endorses this non-normative account of roles—an occupant of box D—needs to explain how occupying roles affects the stock of an agent’s reasons. What we have so far is just the claim that occupying a role is contingently sufficient to provide role-based reasons. In the next section I offer a map of the logical terrain. As we will see, there is a variety of positions available in D, one of which I am particularly keen to defend.

3. Role Pluralism and Its Competitors I begin with a distinction between indirect and direct accounts of role-based reasons (given a non-normative account of roles). The indirect account sees the ends of roles as reason-providing on the condition that some normative fact obtains that (a) is independent of the ends (or functions, etc.) of particular roles and (b) explains why the ends of roles provide normative reasons. For example, if one thinks there is a moral principle in the heavens that says, ‘Thou shalt be a good occupant of the roles one occupies’, and that explains why an agent should realize the ends of her respective roles, then one endorses an indirect approach to the existence of role-based reasons. Similarly, if one thinks one must have a

⁴ None of this implies that we must endorse a uniform account of function acquisition for each of these kinds. It is one thing to attribute a function or end to a kind and another to have an account of how it acquires, or what explains, its having that function or end. It may be that the intentions of designers are necessary in the case of artifacts like hammers but not so in the case of biological traits, for instance. Or it may be that we have a uniform account of function acquisition across artifacts (in those cases in which they have functions), biological traits, and roles. The account of good of kind articulated here is neutral on this point.

 - 

161

normative reason, which is not itself explained by occupying any role, to be a (good) occupant of the role, one endorses an indirect approach to the existence of role-based reasons. And lastly, if one thinks that the role must be good simpliciter in order for its ends to provide reasons, then one endorses an indirect approach to role-based reasons.⁵ The direct approach has at least two ways of being developed; one may take a hierarchical or non-hierarchical approach. The hierarchical approach asserts that some roles are normatively prior to other roles in the sense that the normativity of some roles explains the normativity of others. Aristotle, as I read him, endorsed the hierarchical account. On some approximation of Aristotle’s view, there is no such thing as the Form of the Good.⁶ Instead, there are things that are good of a kind. Good flautists, good statesmen, and good human beings. Each of these kinds has a function or end, and one is a good occupant of a given kind when one realizes the function of that kind. The function of a human being is rationality, for example, and so a good human being is one who exercises the rational function well.⁷ Aristotle’s picture has more structure than this though. He also thinks that human being is a master role in that the normativity of human being explains why subordinate roles are normative. Richard Kraut tells us that Aristotle thought that, ‘when one finds a nested series of functions, they ultimately serve one highest function. The various functions of craftsmen must ultimately serve some higher function—and what else could that be but our functioning as human beings?’ (Kraut 2002, 82).⁸ And on Rachel Barney’s account of Aristotle, ‘Simon’s functioning as a shoemaker can have normative standing only if it realizes or instantiates Simon’s own function as a human being’ (2008, 310). The non-hierarchical approach does not hold that there is any master role that explains the normativity of other roles. When an agent qua role occupant has a normative reason, that is not because of the normativity of any other role. Further, because this is a variety of the direct approach, neither does the agent qua role occupant have a normative reason because of some normative fact that is roleexternal. Occupying the role is, as it were, normative all by itself. However, because this approach assumes a non-normative account of roles, the explanation for why ⁵ It would not be enough to count as an indirect account that the role must pass the test of not being immoral. It may be that an immoral role cannot provide reasons—an issue I will take up—but even if that is the case it does not follow that the lack of immorality explains why the end provides a reason. It would be odd to say, for instance, that what explains why the end of a dance instructor provides the instructor with normative reasons is that being a dance instructor is not immoral. ⁶ See Nicomachean Ethics I.6. ⁷ Aristotle does not think of human being as a role, of course. But he does think that we are human beings, that there is a constitutive function or end of human beings, and that occupants of the kind have reasons to realize the function of a human being. That is all I need to include the version of Aristotle I present as someone who endorses a hierarchical and direct approach to role-based reasons. See Nicomachean Ethics I.7. ⁸ I take the ‘must’ here to be normative, or at least to spell out a necessary condition for some nested function to be normative, viz. it must serve (causally or constitutively) one’s function as a human being.

162



Reasons Externalism & Non-normative Account of Roles (Box D)

Direct Account of Role-Based Reasons

Non-hierarchical Account

Indirect Account of Role-Based Reasons

Hierarchical Account

Figure 8.1. Reason-externalist, non-normative accounts of roles: the options

occupying a role is normative all by itself does not appeal to the ontological claim that roles just are sets of obligations, rights, and normative reasons. We have already covered a fair amount of logical terrain. Figure 8.1 summarizes where we have reached thus far. It is important to stress that the distinction between hierarchical and nonhierarchical accounts turns on what explains why roles are normative. It is sufficient to be a hierarchical account that the account affirms that the normativity of some roles explains the normativity of others. And it is important not to confuse this issue with the issue of the weights of role-based reasons, that is, in cases in which different roles provide normative reasons for incompatible actions, which reasons are the weightiest. More specifically, it is important not to confuse a hierarchical account with what I will call the embedding thesis about role-based reasons. The embedding thesis says that if role A is embedded in role B then the normative reasons of role B are necessarily weightier than the normative reasons of role A. Aristotle endorsed the embedding thesis. On his view, being a human being explains why being a shoemaker is normative, but in addition the reasons attached to the former are necessarily weightier than the reasons attached to the latter; the end of human being is ‘to be preferred’ to the end of shoemaker. And, depending on how ‘embeddedness’ is interpreted, a non-hierarchical account of role-based reasons may endorse the embedding thesis as well. One could think, for example, that occupying the role of human being does not explain why occupying the role of shoemaker is normative—they are both normative all by themselves— but the role shoemaker is embedded in the role human being and, one might maintain, the reasons of the latter always outweigh the reasons of the former. That said, neither the hierarchical nor non-hierarchical account must endorse the embedding thesis. One may maintain that, even though shoemaker is embedded in human being, that says nothing about the relative weights of the respective reasons they provide. I am not sure how we should understand embeddedness on Aristotle’s account but here I stipulate that for role A to be embedded in role B is for it to be the case that if one occupies role A then one occupies role B. For example, the role

 - 

163

professor of the philosophy department is embedded in the role professor of the university; the former entails the latter but not the other way around. Each of these roles has its distinct ends, specifically, those relating to the welfare of one’s department and the students of one’s department and those relating to one’s non-departmental colleagues and non-departmental students, respectively. In cases in which (a) A is embedded in B and (b) B is not embedded in A, the embedding thesis says that the normative reasons of B are weightier than the reasons of A. On this interpretation of embeddedness I think the embedding thesis is false. It strikes me as obvious that one’s reasons as a professor of the philosophy department can outweigh the reasons one has by virtue of being a professor of the university. One can have stronger reason to meet with one’s advisee than to attend the university-wide faculty meeting, for example. And with that pronouncement we have now come, at last, to the account in Box D I will articulate and defend for the remainder of the chapter. I call this view Role Pluralism, and it has four notable traits. It: i. ii. iii. iv.

Is externalist about normative reasons Assumes a non-normative account of roles Affirms a direct non-hierarchical account of role-based reasons Denies the embedding thesis.

According to Role Pluralism, occupying a role gives one normative reasons to realize the ends of that role and it is not the case that the normativity of any nonrole or even any other role explains why occupying a role provides reasons. Further, it rejects the claim that if role A is embedded in role B then the normative reasons of role B are weightier than those in role A.⁹ I call it ‘Role Pluralism’ to indicate that our role-based normative reasons are grounded in each of the various roles we occupy, as opposed to the hierarchical approach, according to which they are ultimately grounded in some ‘master’ role. Role Pluralism, so far as I know, has never received a hearing in the court of philosophical judgment. In my view this is unfortunate, because I think it is not only plausible but a very attractive account of role-based reasons. But so far we have only the sketch of a position. Role Pluralism asserts that occupying a role explains how occupying a role affects the stock of an agent’s reasons, but we have not yet seen what that explanation is. In the next section I am going to offer that explanation. Subsequent sections explain why Role Pluralism is to be preferred to hierarchical and indirect approaches to role-based reasons.

⁹ One may be justly reminded here of David Ross’s (1930) non-hierarchical account of the prima facie duties; in different situations the demands (or at least reasons) of the roles one occupies will vary.

164



4. Role Pluralism as an Analysis of Role-Based Reasons I begin with an articulation of what we are trying to do when we explain normative reasons of any sort, and demonstrate in a very broad way that in trying to explain how occupying a role can explain an agent’s reasons we are doing something of a piece with trying to explain normative reasons in general. I know of no account that has done this with more perspicacity than Mark Schroeder’s Humean account of normative reasons. My goal is to identify in his account a structure for explaining normative reasons and then to exploit that structure for the purposes of developing Role Pluralism. On Schroeder’s view, an agent’s normative reasons are explained by her desires. That is, in essence, what a Humean theory of practical reasons claims. Schroeder’s innovation is to show that the property of being a normative reason has a certain structure, where that structure is realized by a set of properties and relations among those properties. His proposal is as follows: (Schroeder 2007, 59) Desire: For R to be a reason for X to do A is for there to be some p such that X has a desire whose object is p, and the truth of R is part of what explains why X’s doing A promotes p.

In short, to be a normative reason (for action) on his account is to be some fact that explains why doing A would serve X’s desire. For instance, the fact that there is coffee in the next room is Alice’s reason for going there because the fact that there is coffee in the next room explains why her going to the next room serves her desire for coffee. The fact that there is coffee in the next room has the property being a reason to go to the next room (or what amounts to the same, has the property favoring or counting in favor of going to the next room), and what explains why that fact has that property is that being a reason just is to be the sort of thing that explains why an agent’s doing something promotes the object of at least one of the agent’s desires. I want to put to the side the specifically Humean content of Schroeder’s view— that is, the content that involves direct reference to desire—and look at the structure we have once we do that. We have the following. General Structure (GS): For R to be a reason for X to do A is for there to be some p such that X has the end of p, and the truth of R is part of what explains why X’s doing A promotes p.

On Schroeder’s view the end of p is taken to amount to having a desire that p. And this is just what we should expect, since Humeanism is often characterized by the claim that the ends of action are set by (intrinsic) desires. But we can be attracted to GS without being committed to internalism; there

 - 

165

are other ways of filling out GS such that we do not end up with Desire. As the reader may have anticipated by now, we may fill out GS with the notion of the end of a role. Role: For R to be a role-based reason for X to do A is for there to be some p such that X occupies a role with the end of p, and the truth of R is part of what explains why X’s doing A promotes (or realizes) p.

In short, to be a role-based normative reason, on this account, is to be some fact that explains why doing A would realize or promote X’s end qua role occupant. For instance, for ‘that his children need to be picked up at school’ to be a reason for John to pick them up is for there to be an end of ‘caring for one’s children’ such that John occupies a role with that end—father, for example—and the truth of ‘that his children need to be picked up at school’ is part of what explains why his picking his children up at school promotes his end qua father. The fact that his children need to be picked up at school has the property being a reason to pick them up (or what amounts to the same, has the property favoring or counting in favor of picking them up), and what explains why that fact has that property is that being a role-based reason just is to be the sort of thing that explains why an agent’s doing something promotes an end the agent has qua role occupant. In a way, Role Pluralism is fairly simple. Its core claim is that, by virtue of occupying a role, one has normative reasons to do that which is conducive to, or constitutive of, meeting the ends of that role. And that is because, once one occupies a role—that is, once one has met the entry conditions for the role, whatever they are—(a) one has an end and (b) what it is to have a normative reason to act just is for some fact to explain how so acting is conducive to, or constitutive of, realizing that end.

5. Role Pluralism or Its Competitors? I outlined above the various paths one might take in developing an account of role-based reasons that is at once externalist and endorses a non-normative account of roles. The first fork in the road is between direct and indirect accounts. The second fork is between direct accounts that are hierarchical and direct accounts that are non-hierarchical. Let us work our way backwards by starting with the second fork: comparing non-hierarchical with hierarchical direct accounts. I will argue—unsurprisingly, given what Role Pluralism maintains—that the former is a stronger position. I will then go on to compare non-hierarchical direct accounts with indirect accounts arguing, once again, that the former is to be preferred.

166



Hierarchical vs. Non-Hierarchical Direct Accounts of Role-Based Reasons Recall that, according to a hierarchical direct account of role-based reasons, there is some proper subset of roles the normativity of which explains the normativity of all other roles. I referred to the former as ‘master roles,’ and took some version of Aristotle’s view to be an instance of this approach. The case against the approach and in favor of its non-hierarchical sibling (assuming for now that these are the only two positions in play) will have two steps. The first step will involve identifying those elements of the hierarchical view that the nonhierarchical view must reject, distinguishing them from those elements of the hierarchical view that the non-hierarchical view can happily agree with. Step two will be to argue that the elements unique to the hierarchical view are less plausible than the elements it has in common with the non-hierarchical view, and that these more controversial elements are in any case redundant. This does not show the hierarchical view is wrong. The goal here is limited to establishing a prima facie case against the hierarchical view relative to the non-hierarchical direct account. What both positions agree on springs from their both being direct accounts. They deny that the normativity of roles is explained by non-role-based normative facts. At its core, Role Pluralism (our representative non-hierarchical direct account) is the view that there is a non-empty set of things a person can be such that these things have a function or end that is normative reason-giving to those people who are those things. The hierarchical view affirms this too but adds several further elements that Role Pluralism must reject. Since these are the elements that I will be saying are implausible, let me initially cast them in a favorable or at least neutral light. First, we get the assumption that there is a master role (or a suite of master roles, perhaps), the function or end of which sets our moral standard. Second, on a view like Aristotle’s or a suitable development thereof, we get a particular view on what kind of thing that master role is—a human being or a rational agent, say. Third, insofar as the master role is occupied by all (all rational agents, all human beings, as it may be), we get a ready explanation for why moral standards are universal: they reflect the fact that moral standards are grounded in a universally instantiated role. Fourth, and appealing to those who think that there are no normative reasons for immoral deeds, it allows that immoral roles do not provide normative reasons. It allows this because all subordinate roles will inherit their normativity from the master role while no role that is fundamentally incompatible with the master role will do so. Fifth, those hierarchical accounts that incorporate the embedding thesis (which, as we saw, is an optional extra rather than a core requirement of the view) will further claim that the normative reasons of that universal identity are not only distinctively moral reasons but also that they are the weightiest.

 - 

167

Turning now to the second step, it should be evident that the hierarchical view envisages a complex edifice on top of the otherwise simple idea that roles are a direct source of normativity. It is a tempting thought—it would explain so much—but being tempting is not the same as being plausible. The choice between the hierarchical and non-hierarchical versions of the direct account is thus a choice between complex ambition and simple sufficiency. And as with other grand designs in philosophy, there are some familiar dangers facing those attracted to the former option. They would need to defend the notion that there is some specifiable master role and show how the normativity of all roles could be given a perspicuous explanation in terms of the normativity of the master role.¹⁰ Given the enormous variety of normatively potent roles (Barber and Cordell, this volume, Section 2), it is easy to suspect that explanatory vacuity would begin to creep in as we tried to strait-jacket that potency into a single source. It is tempting at this point to recall Popper’s (1957) charges against totalizing explanations in the social sciences, which in their effort to explain everything end up explaining nothing. None of this is to decisively refute the possibility of a hierarchical view. Until such a view has been developed, that is too much to ask. The moral of the reflections above is that we do not in fact need to develop one if we are willing to countenance a pluralistic alternative. Once we accept that one role or proper set of roles—the master roles—are directly normative, insisting that only these roles can have direct normativity, and that all others must derive their normativity from that of this privileged subset, and perhaps of just one in particular, looks philosophically risky and pointless.

Role Pluralism vs. an Indirect Account The indirect approach tells us that role-based normative reasons are explained by some non-role-based normative fact. In fact, this is the most common reply I have received from philosophers in conversation. ‘Surely,’ they say, ‘one must have a non-role-based normative reason to be an occupant of the role, or perhaps to be a good occupant of the role, in order for the standards of that role to provide normative reasons.’ Call this the indirect thought. It is important to highlight that this affirmation of the indirect approach is not an appeal to the need for motivational facts about the agent to obtain, for example that surely the agent must care about being a good occupant if she is to have any role-based reasons. That is to affirm a position in box B, above, which ¹⁰ On the embedding thesis, moreover, since the move from ‘everyone has this reason’ to ‘this reason is stronger than any reason not had by everyone’ is a non-sequitur, we would need an explanation of how universality explains normative strength.

168



I have placed beyond the scope of this chapter. The indirect approach that is a development of a position in box D must proceed in another way. What should lead us to affirm the indirect thought? Role Pluralism, after all, offers an analysis of role-based reasons that, if true, obviates the need for any further non-role-based normative reason to obtain. To be a role-based reason just is, Role Pluralism maintains, to be a fact that explains why performing an action promotes the realization of an agent’s end qua role occupant. Why think there needs to be more to the story than this?¹¹ Defenders of the indirect thought must provide an explanation or an argument for why one first needs to have a normative reason to be a (good) role occupant in order for one to have a rolebased reason. What might such an explanation look like? I am going to answer this question by taking a step back to think about general issues in axiological theory. Doing so will show us a line of thought that, if sound, vindicates the indirect thought. But it will also show us how controversial the underpinnings of the indirect thought are, and thus why we should hesitate to affirm it. A basic distinction in axiological theory is between final value and non-final value. The notion of final value comes to us by way of thinking about what sort of thing ultimately justifies action; if we only do a to get b, and b to get c, and c to get d, and so on, there would be no stopping point, and so no ultimate justification, for our actions. Final value is such a stopping point. When something is of final value, an action is, ceteris paribus, justified by standing in the right sort of relation to that value, such as realizing, bringing about, protecting, etc. How should we understand final value? One way is to think of it as being the property good simpliciter, or good ‘full stop,’ or as Peter Geach famously put it, ‘good in the predicative sense’.¹² If that is the right way to think about final value then that would explain why we need something non-role-based to explain rolebased reasons. Role Pluralism makes no use of the property of being good in the predicative sense. Indeed, it explicitly makes use of what Geach called ‘good in the attributive sense’ or what I have called good of a kind. Now the purveyor of the indirect thought can offer a line of reasoning that vindicates that thought: ‘If we are going to explain normative reasons then we need the property of being finally valuable. Further, being finally valuable is a matter of being good in the predicative sense and so only those roles that are good in the predicative ¹¹ In conversation I have found that the indirect thought is motivated by judgments about particular cases, specifically, cases in which someone seems to be engaged in trivial or just silly endeavors. ‘We may create a professional grass-counting league,’ our objector continues, ‘and there could be professional grass-counters, some of whom are quite good. But even though agents have an end qua grasscounter, surely there is no reason to count grass even if one is a grass counter. That would require that there be something to recommend about being a grass counter in the first place. But of course there is no normative reason to do something—to be something—so trivial.’ ¹² Geach (1956). G. E. Moore famously conceived of good simpliciter as that which has intrinsic value (Moore 1903, 1933). Korsgaard (1983) helpfully corrected him.

 - 

169

sense are reason providing. Put differently, there is some non-role-based normative fact—that the role is good in the predicative sense—that explains why the standards of the role are reason-giving to its occupants. Thus, RIP Role Pluralism.’ That line of reasoning hinges on a highly contentious premise, though: that final value must be analyzed as that which is good in the predicative sense. We certainly do not need to take on board such a bold assumption. Indeed, in the contemporary literature it is far more common to think of final value in terms of non-instrumental reasons. For x to have final value, on that view, just is for x (or properties of x) to give non-instrumental normative reason to agents to, say, promote or realize x.¹³ This newer conception of final value is compatible with Role Pluralism, for according to Role Pluralism an agent has normative reasons to do that which realizes the ends of the roles one occupies where the ends of those roles are the final values for the agent qua role occupant. When an agent qua parent, for example, does that which is non-instrumentally related to realizing the end of a parent—performing an act that is constitutive of realizing that end—that action is not a means to some further end the agent has qua parent. Realizing that end is the end for the agent qua parent; it is that which ultimately justifies the agent’s action qua parent, and that is exactly what we were looking for final value to do, viz. to provide a non-instrumental justification for action. We went down this path because we were searching for a vindication of the indirect thought, the thought that some non-role-based normative fact must explain why occupying a role affects the stock of an agent’s reasons. One way to vindicate the indirect thought is first, to assert a deep connection between final value and normative reasons, and second, to insist that final value must be analyzed as that which is good in the predicative sense. I have accepted the first claim and denied the second, maintaining instead that (i) we may—as many others do—understand final value in terms of non-instrumental normative reasons, and (ii) understand the ends of agents qua role occupants as final values for those agents. Role Pluralism is thus at least as plausible as the indirect account. Whether one prefers the indirect account will be a function of whether one thinks final value must be understood in terms of what is good in the predicative sense (a kind of goodness, I hasten to add, that gives rise to the set of debates that characterize twentieth-century metaethics, e.g. realism vs. anti-realism, cognitivism vs. noncognitivism, naturalism vs. non-naturalism, and so on). Role Pluralism, in not appealing to good in the predicative sense, avoids those highly contentious issues; perhaps Role Pluralism is more plausible for that reason.¹⁴

¹³ See e.g. Scanlon (1998).

¹⁴ See Blackman (2015).

170



6. The Immorality Objection At last we come to the objection we’ve all been waiting for, the objection that is standardly taken to be the definitive refutation of the claim that roles, by themselves, ground reasons. We can call it the immorality objection: ‘If Role Pluralism is true then occupants of all roles, including immoral roles like thief and Nazi camp commandant, have normative reasons to realize their respective ends. But surely agents do not have normative reasons qua thief or qua Nazi. Thus, Role Pluralism is false.’¹⁵ The immorality objection rests upon a more general claim about attributions of normative reasons. I call this claim the ‘morality constraint’. Morality constraint: If X has a normative reason to A, then it is not the case that X’s A-ing is immoral.

With the morality constraint made clear, we can see two ways of fending off the immorality objection. The first is simply to deny the morality constraint; we can endorse the view that, for instance, thieves and Nazis have normative reasons to realize their respective ends, even if we also think those normative reasons are (massively) outweighed by moral considerations; being outweighed is simply not the same as being cancelled. I happen to think the morality constraint is false, and many other philosophers do as well. But some philosophers think it’s true. And while we could argue over the morality constraint, it would be better if there were another route to responding to the immorality objection, and indeed I think there is one. This second and more appropriate reply to the immorality objection, and to the morality constraint that supports it, is that it is irrelevant. In its structure, the morality constraint is similar to other potential constraints on attributions of normative reasons. For instance, one might think that ‘ought/has-areason implies can’ is one such constraint: OHR constraint: ‘If X ought/has-a-normative-reason to A, then it is not the case that X cannot A.’

The OHR constraint, it must be stressed, does not tell for or against any particular account of normative reasons. For instance, on the desire-based account of normative reasons, X may have a reason to A grounded in X’s desire to Y on ¹⁵ Notice that this objection applies with equal force to any view that incorporates a normative account of roles. Since roles are, on that view, sets of obligations, rights, and reasons, and thief and Nazi are roles, then occupants of those roles instantiate the normative relations that constitute those roles. Thus, anyone who endorses a normative account of roles must also deny the immorality objection on pain of otherwise having to deny that Nazis have existed or even could exist.

 - 

171

the condition that they can A. Or on the value-based approach, X can have a reason to A grounded in the value of Y on the condition that X can A. Or on the role-based account, X can have a reason to A grounded in X’s occupying role Y on the condition that they can A. Or, if the desire/value/role-based theorist thinks the OHR constraint is false, then they will simply say that the desire/value/ role sometimes grounds reasons for actions that X cannot perform. Again, the OHR constraint is neutral on the truth or falsity of any particular account of normative reasons. If the OHR constraint is neutral on the truth of these theories, mightn’t the morality constraint also be neutral? After all, desire-based and value-based theorists can disagree between and amongst themselves whether the morality constraint is true. For example, one desire-based theorist may affirm that X has a reason to A on the condition that A-ing is not immoral, while another desirebased theorist can deny that that is a constraint on when desires ground normative reasons. One value-based theorist may think that X has a reason to A on the condition that A-ing is not immoral, while another may deny any such condition. And finally, one Role Pluralist may think that X has a reason to A on the condition the role that grounds that reason is not itself immoral, while another Role Pluralist can deny any such condition. The morality constraint, like the OHR constraint, is just that: a constraint on the conditions under which an agent can have a normative reason. Neither constraint tells for or against any particular account of what grounds normative reasons generally. Insofar as the immorality objection depends for its force on the morality constraint, we don’t need to argue over its truth or falsity for the sake of determining the truth or falsity of Role Pluralism; the immorality objection, underpinned by the morality constraint, is irrelevant.

7. Conclusion Role Pluralism assumes a non-normative account of roles, which in turn allows for a unified account of the concept and property good of a kind. Since the analysis of role-based reasons it affirms makes no mention of the motivations of agents it is an externalist account of normative reasons, and since the property of being a role-based reason is realized when a set of natural properties and relations obtain, it is naturalist. Further, since the analysis of role-based reasons explains how occupying a role with ends gives reasons to role occupants without referring either to something role-external or to ‘master roles’, it is a direct, non-hierarchical account of role-based reasons. Finally, because there seem to be clear counterexamples to the contrary, Role Pluralism denies the embedding thesis. We cannot conclude from the above arguments in defense of Role Pluralism and the arguments against its competitors that it is true. In order to conclude that,

172



not only would we have to conclude that Role Pluralism is the best option in box D, but also that it is the best option as compared to the options in boxes B and C. That said, a view with all of these features is, I think, an attractive account of role-based reasons generally, not just attractive for the kind of account of rolebased reasons that falls into box D.

References Aristotle. 2002. Nicomachean Ethics. Sarah Broadie and Christopher Row (trans. & eds.). Oxford: Oxford University Press. Barber, Alex and Cordell, Sean. This volume. ‘An Overview of Social Roles and Their Ethics’. Barney, Rachel. 2008. ‘Aristotle’s Argument for a Human Function’. Oxford Studies in Ancient Philosophy 34: 293–322. Blackman, Reid. 2015. ‘Meta-Ethical Realism with Good of a Kind’. The European Journal of Philosophy 23 (2): 273–292. Blackman, Reid. 2017. ‘Roles Ground Reasons; so Internalism is False’. Pea Soup . Brandt, Richard. 1979. A Theory of the Good and the Right. New York: Clarendon Press. Cole, Julian. 2015. ‘Social Construction, Mathematics, and the Collective Imposition of Function onto Reality’. Erkenntnis 80: 1101–1124. Cordell, Sean. 2011. ‘Virtuous Persons and Social Roles’. Journal of Social Philosophy 14: 254–273. Dancy, Jonathan. 2000. Practical Reality. New York: Oxford University Press. Dare, Tim. 2016. ‘Robust Role-Obligation: How Do Roles Make A Moral Difference’. Journal of Value Inquiry 50 (4):703–719. Darwall, Stephen. 1983. Impartial Reason. Ithaca, NY: Cornell University Press. Enoch, David. 2011. Taking Morality Seriously: A Defense of Robust Realism. Oxford: Oxford University Press UK. Finlay, Stephen and Schroeder, Mark. ‘Reasons for Action: Internal vs. External’. The Stanford Encyclopedia of Philosophy (Fall 2017 Edition). Edward N. Zalta (ed.), . Geach, Peter. 1956. ‘Good and Evil’. Analysis 17: 33–42. Hardimon, Michael O. 1994. ‘Role Obligations’. Journal of Philosophy 9: 333–363. Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge University Press. Khalidi, Muhammad Ali. 2015. ‘Three Kinds of Social Kinds’. Philosophy and Phenomenological Research 90: 96–112.

 - 

173

Korsgaard, Christine M. 1983. ‘Two Distinctions in Goodness’. Philosophical Review 92: 169–195. Korsgaard, Christine. 1986. ‘Skepticism about Practical Reason’. Journal of Philosophy 83: 5–25. Kraut, Richard. 2002. Aristotle: Political Philosophy. Oxford: Oxford University Press. Manne, Kate. 2014. ‘Internalism about Reasons: Sad but True?’ Philosophical Studies 167 (1): 89–117. Markovits, Julia. 2011. ‘Why Be an Internalist about Reasons?’ In Russ Shafer-Landau (ed.), Oxford Studies in Metaethics (Volume 6). Oxford: Oxford University Press, pp. 79–106. Moore, G. E. 1903/1993. Principia Ethica. Cambridge: Cambridge University Press. Moore, G. E. 1922. ‘The Conception of Intrinsic Value’. In his Philosophical Studies. London: Routledge & Kegan Paul, pp. 253–275. Parfit, Derek. 2011. On What Mattters, Vol. 1 and 2. Oxford: Oxford University Press. Popper, Karl R. 1957. The Poverty of Historicism. London: Routledge. Railton. Peter. 1986. ‘Moral Realism’. Philosophical Review 95 (2): 163–207. Ritchie, Katherine. 2015. ‘The Metaphysics of Social Groups’. Philosophy Compass 10 (5): 310–321. Ross, W. D. 1930. The Right and the Good. Oxford: Clarendon Press. Scanlon, Thomas. 1998. What We Owe to Each Other. Cambridge, Mass.: Harvard University Press. Schroeder, Mark. 2007. Slaves of the Passions. Oxford: Oxford University Press. Sciaraffa, Stefan. 2011. ‘Identification, Meaning, and the Normativity of Social Roles’. European Journal of Philosophy 19:107–128. Searle, John. 1995. The Construction of Social Reality. New York: The Free Press. Searle, John. 2010. Making the Social World. Oxford: Oxford University Press. Smith, Michael. 1994. The Moral Problem. Oxford: Blackwell. Swanton, Christine. 2007. ‘Virtue Ethics, Role Ethics, and Business Ethics’. In Rebecca L. Walker and Philip J. Ivanhoe (eds.), Working Virtue, pp. 207–244. Oxford: Oxford University Press. Thomasson, Amie. 2003. ‘Realism and Phenomenological Research 67: 580–609.

Human

Kinds’.

Philosophy

and

Velleman, David. 1996. ‘The Possibility of Practical Reason’. Ethics 106: 694–726. Williams, Bernard. 1981. ‘Internal and External Reasons’. In his Moral Luck. Cambridge: Cambridge University Press, pp. 101–13. Williams, Bernard. 1995. ‘Internal Reasons and the Obscurity of Blame’. In his Making Sense of Humanity, and Other Philosophical Papers. Cambridge: Cambridge University Press, pp. 35–45.

PART THREE

RO L ES, I N STITUTIONS, AN D OR G AN IZA T IONS

9 Role Ethics and Institutional Functions Sean Cordell

1. Introduction In Michael Hardimon’s ‘Role Obligations’ (1994), three claims together present a problem of indeterminacy, and a fourth offers a possible solution: (1)

At least in modern societies, many of our institutional social roles are ‘interpretative’ (1994, 336), such that we can reasonably disagree about what the duties of a particular role are or should be.

(2)

‘In order to settle what the requirements of a given role are, it is often necessary to determine how the institution of which it is a part is—or ought to be—structured’ (1994, 337).

(3)

‘Institutional concepts are no less interpretative than role concepts’ (1994, 337).

The problem is: (1), so (2), but (3). Regarding (1), Hardimon cites for example the ‘remarkable indeterminacy’ even of some social roles well established by tradition. ‘What, for example, are the duties of a wife today? What duties does a present-day husband have?’ (1994, 340). Given this, we might, (2), look for answers at how the institution of marriage is or ought to be structured. But (3), the indeterminacy problem shifts to the level of the institutions which specify those roles. The institution of marriage itself seems no less interpretative than the spouse roles. So where next? The germ of a solution is in Hardimon’s characterization of institutional social roles as: (4)

‘[C]onstellations of institutionally specified rights and duties organized around an institutionally specified social function’ (1994, 334).

The germinal idea is that institutions themselves have such functions—they each have some purpose or point—which determine those roles for individuals within those institutions. Consider some apparently more obvious, less interpretative, cases. The professional role-obligations of an optometrist are determined by the eye-hospital which employs her, the function or point of which is eye-health. The

Sean Cordell, Role Ethics and Institutional Functions In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Sean Cordell 2023. DOI: 10.1093/oso/9780192843562.003.0009

178



role-obligations and rights of the civic role of juror are specified as they are, and not some other way, because the point or function of the jury is (or should be) to reach a verdict fairly. The thought, then, is that any institution which defines roles—including less obvious, more interpretative ones (‘marriage’, ‘family’, ‘the state’)—will have some point, purpose, or function, and this is because all such institutions are the results of purposive activity. As with other things we make, such as buildings or tools, institutions are ‘created [or sustained] with a point in mind or to serve some specific purpose’ (Ehrenberg 2013, 448).¹ At the pre-theoretical level, this reflects much of our ordinary thinking and speech about how an institution ‘is or ought to be structured’. ‘The IT department is not fit for purpose’ or ‘this is a dysfunctional family’ have more than metaphorical sense. Regarding the ‘is’, we identify institutions using functional or purposive concepts. Several different hospitals, for example, may have many different attributes—different sizes and shapes, sets of staff, locations, and so on—but we recognize them all as hospitals because of one thing they all do, namely administer patient care. Regarding the ‘ought’, functional notions readily figure in ‘internal’ evaluations of social institutions, their policies and practices, and their roles. For example, when the conduct of press or other news media institutions is under public scrutiny, questions get asked about whether their practices (such as privacy invasion) are or should be outside the reach of law, or whether they are contrary to common standards of decency or morality. But the different internal question is that of whether news institutions and organizations, and their roles such as reporters and editors, are operating and acting as they should qua press institutions and qua those roles. At the theoretical level too, the need for a developed view of this kind—which for convenience I will call the functional view—is sometimes raised in ethical studies of professional roles. Writing on business ethics, for example, Christine Swanton says that: any work on professional ethics needs to supply a view about the point or function of the institutions in which roles are embedded. For . . . the point or function of those institutions determines the nature of the roles which individuals in those roles occupy. (Swanton 2007, 210)²

In this chapter I consider how far the functional view of institutional social roles could be generalized. For, if developed successfully, such a view could be central to a socially enriched ethics not just of professional roles as Swanton suggests, but of social roles per se. To do so would be a considerable undertaking given the variety ¹ See also Searle 2010, 7. ² See also Swanton 2016, 688 and 2020, 51. Seumas Miller (2019) points out that this too has been overlooked in contemporary political philosophy. Rawls, no less, proceeds ‘in the absence of a developed theory of the nature and point of the very entities [social institutions] to which the principles of justice in question are supposed to apply’.

    

179

of institutional social roles individual moral persons occupy, but a worthwhile one given that social roles are, as Robin Zheng puts it, the ‘site where [social] structure meets agency’ (2018, 870).³ I will say more about the role-indeterminacy problem in Section 2. In Section 3 I discuss three ways in which the functional view might be formulated in response: a backward-looking etiological approach (roughly, what some institution is ‘there for’ in the first place); a present-focused ‘practical’ view (roughly, what it is ‘used for’); and a forward-looking teleological account in terms of institutional ends (roughly, what it is ‘good for’). My aim is to use the shortfalls and strengths of each of these candidate accounts⁴ to arrive at a more coherent view. In Section 4 I propose such a view in terms inspired by Aristotle, that of function as characteristic activity.⁵ Before that, a note on ‘institutional social roles’. Following Hardimon, I take institutional social roles to include not only professional and occupational roles (professor, HR officer, train driver) but also roles and offices in political associations or recreational clubs (trade union representative, chess club treasurer), civic roles (citizen, juror), and those of families (sibling, grandparent). I will use these and similar examples to illustrate the indeterminacy problem outlined above and the potential for a solution framed in terms of institutional function. Beyond this I do not attempt to sort or define. I grant, and keep out of, reasonable controversy over whether some particular role is or is not institutional in the stipulated sense.⁶ Maybe there is a sense in which any non-idiosyncratic social role is part of some ‘institution’, conceived broadly as ‘systems of established and embedded social rules that structure social interactions’ (Hodgson 2007, 98). And if institutional roles do form a category within that of social roles more broadly, then I am happy with the possibility of borderline cases.⁷

2. The Role-Indeterminacy Problem Role-indeterminacy can arise in at least three different ways. First, some social roles or their requirements are interpretable because they are not well codified. But ³ I share Gregory Cooper’s hope that ‘a functional approach to institutional roles opens the door to a deeper analysis and, one would hope a fruitful debate, about the nature of role-based obligations’ (2020, 85). I am indeed indebted to Cooper here, his aim being to ‘start a conversation, not end one’ (85). ⁴ Though these are not the only candidates. Another such account, one which I do not have space to do justice here, could be developed in social contractual terms. See Sachs 2019. ⁵ I have discussed this elsewhere (Cordell 2011) in the context of neo-Aristotelian virtue ethics. ⁶ For discussions and criticisms of Hardimon’s categorization of social roles see Sciaraffa 2011; Simmons 1996, 28, esp. fn 2. Dare 2020, 35, Kisolo-Ssonko 2019; Jeske 2019, and this volume. ⁷ For example, Hardimon thinks ‘friend’ is not an institutional role, instead claiming friendships to be relationships (1994, 336). I am inclined to think that ‘friend’ may be a borderline institutional case, inasmuch as some requirements of being a friend are circumscribed not just by individuals in particular friend-relationships but by ‘friendship’ as a bunch of social rules and norms which stand outside any particular relationship (Dare 2020, 34–35). Whether anything is at stake in this disagreement, and if so then what, are questions for another day.

180



secondly, many well-codified role-obligations also require interpretation by those acting in the role. Thirdly, there are cases of ‘ersatz’ role-obligations which may be quite clearly, though incorrectly, specified by institutions. Let us look at these in turn. First, one way a role can be ‘interpretative’, as Hardimon claims, is when at least some of its requirements or obligations are non-codified, or not clearly codified. The most obvious example is of informal behavioural norms of roles not being specified explicitly in rules, regulations, codes of conduct, or role-descriptions. We reasonably expect a schoolteacher to behave in certain ways qua holder of their role, for example to be both pastoral and authoritative with students. Employees in organizations may be supposed qua members of that organization to cultivate certain attitudes and manners of collegiality with each other. There may be certain commitments and a team-spirit which players are expected to engender by virtue of playing not just football but for this, their, football club. These norms and expectations are role-requirements and they are institutional: specific to and determined by professional standards, or an organization or club’s culture or ethos (Miller 2010, 49–50). Note also that a role’s codified or formal requirements are not necessarily stronger than its non-codified or informal ones. To continue with Hardimon’s example, marriage typically incorporates some well-codified legal rights that are signed up to and some duties that are sworn ceremonially. But it is unclear to what extent these determine what is required of individual spouses. Assuming social consensus that marriage is, minimally, a mutual commitment to a loving partnership, different people can and do have different ideas about what is ‘in the deal’: which if any rules are set by such commitment, what if any duties it prescribes. Over time, too, the broadly consensual minimal view of a role can itself shift and vary according to prevailing norms and attitudes, such that the basic social conception of the role and its requirements can come to supersede or dwarf a previous one, even if the codified legal ones stay much the same. Perhaps in many societies a widely held minimal perception of the spouse role now is that ‘wife’ and ‘husband’ stand either side of a mutually reciprocal and equal partnership, and not that the former is subordinate to the latter. Secondly, some clearly codified and well specified role-obligations can be interpretable, to which extent they are at least not fully determinate. Within professional roles’ directives and parameters, which may be clearly laid out for role-holders, there can be significant variation of opinion about when ‘getting the job done’ remains within the rules. The legal profession is a source of good examples, which is perhaps one reason why it has generated a role-ethical literature of its own. A defence lawyer can pursue litigation within the law and according to explicit legal regulations, and in doing so coach her clients to give testimony by ‘jogging’ their memories to identify details of which they almost

    

181

certainly will have had no recollection (Wendel 2020, 156–157).⁸ This, or other less than honest practices, may be permissible within professional regulations. And the interesting role-ethical problem is when or whether such practices appear to be role-requirements, these being sometimes an expedient—perhaps necessary— means to fulfil the role’s central task of pursuing a client’s defence. In professional ethics this kind of problem tends naturally to be cast as a conflict between role-obligations on the one hand and the requirements of ‘ordinary’ or ‘general’ morality on the other (Walker and Ivanhoe 2007, 26). The problem is that: if there are dirty deeds the good person should not do, yet their commission is sometimes required of the ‘good lawyer’, then seemingly someone cannot be a both a good person and a good lawyer. Straightforwardly interpreting the role and its requirements fails to solve the problem. On the contrary this is what causes it, deception being sometimes required to fulfil that basic—perhaps constitutive—roleobligation to defend clients. One kind of response to this is to argue that a genuine role-requirement of a generally good role will thereby not—all things considered—be an immoral requirement. Prima facie immoral deeds of deception may sometimes be required properly as a legitimate part of the lawyer’s job within its rules of engagement, when (and only when) consistent with its purpose of promoting just outcomes for clients. The response from the other direction is that a generally good, just, role such as one in the legal profession cannot genuinely require or oblige its occupants to do immoral things such as deceive. On a better understanding of what the good lawyer is within the context of good law and good legal practice, we will see that mendaciously jogging clients’ memories—to ‘recollect’ things they probably never even collected—is not something that could be required of a good and conscientious lawyer. Lawyers, we might think, are a long way from spies or undercover intelligence agents, for whom deception very much is part of their role (Rawls 2000, 354). Thirdly, though, there is another quite different perspective on this problem. We could agree in part with the thought that sometimes deceptive chicanery is in fact a requirement of the defence lawyer’s role. At the same time we could agree with the other response, that it should not be required or even permitted of good lawyers qua lawyers. From this perspective, what has gone wrong is precisely that ‘be deceptive-when-expedient given certain circumstances’ is, or has become, a role-requirement for the lawyer. This then is not the study of how good moral individuals should best interpret role-requirements if they want also to be good lawyers. Rather, we are at the next level up, at the institutional misspecification of those requirements. The substance of such a misspecification claim could be that the adversarial legal system itself, or a culture within it of win at all costs, or

⁸ Wendel cites the case of Jackson v Johns-Manville Corp., 1336–1341, including appendix.

182



financial incentives, or ready hiring and firing, or the dog-eat-dog greasy career pole, or what have you, have distorted the defence-lawyer’s proper role. Institutions which define and maintain the role have sent it, or allowed it to veer, off course by specifying the wrong or ‘ersatz’ role-requirement to act deceptively when expedient. On this view the normative focus is not that of individuals’ rightly or best interpreting their institutionally determined role-requirements, but of those requirements being wrongly institutionally determined. There are two striking things about this kind of institutional misspecification or ‘ersatz role-requirements’ problem. For one, it can arise in all sorts of roles and, for another, it is not confined to conflicts between roles and general morality. Prosaic examples abound, at least in modern role-pluralistic societies with which I assume readers will be familiar. Take an exasperated schoolteacher and her colleagues who all agree that what they should be doing in their roles is working with their school students, rather than administrating statistics and paperwork to meet targets. They come to express this agreement when, and because, they are all now required to spend most of their time doing exactly the latter task and far too little doing the former. The problem here is not that their institutional role might require them to act contrary to a moral principle, but that their institution is requiring them to act contrary to the role’s principles—or what they understandably see as something like that role’s principles. Someone may want to resist elevating the critique to the institutional level here, pointing out instead that the essential—or real—requirements of ‘schoolteacher’ are just there in the activity comprising the role’s title or role concept: that of teaching school pupils. This may well of course illustrate the worry about misspecification of role duties, as for example in the satirical air of ‘a schoolteacher too busy in her role do any teaching’. But a role title or its concept could itself come to incorporate things it used not to incorporate. So, the title ‘schoolteacher’ could have come now to describe a set of duties including administration. That this has happened to them may be a glaring reflection of what we think has gone wrong with the role, but it does not by itself substantively explain the injury the role has sustained. A more substantive diagnosis lies in the role somehow just not working or functioning as it should, where at root the fault is not semantic but occurs at some systemic institutional level. In this mode the teacher might instead then ask rhetorically: ‘what is the point of the schools and school-teachers?’ or ‘why are schools there in the first place?’ and so on. As we will see, those questions do not admit straightforward answers, but their rhetorical force is in their appeal to dysfunction or a failure to fulfil some purpose. Regarding the misspecification or distortion of role-requirements, we can see further use for the functional view in the potential amelioration of institutions. Rosalind Hursthouse, for example, sees the ‘good lawyer versus good person’ scenario as allowing rare but possible dilemmas in which someone may qua good lawyer be forced to do what is best overall though not ‘right without

    

183

qualification’ (2007, 249).⁹ Here she argues that doing so will then leave the conscientious lawyer with some responsibility to do what she can to put right the legal institution(s) whose requirements gave rise to that dilemma.¹⁰ In such a case, what blueprint or criteria are we to use for modifying this specific institution for the better? In what direction is one to go the extra mile to improve it, not just in general moral terms but qua that sort of legal institution? What is needed is some notion of what this, specific, legal institution is supposed to be doing: what it is good at or good for. And some notion of the function or purposes or point of this or that institution appears salient.

3. Construing Institutional Functions I have suggested so far that such functional ideas are intuitive and latent in much of our thinking about institutions and roles, and so far the idea has been coarsegrained, with the different terms and concepts ‘function’, ‘purpose’, and ‘point’ in play. We can bring some focus by considering functional explanations. These have the form: ‘the function of x is to f ’; or ‘x is there for f ’. This kind of explanation has been discussed extensively in the philosophy of biology,¹¹ where functional accounts of natural phenomena are in one sense puzzling. To say that ‘the function of birds’ wings is to aid flight’ seems to be to say ‘wings aid birds’ flight’ and that ‘they should aid birds’ flight’, the puzzle being how things like wings have or could acquire such normative properties.

An Etiological Approach An apparent advantage for institutions in this respect is that their having a function seems not to generate the same puzzle, because they are the products of purposive action, and as such they are artifacts. Corporations, to take one example, look to have been designed: ‘humans make and could make them differently if they wanted’ (Donaldson 2019, 857).¹² So here we might turn to the notion of design function which fits artifacts. A teacup’s design function is to hold hot liquid for people to drink, so its normativity is less puzzling: it ‘should’ hold hot drinks because that is what it was designed for. ⁹ See also Dare 2009, 53–54. ¹⁰ Occupants of roles may be well placed to effect that reform (Collins, this volume). Zheng (2018) argues that social role-occupants can have wider societal obligations to reform their own roles and institutions. ¹¹ Though also in the philosophy of scientific explanations more generally (Wouters 2005); the social sciences (Kincaid 1990); the philosophy of mind (Price 2002); epistemology (Plantinga 1993); and aesthetics (Parsons and Carlson 2008). ¹² See also e.g. Crowe 2014.

184



Of course, one problem with ascribing this kind of design function to institutions in general is that some of them seem never to have been designed in the relevant sense, that is, in the way that teacups have been. However, even in cases where we can find no explicit design—no founders’ statement, no charter, no constitution—we may nonetheless be able to infer certain intentions behind the creation of an institution. Take a cooperative retail organization which was founded in the late nineteenth century and endures to this day. Suppose that through good historical sources we can infer that it was originally intended to provide fair and affordable trade for those in abject poverty who were at the mercy of ruthless profiteers. This might give us ‘what it is for’ in a sense close enough to the ‘what it was created to do’ sense which also underlies design function. This still seems to leave out important cases of institutions which clearly came about at some point—as social institutions their existence cannot fail to be due to intentional human activities (Leiter 2011, 666)—but to which we cannot easily attribute or infer any human design or intentions (Green 1998, 117–118). Even in these cases, however, we can impute functions through reference to their creational origins, without having to find or infer the intentions of any specific human agents. Absent identifiable plans, designers, makers, or even particular intenders, ‘design’ may be too far a metaphorical stretch here. Nonetheless the idea of original function—the function for which the institution came into being in the form it did—might remain explanatorily useful. We could find certain sociopolitical, economic, conditions which gave rise to its creation and explain how it came about in the form it did, in terms of it serving the interests of a dominant social class in those conditions (Cohen 1978). Suppose for the sake of argument (or else just because it is entirely plausible) that in such conditions the traditional nuclear family arose in the form of a particular social structure which facilitated the subjugation of women through their role in that family structure. With the advent of civilization and class society, the twin changes [the rise of private property and the unleashing of new social forces] were consummated: the private property system was victorious, and women plummeted from their former high status to become the oppressed sex of patriarchal society. (Reed 1972, 10)

Design, intention, or origins might then be grouped together under a broadly causal or etiological—what it is ‘there for’ in the first place: what made it happen— view of institutional functions. An advantage of this way of thinking about institutional function is that it accounts for institutions all having been created: their being socially instituted at some point and for some point. The main problem with the etiological view as a general approach, however, is its failing to track or grasp the common-sense concept of institutional function by which we identify and evaluate institutions,

    

185

and which is what we are after in the context of the role-indeterminacy problem. In some cases, the original function may relate closely to later understandings of its function, whereas in others it does not. It is thus a contingent matter whether the design or original function of some institution will map on to the putative identifying and evaluative sense—roughly, ‘what it is for now’. How the current cooperative retail organization operates, and what it is widely understood to be, accords directly with its founding purpose and equitable principles, and quite possibly with some explicit design in a charter or mission. By contrast, assuming the original function of the nuclear family to be the patriarchal one, there seems no good reason to suggest that this has anything to do with the current function(s) of this or any form of the family, still less with how we value the family. Whatever else it may do or wherever it came from, families function as units of social support, love, bonding, and nurturing. This is what makes sense of our calling the ones which fail to do these things ‘dysfunctional families’.

A Practical Conception: Use Function We might therefore move from the past to the present, from an etiological conception to what I will dub a practical one. On this view we equate talk of an institution’s function, not with its origins or design but with how that institution is used, its use function (Ehrenberg 2016, 12; Achinstein 1977). In the case of artifacts, for example, an old tyre can have a use function as a garden swing on the end of a rope. This is clearly different from its design function, which is to cushion and roll a wheel on the ground (Preston 1998, 220). To take another example, a castle built as a military fortress in medieval times can today have the use function of attracting historically interested tourists. Accordingly, with institutions, it could be that the concept of function best employed when thinking of the cooperative is original design function, while for the family we should look to its use function(s), that is, the function(s) it has since come to acquire. An advantage of the practical, use-functional idea is that seemingly any extant institution will have one (or more) on show, even in cases where original functions are obscure. Where an institution has not gone out of existence and is thus ‘working’, it is being put to some use(s), thus fulfilling some use function regardless of any for which it may have been created or came about. One problem now, however, comes in disentangling use function(s) from design or original ones. Since its inception the nuclear family could have always fulfilled the functions we now want to distinguish as its ‘use functions’—mutual support, care, raising children, love, and so on. Plausibly there will long have been many people who formed families and raised children with intentions of fulfilling exactly these putative functions. But what we want to bracket here as use functions of the family will also have played some causal-constitutive role in the family alongside

186



the original patriarchal function (which may continue to be constitutive insofar as the modern family perpetuates patriarchal structures).¹³ To illustrate the problem consider how, by contrast, we seem better able to delineate and ascribe separately the design and use functions of artifacts such as castles. The castle’s use function as a historical tourist attraction is of course strongly connected to its original design function as a fortress, after all it only functions now as a historical attraction to tourist because of the design function it had and once fulfilled. But it clearly does not now fulfil its original fortressfunctions of military defence, the ones for which it came about in the first place. In this case we can much more clearly distinguish and ascribe these two different types of functions (design and use), which the castle has had at different times, than we can in the nuclear family case (with its original and use functions). This is not to suggest that it is always straightforward to ascribe function or distinguish function types in cases of non-institutional artifacts.¹⁴ But the contrast between the examples is meant to reveal that especially acute difficulties attach to ascribing artifact-type functions to extant social institutions. These difficulties are due to extant institutions comprising sets of human activities, not just being the result of them. The key difference is between artifacts such as tyres or buildings that are created and then endure, and institutions such as the family, whose creation endures. Insofar as some institution typically endures as an identifiable entity, its causal story includes its continued ‘use’ by intentional human agents over time. In typical cases of enduring institutions, then, ‘use’ is or can be part of design and creation. Compare again biological functions such as those of bodily organs where, as Fred Dretske has observed, ‘biologists discovered these functions; they didn’t invent or assign them. We cannot, by agreeing among ourselves, change the functions of these organs.’¹⁵ By contrast it is far easier to invent, assign, or change social institutions’ functions. Institutions are of course historical phenomena the different causes and effects of which we can interpret in various functional terms, as we might do with teacups or buildings. But unlike other artifacts made by and then used by humans, as constellations of human activities and projects they are something like ‘social works in progress’. Created and sustained by human projects and intentions, they both came into being at some point and continue to happen. Institutions are to that extent themselves apt to modification and susceptible to new kinds of activity, and thus open to normative arguments about what their activities and functions are or should be.¹⁶ To put some flesh on these bones and re-centre on the role-indeterminacy problem, think back to the example of the non-codified requirements of marriage and its roles. Different and changing interpretations of its requirements including, ¹³ See Moller-Okin (1989). ¹⁵ In Plantinga (1993, 4).

¹⁴ See Preston 1998; Kelemen and Carey 2007, 214–215. ¹⁶ See MacIntryre (1973, 7).

    

187

importantly, interpretations by those occupying its spouse roles, have played no small causal part in how we have come in time to understand the institution and its roles. Other interpretations may continue to emerge as long as we have the institution of marriage and the spouse roles it circumscribes. The problem is shown too when we return to wrongly specified or ‘ersatz’ rolerequirements. Imagine, if you will, that lurking within some university there are pressures towards managerialism, marketization, and all that goes with them. A traditionalist academic, seeking to defend her role against this perceived encroachment, might insist: ‘that is not what a university is for’ (much as the schoolteacher did in the earlier example). But appealing as her view of the university may be for other reasons, how much weight here can we put on such purposive or functional foundations? As Michael Blake makes clear, even if we were to establish that an institution has some purpose or function, that would not necessarily be sufficient reason to preserve that purpose or function: What it means to be a university might be subject to social and political alteration. If a democratic citizenry voted to make universities focus exclusively on jobs training and abandon a commitment to the production of knowledge, I think that citizenry would be making a mistake, but I do not think they could be shown to be mistaken with reference to the purposes of the university itself; it is precisely those purposes that the citizenry is proposing to change. (Blake 2011, 283)

In short, why think that institutions’ purposes are especially normatively robust, given how we can contend and change them?

A Teleological View: Functions as Good Ends We might address this problem with yet another way of thinking about institutional functions, this time focusing on specific good ends or goals for humans which social institutions promote. As has just been emphasized, any social institution comprises human activities and purposes. So, in cases of pro-social roles and their determining host institutions, we could construe the function or purposes of those institutions in terms of certain good ends they promote for humans. This approach would also fit with another kind of general functional explanation put in terms of ‘good-consequences’. ‘The (a) function of x (in S) is to y if and only if x does y (in S) and doing y (in S) confers some good (upon S, or perhaps upon something associated with S)’ (Achinstein1977). Such a view could, for example, be articulated in terms of a type of good which all and any pro-social institutions should serve, such as human well-being or flourishing. The standard by which we judge the good and well-functioning family

188



could be the extent to which it facilitates, for example, flourishing for family members, and has structures and practices which tend to promote, or at least not inhibit, it. Flourishing could be considered broadly to include a flourishing community or society, rather than only those individuals within or directly affected by a particular institution at a given time. Having families might be conducive to societal flourishing, not just the flourishing of those within families.¹⁷ Versions of something like this ‘good ends’ view of institutional function have been proposed in discussions of specific roles and in applied ethics. In legal ethics W. Bradley Wendel suggests that on a functional understanding, when we ask what makes some professional role-holder a good one, we will look at how it contributes to systems and institutions of which it is part and ‘the human good for which [these systems and institutions are] constituted’ (2020, 151). Gregory Cooper claims similarly that ‘[t]he point behind [social] institutions is that they are conducive to (or in some cases necessary for) some good ends or set of ends’ (2020, 79). Justin Oakley and Dean Cocking (2001) propose that valuable professional roles within institutions of medicine and law are each subject to certain regulative guiding ideals of health and justice. A strength of this approach in general is that it links goods or valuable goals to widely understood institutional functional categories. We rightly suppose that the goal of medical institutions is health, educational institutions are for knowledge, legal institutions serve justice, and so on. The problem with this view, as a general approach in role ethics, is its limit on how far its lens can zoom in on the many particular institutions which determine our many social roles. To a great extent the general or abstract human goods or ideals, to which we link institutional categories, are particularized and made concrete, and in different specific forms, by different institutions within those categories. Clinics, hospitals, or surgeries may all have or serve ‘health’, but so do gymnasiums or leisure health clubs for example, so to identify the medical institutions we need to be more specific. So, we may then refine the analysis and cite the human good or goal of the medical institutions in question as ‘healthcare’. But then, on closer examination we also have more specifically ‘clinical healthcare’, ‘hospital healthcare’, ‘surgery healthcare’, ‘mental healthcare’, or ‘dental healthcare’, and so on. Each of these are ‘modes’ of healthcare (Ehrenberg 2013, 453) which individuate different health institutions, and there are further fine-grained modes. Psychiatric healthcare is part of mental healthcare, and one hospital may have within it both an end-of-life care unit and a post-natal clinic, for example. For each institution and sub¹⁷ Someone sceptical of the link between social institutions we want or need to maintain on one hand and human goods such as flourishing on the other might offer, for counter-example, penal institutions of law which mete out hard punishment. But, on the assumption that prisons or legal punishments (or at least the authority and means to impose punishment) are needed in a well-ordered society, the function of a punitive institution might be construed as conducing to the flourishing of the citizenry considered as a whole.

    

189

institution, its distinct way of working will serve the good end of health or healthcare, but its ways of working are not specified by the good ends that all of them qua healthcare institutions serve. Consider this granularity through an analogy with tools. If we selected at random any tool from a bicycle toolbox and asked ‘what does this do?’, one correct answer would be: ‘it maintains bicycles’. But where we have picked up say, a chain-link remover then that answer would not get us far with working out what it is supposed to do: its function or purpose. And if the question were put in terms of ‘what it this tool good for?’, then recognizing its necessary connection with the multiply realizable good of bicycle maintenance would also be true but similarly unhelpful. We might then try to zoom in further on some more specific distinctive good(s) or benefit(s) of certain institutions. For example, Elaine Sternberg has appealed to the function of business and argued that ‘[t]he defining purpose of business is maximizing owner value over the long term by selling goods or services’ (2000, 32). A manufacturing business is not a hospital or care institution, or a charity, nor should a business be taken to share the same purpose(s), even if they might share some coincidental benefits. ‘To incorporate extraneous elements into the definition of business simply because they are perceived to be good, or important, is to distort the truth. It is to confuse the desirable with the essential’ (2000, 33). Sternberg’s methodology is to identify what any business should be doing or not doing in terms of business’s individuating sine qua non. Notwithstanding some disagreement over what the defining purpose of business is (Donaldson 2019), in those terms this may be successful in getting at the key function of any and all businesses qua businesses. What it cannot do is tell us much more about different particular business organizations’ functions or purposes qua whatever they each do. There are of course many different sorts of business, and what they do will in important ways simply not be the same, even if qua businesses they all have the sine qua non. A local pub is a business, for example, but this is surely not its function even though it functions as a business in much the same way as a dogfood manufacturer business. To overlook the pub’s function qua space for certain social activities (and not others) would be to fail to grasp the sine qua non for pub and seem simply to ‘distort the truth’ about what it is there to do, or what it is there for. So much for what might seem banalities about differences among institutions and organizations. But this is of particular importance when we are looking for a way of determining role-obligations and requirements by considering the function of the particular institutions that host and circumscribe those roles. Think of the diverse activities of the role-defining organizations nested within larger institutions, such as the IT department within the hospital, within the local health authority, within a national health service. Or consider that we can think of ‘professor’ as a role within a specific educational institution or within a

190



research institution. Narrowing down to ‘university professor’ depending on the role-requirements we want to focus on in our analysis, we can identify the role at the specific level of a particular university—or perhaps narrower still, index that role to a university with particular subject specialisms—as an organization which specifies its professors’ roles. Or on the other hand we can think of ‘University Professor’ as a role-type whose requirements derive at some general level from the institution of ‘the university’, or from ‘academia’, or even the overarching social institution of ‘higher education’. The upshot for the human goods view is not that it will fail to tell us important things about the ways institutions we value do and should work—it does this. But it only gets us so far towards addressing the role-indeterminacy problem. Many different institutions can share goals or serve the same goods but each of them particularizes its own goods and functions—does the things it does—quite differently. And these are many and various given the sheer variety of institutions which circumscribe different roles.¹⁸ The point about specificity of role-defining institutions can be seen by comparison with other functional accounts of large-scale social phenomena. For example, functional explanations or definitions of law seek to identify some function(s) which any legal institution will necessarily serve. But an adequate functional account of the many different particular legal institutions that determine roles (‘jury’, ‘judiciary’, ‘high court’) would not follow necessarily from such an account or definition, however successful it may be in its own right.

4. Aristotle and Characteristic Activity In the discussion in Section 3 of these three broad approaches to institutional function, one etiological, one practical, and the other teleological, I have sought to show how they capture our intuitive functional thinking about institutions but also to highlight key difficulties. The past-facing etiological approach runs with the thought that institutions, like artifacts, are intentional human creations, and the present-facing practical view recognizes that they are things we put to purposeful uses which may diverge from their design functions or original functions. But together these accounts face the problem that institutions are typically, unlike many other kinds of artifacts, social works in progress. This makes it difficult in certain cases to establish which purpose(s) or kind(s) of function(s) we should look at to individuate, evaluate, and recognize, some institution qua that institution, which is what is needed to address the role-indeterminacy problem. The forward looking teleological story assumes that institutions are or should be oriented to serve human goods and good ends, but as a general approach this ¹⁸ See Moore 1992; Green 1998; Murphy 2003, 257–260. See Ehrenberg 2016 for a comprehensive discussion of functional accounts of law, and a defence of his own.

    

191

view cannot accommodate the specificity of the many different institutions which in turn specify social roles and how they operate. This is something the functional view will need to do if it is to be of use in tackling the role-indeterminacy problem we began with. At this point we might want to resign ourselves to the thought that functional or purposive thinking about institutions and their roles is necessarily vague, and leave things there. Perhaps our talk of what things like universities ‘are for’ is usually mostly rhetorical and always fairly loose. And perhaps that is rightly the case because complex, diachronic, and contestable institutions (Miller 2010, 25) are just not the sort of thing to which we can attribute anything like a fixed or stable function or set of functions, at least in any sense that is useful for thinking about role-indeterminacy. But given the problem we started with and the prevalence and normative sway of the many institutional social roles we may occupy, it may be worth considering how we could start to conceptualize function, purpose, or point in a different way. I will therefore give a hearing to Aristotle’s ‘function argument’,¹⁹ which in the Nicomachean Ethics he (in)famously applies to the subject of the human being. Aristotle’s claim is that what counts as a good human life may become apparent once we gain an understanding of what a human being qua human being characteristically does, thus hitting upon the notion of a thing’s being good as an instance of its kind by virtue of its fulfilling its function. The form of Aristotle’s ergon argument—‘ergon’ being variously translatable as ‘work’, ‘task’, or ‘function’, though rendered most usefully as ‘characteristic activity’—is applied by him to natural life forms’ biological processes of development and survival. A plant does well by doing what it characteristically does, by soaking up nourishment from soil that bears the appropriate nutrients. A plant that fails to do these things will be a poor plant.²⁰ Precisely what Aristotle was getting at by applying the ergon argument to humans is controversial.²¹ It seems that, as Christopher Megone observes, ‘Aristotle presents his function argument . . . as a resource to help him clarify what constitutes eudaimonia’ (1998, 188), that is, the characteristically human, well conducted, ‘flourishing’ life. But in any case, the form of Aristotle’s argument— on plausible interpretations—offers insights for us in the present context if we apply it to institutions. ¹⁹ Aristotle 2002, Book I.7. ²⁰ Cf. Foot (2001), esp. chapter 2 and Rosalind Hursthouse (1999), chapter 9. See also Thomson 1997. ²¹ For example, it has been debated whether Aristotle is contending that the life of perfect virtue would, ultimately, be the life of theoria—purely intellectual contemplation in pursuit of sophia— intellectual wisdom; or whether the non-intellectual virtues such as courage also figure essentially in such a life, even if a life fully devoted to intellectual contemplation was achievable. The former interpretation of Aristotle has been called the ‘intellectualist’ or the ‘dominant’ reading, the latter the ‘inclusive’ reading: see Richard Kraut (2000), and Hughes (2001, 43–44).

192



The first way it can do this is allow us to conceive of ‘function’ not as purpose but as ‘functioning’, as Christine Korsgaard has in her exposition and defence of Aristotle’s function argument. I think it is helpful here to distinguish two possible senses of ‘function.’ In many cases it is quite natural to identify a thing’s function with its purpose, with what it is for or simply what it does. . . . Function can refer to the way a thing functions or how it works, to its function-ing. If we use ‘function’ in this sense—‘how a thing does what it does’—it will diverge from ‘purpose,’ which is simply ‘what it does.’ Consider, for example, a complicated machine. Such a thing might have many purposes, but in the sense I am discussing now it has only one function— one way of functioning. For instance, a computer serves a great variety of purposes, things as different as word processing, solving mathematical problems, writing music and playing chess. But to describe its function, in this second sense, is to describe what we might call its functional construction, the mechanisms that enable it to do all these things. Superficially, we might say that its function is the electronic storage and retrieval of information according to a program, or some such thing. But in the strict sense, only someone who actually understands how computers work can tell you what their function is. Or, to take another example, you could say of a radio that among its purposes is to broadcast music and live entertainment, provide a medium for advertisement, keep people up to date on the news and serve as an early warning system in an emergency. These are all ‘what it does.’ But if we wanted to talk about ‘how it does what it does’ we would have to talk about transmitting electromagnetic waves of certain frequencies and rendering them audible, and about the mechanisms that make this possible. The various things the device does are its purposes; the second thing, how it does all this, is its form or function. (Korsgaard 2008, 138)

How might this help our thinking about institutions? The insight here is that on this view the ergon of some institution would not be its purpose or bunch of purposes. Rather, it is understood as a certain configuration of activities and structures that fulfil those purposes in a certain way. One advantage of this understanding is that it takes us away from linking function to intentions or purposive design, where intentions can be various and purposes can be manifold and changing. Another is that, whilst we may understand a number of particular institutions in terms of their serving the same good end, conceiving function this way, as something like ‘characteristic activity’, is to conceive it in terms of the specific means whereby, and the specific ways in which, it serves that good end. Emergency hospitals, clinics, care hospices, and dental surgeries all serve healthcare, and each does so in their characteristically distinct ways. The problem now, of course, is saying which of any given institution’s ‘ways of doing things’ comprises its functioning in this ergon sense—its characteristic

    

193

activity—and which do not. On which there may be a second helpful insight from consideration of Aristotle. A potentially misleading aspect of the ergon argument as applied to humans is its appeal to uniqueness, that is, the idea that humans’ capacity for reason is our distinctive function by virtue of its not being shared by other lower life forms. Aristotle’s identification of the ergon of humans is based on categories in nature, where a certain form of life has a feature that is distinctive (idion) if and only if all normal members of that life form have that feature and no members of a lower form of life have it.²² Humans, like other forms of life, have functions of nutrition, perception, and growth, but the feature which distinguishes a rational agent from lower forms of life is the activity of theoretical and practical reasoning. Now, as has been noted in criticisms of Aristotle’s argument, a unique feature of a thing seems not to equate to or track its function. Applying the criticism to the institutional case, an enduring feature of a club could be that it meets on a particular day at a particular time of day which no other club does. But it would seem a conceptual contortion to think that this feature is thereby part of its function. As Stephen Everson notes, however, it is a mistake to suppose that Aristotle saw a property’s being ‘unique’—idion—to a species only as its peculiarity to that species.²³ Had he done so while also claiming that what is idion to something is also its characteristic function, its ergon, he would be committed to treating peculiarly human activities such as Morris-dancing or prostitution as part of the human ergon (1998, 88–89). But for a feature to be idion is, as Aristotle says, for it to be ‘ “convertibly predicable” with what possesses it’ (89). Everson explains that ‘it is not sufficient that a capacity should be possessed only by humans: it must also be the case that anything which is human has that capacity. This will knock out prostitution . . . and quite possibly Morris-dancing as well’ (89).²⁴ Taking a similar line, we could include in the ‘characteristic activity’ of some institution (i) the good or set of goods it serves and (ii) the particular configuration of its features and activities by which it serves that good or those goods. This will individuate the institution through an understanding of its sine qua non features and looks to accommodate both a broad typology such as we saw in Sternberg’s analysis of business—any sports club will do sports—and more fine-grained individuation of institutions and organizations that define roles. Two sports clubs may have distinct ways of playing the same game, different missions, structures, and procedures, rules of membership, and so on. One club’s function

²² See NE 1, 7 (1097b22–118). Cf, Santas (1997, 267–268). See also Hughes (2001, 36ff.). ²³ This paragraph on Everson’s interpretation draws on Cordell 2011, 267. ²⁴ Everson (1998, 89) takes the Aristotle quotation ‘convertibly predicable’ from Topics I.5. Everson here goes on to argue that because pure intellectual contemplation—theoria—is, for Aristotle, a property we share with the gods, it cannot be idion to humans. For Aristotle, ergon may not anyway be identical with eudaimonia: there may be more to a good and characteristically human life than reasoning well.

194



in the ergon sense might be to provide access to that sport for young people and marginalized communities, another’s might be success at the highest competitive level. A third club might incorporate, say, a democratic constitution and an egalitarian ethos including rules ensuring fair team selection by elected committee. These are examples of ways in which each club functions differently to participate in the same sport. Function as characteristic activity in this way individuates an institution or organization in terms of ‘its own way’ of doing things. Some of these features may amount to a club having a quite specific function—we might also want to refer to this as its identity or, more figuratively perhaps, its ‘character’. In principle this allows for characterization at different levels of institutional granularity, depending on which institution we want to characterize. For example, I am employed by a university. Qua ‘university’ it provides the education, courses, and qualifications, and academic standards of other universities, and has similar structures and roles to them. Zooming in on my particular university, however, it is distinguished from these others by its function—its function-ing in Korsgaard’s sense—by having (a) open access to students which is effected by (b) its specific mode of distance learning using correspondence teaching. These two features comprise its ergon. That, then, is a conception of some institution’s function as its serving good ends by a particular configuration of activities that only it has but which single it out as that kind of thing. I think that it may, at least, refine and articulate some common-sense ideas about what institutions are ‘meant to do’ or ‘are for’, but with more particularity than thinking only or primarily in terms of their purpose(s) or ends. In the case of the family example, its original patriarchal function can be excluded from its characteristic activity on the grounds that this is inimical to the good ends the family can serve.²⁵ This is so even though—or especially as—the family, or certain kinds of family structure, still tends to instantiate something of its patriarchal function. On the other hand, in the cooperative retailer case, its original fair-trade function can be included, both because it serves the end of providing affordable goods and because this is a key feature that marks it out as the cooperative retailer and not some other type of institution, retail or otherwise. Where, however, does this get us with contested cases given the problem of institutional change? If, as social works in progress, institutions are typically sets of human activities subject to different human purposes, and there are cases of disagreement as to which purposes an institution should serve, then how do we

²⁵ See Kane 2019, who argues that ‘family’ should be reconceptualized in normative terms of its ‘unique primary purpose’ of nurturing and furthering the well-being of its interdependent members. Kane offers this as a functional definition, a commitment of which is that some ‘de facto’ family (say, a blood-related social group) is not a family if it fails to fulfil that purpose (2019, 72). I diverge from Kane’s approach insofar as a general view of role-determining institutions must, as I have claimed, capture cases where institutions go wrong—malfunction—in misspecifying certain obligations for their role-occupants.

    

195

cite a set of these activities as the distinctive individuating characteristics? After all, cases where we would have most call to substantiate ‘what it is for’ type claims, will be those in which an institution’s activities are contested. Imagine a debate involving those who favour reforming a university, embracing (or at least tolerating drift towards) marketization and so a shift towards more lucrative vocational training. Their academic opponents robustly defend the traditional conception of the university, knowledge for knowledge’s sake, free enquiry, and the like. Can the Aristotle-inspired framework sketched here help resolve such cases? One thing the characteristic activity view can do, here, is better articulate the quite intuitive sense of function which should rightly figure in the argument. For example, in his fine account of the university as an institution built on moral foundations, Seumas Miller claims (2010, 227–228), and rightly in my view, that full ‘corporatization’ of universities would mean not merely a change to but rather the death of the university as we know it. I suggest that the characteristic-activity view of function refines the purposive or functional sense of such a statement, thus getting clearer about what the institution is, and what is its special value, which would be saved or lost in allowing or resisting changes. The focus of the argument on this view is on a particular configuration of features in the form of the university, and on why and whether we should value that configuration. The traditionalist can argue that there are distinctive things that only ‘the university’ as we know it does in certain ways, and these are distinct from the specific ways training colleges, apprenticeships, business schools, and others work. Insofar as the reformer has in mind a different configuration, their proposal is, in a sense, to change ‘the university’ into something else, an institution that will have some other sort of function as its characteristic activity. It should be obvious, I hope, that this has knock-on implications for the shape of roles, and especially academic roles, within these various educational institutions. This then, is not to suggest that essential and unique features of institutions such as universities will be transparent, such that one could achieve knock-out wins in disputes about institutional reform by revealing these features. It is to say that where a normative argument about an institution appeals to its functions, purposes, or points, the subsequent debate can better proceed with the enriched conception of function as characteristic activity outlined here.

5. Conclusion I began, via Hardimon, with the problem of indeterminacy within the normative content of roles. Then in considering the prospects for a general purposive or functional view of their host institutions as a way of approaching that problem, I have proposed an Aristotelian-flavoured account of characteristic activity.

196



Beyond this, I hope also that the discussion has provided an example of when role ethics—conceived broadly as enquiry into the normative dimensions of roles—will necessarily overlap with normative enquiry into the structure of institutions. As with standard cases of professional and applied ethics, the roleindeterminacy problem starts with thinking how an individual should act in relation to the requirements of this or that role. But then, whether this or that role-determining institution is (or is not) correctly defining these role(s) and prescribing their requirements and obligations, and if so why (or why not), are themselves questions within the purview of any adequate answer. Role ethics will, and should, also be political philosophy.²⁶

References Achinstein, Peter. 1977. ‘Function Statements’. Philosophy of Science, 44: 341–367. Aristotle. 2002. Nicomachean Ethics. Sarah Broadie and Christopher Row (trans. & eds.), Oxford: Oxford University Press. Blake, Michael. 2011. ‘Review of Miller, Seumas. The Moral Foundations of Social Institutions. Cambridge: Cambridge University Press, 2009’. Ethics 121: 820–824. Cohen, Gerry. 1978. Karl Marx’s Theory of History: A Defence. Princeton: Princeton University Press. Collins, Stephanie. This volume. ‘Role Obligations to Alter Role Obligations’. Cooper, Gregory. 2020. ‘The Role of Roles in the Normative Economy of a Life’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons, and Obligation, pp. 72–91. Abingdon: Routledge. Cordell, Sean. 2011. ‘Virtuous Persons and Social Roles’. Journal of Social Philosophy 42: 254–272. Crowe, Jonathan. 2014. ‘Law as an Artifact Kind’. Monash University Law Review 40 (3): 737–757. Dare, Tim. 2009. The Counsel of Rogues: A Defence of the Standard Conception of the Lawyer’s Role. Farnham: Ashgate. Dare, Tim. 2020. ‘Roles all the Way Down’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons, and Obligation, pp. 31–44. Abingdon: Routledge. Donaldson, Thomas. 2019. ‘Androids and Corporations: Why Their Rights Derive from Purpose’. The Georgetown Journal of Law and Public Policy 17: 853–864.

²⁶ For their comments on drafts of this chapter I am grateful to Alex Barber, Derek Matravers, Mark Pinder, and Dan Cavedon Taylor. For discussions of much earlier iterations I thank Christopher Bennett, Sophie Grace Chappell, Joe Morrison, Carolyn Price, Robert Stern, an audience at Queens University Belfast in 2014, and all our colleagues in the Role Ethics Network from 2015 to 2018.

    

197

Ehrenberg, Kenneth. 2013. ‘Functions in Jurisprudential Methodology’. Philosophy Compass 8: 447–456. Ehrenberg, Kenneth M. 2016. The Functions of Law. Oxford: Oxford University Press. Everson, Stephen. 1998. ‘Aristotle on Nature and Value’. In Companions to Ancient Thought 4, Ethics, pp. 77–106. Cambridge: Cambridge University Press. Foot, Philippa. 2001. Natural Goodness. Oxford: Oxford University Press. Green, Leslie. 1998. ‘The Functions of Law’. Cogito 12: 117–124. Hardimon, Michael O. 1994. ‘Role Obligations’. Journal of Philosophy 9: 333–363. Hodgson, Geoffrey M. 2007. ‘Institutions and Individuals: Interaction and Evolution’. Organization Studies 28: 97–116. Hughes, Gerald. 2001. Aristotle on Ethics. London: Routledge. Hursthouse, Rosalind. 1999. On Virtue Ethics. Oxford: Oxford University Press. Hursthouse, Rosalind. 2007. ‘Two Ways of Doing the Right Thing’. In Colin Farrelly and Lawrence B. Solum (eds.), Virtue Jurisprudence, pp. 236–255. Basingstoke: Palgrave MacMillan. Jackson v. Johns-Manville Sales Corp., 750 F.2d 1314, 1336–1341 5th Cir., 1985.

Jeske, Diane. 2019. ‘Special Obligations’. In Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy . Jeske, Diane. This volume. ‘Are Obligations of Friendship Role Obligations?’ Kane, Laura Wildemann. 2019. ‘What is a Family? Considerations on Purpose, Biology, and Sociality’. Public Affairs Quarterly 1: 65–88. Kelemen, Deborah and Carey, Susan. 2007. ‘The Essence of Artifacts: Developing the Design Stance’. In Eric Margolis and Stephen Laurence (eds.), Creations of the Mind: Theories of Artifacts and Their Representation, pp. 221–230. New York: Oxford University Press. Kincaid, Harold. 1990. ‘Functionalism in the Social Sciences’. PSA: Proceedings of the Philosophy of Science Association 1: 341–354. Kisolo-Ssonko, Joseph. 2019. ‘Race and the Responsibility to Abide by the Norms of Unchosen and Unjust Social Roles’. The Monist 102: 172–186. Korsgaard, Christine M. 2008. ‘Aristotle’s Function Argument’. In The Constitution of Agency: Essays on Practical Reason and Moral Psychology. Oxford: Oxford University Press. Kraut, Richard. 2000. Aristotle. Oxford: Oxford University Press. Leiter, Brian. 2011. ‘The Demarcation Problem in Jurisprudence: A New Case for Scepticism’. Oxford Journal of Legal Studies 31: 663–677. MacIntyre, Alasdair. 1973. ‘The Essential Contestability of Some Social Concepts’. Ethics 84 (1): 1–9. Megone, Christopher. 1998. ‘Aristotle’s Function Argument and the Concept of Mental Illness’. Philosophy, Psychiatry and Psychology 5: 187–201. Miller, Seumas. 2010. The Moral Foundations of Social Institutions: A Philosophical Study. New York: Cambridge University Press.

198



Miller, Seumas. 2019. ‘Social Institutions’. In Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy (2019 edition). . Moller Okin, Susan. 1989. Justice, Gender, and the Family. New York: Basic Books. Moore, Michael. 1992. ‘Law as a Functional Kind’. In Robert P. George (ed.), Natural Law Theory: Contemporary Essays, pp. 188–242. Oxford: Clarendon Press. Murphy, Mark C. 2003. ‘Natural Law Jurisprudence’. Legal Theory 9: 241–267. Oakley, Justin, and Cocking, Dean. 2001. Virtue Ethics and Professional Roles. New York: Cambridge University Press. Parsons, Glenn, and Carlson, Allen. 2008. Functional Beauty. Oxford: Oxford University Press. Plantinga, Alvin. 1993. Warrant and Proper Function. New York: Oxford University Press. Preston, Beth. 1998. ‘Why is a Wing Like a Spoon? A Pluralist Theory of Function’. The Journal of Philosophy 95: 215–254. Price, Carolyn. 2002. Functions in Mind: A Theory of Intentional Content. Oxford: Oxford University Press. Rawls, John. 2000. A Theory of Justice. Oxford: Oxford University Press. Reed, Evelyn (ed.). 1972. ‘Introduction’. In Frederick Engels, The Origin of the Family, Private Property, and the State, pp. 7–22. New York: Pathfinder Press. Sachs, Benjamin. 2019.‘Teleological Contractarianism’. Journal of Social Philosophy 50: 91–112. Santas, Gerasimos X. 1997. ‘Does Aristotle Have a Virtue Ethics?’ In Daniel Statman (ed.), Virtue Ethics: A Critical Reader, pp. 260–285. Edinburgh: Edinburgh University Press. Sciaraffa, Stefan. 2011. ‘Identification, Meaning, and the Normativity of Social Roles’. European Journal of Philosophy 19: 107–128. Searle, John. 2010. Making the Social World: The Structure of Human Civilisation. Oxford: Oxford University Press. Simmons, Andrew John. 1996. ‘External Justifications and Institutional Roles’. Journal of Philosophy 93 (1): 28–36. Sternberg, Elaine. 2000. Just Business: Business Ethics in Action. Oxford: Oxford University Press. Swanton, Christine. 2007. ‘Virtue Ethics, Role Ethics, and Business Ethics’. In Rebecca L. Walker and Philip J. Ivanhoe (eds.), Working Virtue, pp. 207–244. Oxford: Oxford University Press. Swanton, Christine. 2016. ‘A Virtue Ethical Theory of Role Ethics’. Journal of Value Inquiry 50: 683–819.

    

199

Swanton, Christine. 2020. ‘Expertise and Virtue in Role Ethics’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons, and Obligation, pp. 45–71. Abingdon: Routledge. Thomson, Judith Jarvis. 1997. ‘The Right and the Good’. The Journal of Philosophy 94: 273–298. Walker, Rebecca L. and Ivanhoe, Philip J. 2007. ‘Introduction’. In Rebecca L. Walker and Philip J. Ivanhoe (eds.), Working Virtue: Virtue Ethics and Contemporary Moral Problems, pp. 1–40. Oxford: Oxford University Press. Wendel, Bradley W. 2020. ‘Crossing the Bridge’. In Tim Dare and Christine Swanton (eds.), Perspectives in Role Ethics: Virtues, Reasons, and Obligation, pp. 146–159. Abingdon: Routledge. Wouters, Arno. 2005. ‘The Function Debate in Philosophy’. Acta Biotheoretica 53: 123–151. Zheng, Robin. 2018. ‘What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice’. Ethical Theory and Moral Practice 21: 869–885.

10 Role Obligations to Alter Role Obligations Stephanie Collins

1. Introduction Many roles are situated within collective agents. These include occupational roles (e.g., ‘financial advisor’), community roles (e.g., ‘chair of the condominium owner assembly’), and political roles (e.g., ‘citizen’). There are many situations in which the demands of such roles appear to conflict with the demands of morality. For example, a financial advisor might face the demand to satisfy criteria for probation, where this requires putting advisees into high-risk investments, which in turn requires giving advice that is not in advisees’ best interests. Or the chair of a condominium owner assembly might face the demand to lower the costs of rubbish disposal within the condominium building, where this requires disposing of rubbish in environmentally unsustainable ways. Or a citizen might face the demand to pay taxes that will fund a clearly unjust war. When such conflicts arise, role-occupants face the question: does morality trump roles, or do roles trump morality? Many forms of role-based protest—such as employees’ whistle-blowing or citizens’ taxation boycotts—can be seen as motivated by the conviction that morality trumps roles. This chapter argues that the question, as just stated, often contains a false presupposition. The false presupposition is that there is a fundamental conflict between the demands of the role and the demands of morality. I suggest we need a capacious understanding of the ‘role’ at issue and the ‘demands’ the role produces. With a sufficiently capacious characterization of the role and its demands, we can see that the demands of the role are often (though not always) concordant with the demands of morality, such that there is no conflict between the role and morality. The result is that whistle-blowing or taxation boycotts (as well as more conservative forms of role protest) can be seen as the fulfilment of role obligations—not the flouting of role obligations in the service of a morality that is ‘external’ to the role and its obligations. The general proposal is this. When a role-versus-morality conflict appears to arise, role-occupants acquire an obligation to use (what I will call) the ‘fundamental’ demands of their role to do what they can to challenge or alter (what I will call) the Stephanie Collins, Role Obligations to Alter Role Obligations In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Stephanie Collins 2023. DOI: 10.1093/oso/9780192843562.003.0010

     

201

‘imposed’ demands of their role. These ‘role obligations to alter role obligations’ are distinctively role obligations, rather than garden-variety moral obligations, insofar as the role plays an ineliminable part in justifying (and determining the content of) the obligation to alter the role obligation. The precise demands of role-obligations-to-alter-role-obligations will vary with context. Sometimes, these obligations might require whistle-blowing or civil disobedience; at other times, they might require ‘working within the system’ in more incremental ways. I’m neutral on those details. The more basic, and theoretical, point of the chapter is this: insofar as there appears to be a tension between organizationally embedded role obligations and moral obligations, this tension can often be resolved via a proper conceptualization of what the ‘role’ amounts to. Embracing the role’s fundamental demands might require rejecting the role’s imposed demands, but it (often) doesn’t require rejecting morality. I begin in Section 2 by laying out a view of what organizations are and how roles relate to them. This demonstrates the breadth of the account and the features of organizations that produce (seeming) role-versus-morality conflicts. Section 3 elaborates on how (seeming) role-versus-morality conflicts arise. To resolve the conflicts, Section 4 distinguishes between the ‘fundamental,’ ‘imposed,’ and ‘derived’ obligations of an organizationally embedded role, using this distinction to explain how the seeming conflict can be resolved in a wide range of cases, such as the examples mentioned above. Section 5 concludes.

2. Organizational Roles As I’ll use the term, an ‘organization’ is a collective agent that is composed of over 150 people, where those people realize a structure that coordinates their divided labour. The coordination occurs via rules and hierarchical command relations. All of this is guided by a collective decision-making procedure.¹ Organizations are a species of the genus that is ‘collective agents.’² A collective agent is any entity that is composed of people who are united under a rationally operated collective decision-making procedure. Amongst collective agents, organizations are distinctive in a few ways: organizations have a large size, while other collective agents might have as few as two members; organizations have a structure, which less ¹ I develop this definition in Collins (2023, ch. 1). This definition follows a long tradition in sociology, including Max Weber’s (1968, vol. I, 223ff.; 1968, vol. III, 956ff.) characterization of bureaucracies, Rom Harré’s (1979) characterization of institutions, and Jonathan Turner’s (1997) characterization of institutions. See also Geoffrey M. Hodgson’s (2007) definition of organizations in management theory. Within philosophy, similar characterizations can be found in the work of Peter French (1984, 13ff.), Raimo Tuomela (2013), and Lisa Herzog (2018). ² The latter have been theorized by, e.g., Carol Rovane (1998), Christian List and Philip Pettit (2011), Deborah Tollefsen (2015), Kendy Hess (2018a, 2018b), Frank Hindriks (2018), Collins (2019), and others.

202



organized collective agents might lack; organizations divide role-occupants’ labour, whereas other collective agents might be less prescriptive on this front; and organizations have rules and hierarchical command relations, which more egalitarian collective agents might lack. As I’ll now explain, these various distinctive features imply that organizations are particularly apt to produce the roleversus-morality conflicts that this chapter aims to address. To begin to see how organizations are ripe for role-versus-morality conflicts, consider the sense in which organizations have a ‘structure.’ A structure is a collection of roles that stand in relations. An organization’s structure thus produces a differentiation between various roles: the roles are differentiated by the relations they stand into one another. (In those collective agents that lack a structure, there is therefore also a lack of ‘roles,’ in the sense I’ll use that term.) In an organization, the roles are jobs or tasks, while the relations between the roles are usually reporting and delegation lines. The roles and their relations can be highly diverse, and more than one person can occupy each role. Both the roles and the relations are representable in an organization chart. I follow Katherine Ritchie (2013) in identifying the organization with a realization of a structure. A structure becomes realized—and, thus, an organization exists—when a sufficient number of the roles in the structure’s chart are occupied. Exactly what number of (and which) roles must be occupied, in order for the organization to exist, is a matter determined by the structure itself. By viewing organizations as realizations of structures, we can integrate them into a naturalistic metaphysics. That is, by viewing organizations as realized structures and by including those realized structures in our ontology, we do not posit anything ‘spooky’ or ‘mysterious.’ Yet including organizations in our ontology allows us to attribute purposes or goals to them. Those purposes and goals are important for role obligations, since, as we’ll see, organizations’ purposes and goals are the grounds of role-occupants’ role obligations. An organization’s purposes and goals are dictated by its structure, including when that structure is used to produce new purposes and goals. And an organization’s purposes and goals produce a kind of internal normativity for organizations—which, in turn, produces (seeming) role-versus-morality conflicts. I said above that roles are jobs or tasks, which are related by delegation and reporting lines. For the argument that follows, it’ll be important to appreciate the breadth and diversity of norms that are entailed by any given ‘job’ or ‘task.’ I’ll assume that the demands of a job or task include all norms that are operative within the organization, as well as the more specific requirements of the particular role. Following Amie Thomasson (2019), the demands of an organization can broadly be divided into ‘internal,’ ‘structuring,’ and ‘external’ demands. As Thomasson puts it, internal demands regard ‘how members of the group are to behave, regard themselves and other group members, and so on (what they are to wear, how they are to eat and prepare food, what they are (and are not) to do . . . )’

     

203

(2019, 4838). Structuring demands ‘place different group members at different nodes, with different norms regarding those who occupy different nodes (members of the congregation, the priest, the organist, the bishop; the president, secretary, treasurer, member-at-large, new initiate; privates, corporals, lieutenants, colonels . . . )’ (2019, 4838). Finally, external demands are ‘norms regarding how members of that group are to be treated, regarded, behaved towards by those who are not members of the group. . . . one way social groups may be constituted is by their members having shared external norms of treatment (privileging, subordinating, or just different), based on any of many kinds of different ‘markers’ . . . ’ (2019, 4839) For example, an external demand on members of a bank might come from existing and potential customers: all members of a bank might face the external demand from non-members to do what they can to satisfy (potential) customers’ banking needs. Thomasson is thinking of social groups, such as genders, races, and classes— not organizations in the present sense. But these categories of demands or norms also apply to role-occupants in organizations. I will use a broad concept of a role’s ‘demands,’ which subsumes all the internal, structuring, and external norms associated with that role—not just the structuring norms, which are perhaps what would first spring to mind. As I hope is clear, the internal norms are perhaps the most ready source of role-versus-morality conflict. Internal norms in an organization might include norms of competitiveness, selfishness, ambition, cost-cutting, and so on. Such internal norms can often be rationalized by their usefulness in achieving the organization’s goals (such as selling bank loans)—even if the internal norms, and the goals they serve, conflict with morality. So much for the notion of organizations as constituted by role-occupants in a structure. When characterizing organizations, I mentioned other ways in which organizations are distinctive amongst collective agents: organizations have over 150 role-occupants; organizations divide those role-occupants’ labour; and organizations have rules and command relations. These three features compound to create a high propensity to role-versus-morality conflicts. First, an organization’s size—over 150 role-occupants—means that roleoccupants cannot all be known to each other personally. (150 is ‘Dunbar’s number’: the maximum number of humans with which any human can have meaningful contact.) Due to simple human cognitive limitations, many role-occupants will therefore tend to interact with one another in ways shaped by each other’s role. Many role-occupants in the organization will not appear to one another as unique individuals worthy of particularized engagement, but rather as faceless and nameless occupants of roles or role-types. For an example familiar to those in university settings, we can think of the specific role ‘lecturer in philosophy’ or the general role ‘academic.’ Both of these roles mediate one’s relationship with many of one’s non-immediate colleagues. Such mediation often determines the demands, expectations, and hopes that we place upon one another. The demands or expectations that come from others

204



within our organization are unlikely to be tailored to our specific circumstances, but rather derive from our role. Second, organizations’ division of labour further entrenches the salience of roles, when it comes to an individual’s own reasoning about what she should do in the organizational context. The division of labour (combined with organizations’ size) implies that role-occupants don’t know all the details of what all other role-occupants are doing vis-à-vis the organization’s goals and plans. (This differs from not knowing them personally, which was the point made immediately above.) This epistemic blind spot makes it difficult—if not impossible—to work out what one should do: if I don’t know what others in my organization are doing, then I can’t take it upon myself to confidently work out exactly what I should do in response to what they are doing. In the face of limited information about others’ precise tasks, it may be perfectly reasonable for individuals to trust ‘the system’—at least if they have some independent reason to believe that the system is overall abiding by moral principles. Trusting the system means performing one’s role while trusting that others are performing complementary roles and trusting that the roles together abide by morality’s demands. Such trust can be reasonable even when morality’s demands are in fact being violated (either by other role-occupants, or by the system as a whole). Third, organizations have rules and command relations. That is, not only are people structured into a role-based system, and not only are there are a lot of those people, and not only do those people not know what one another is doing— additionally, all of that happens in the presence of hierarchy. I mentioned this when characterizing an organization’s structure, when I noted that the relations between roles will often be ‘reporting and delegation’ lines. Hierarchies—that is, rules and command relations—have the function of turning expectations or hopes into instructions or demands. From the perspective of any given role-bearer, others within the organization have the role of commanding, instructing, admonishing, reprimanding, and sanctioning that role-bearer (except, perhaps, the person at the top—though even here, there are usually shareholders or other top-dogs that can remove an errant director). The result is that the ‘role’ side of role-versusmorality conflicts are explicit demands or instructions from particular other rolebearers, where those demands carry sanctions. This increases the psychological force of role-based demands. These special features of organizational roles are not confined to the commercial world, or even to organizations in which role-bearers are formally employed for pay. As Section 1’s examples demonstrated, organizational roles (in my sense of that term) cross-cut Michael Hardimon’s (1994, 353) distinction between roles in civil society (which he views as essentially voluntary) and roles in the family and state (which he views as essentially non-voluntary). As I have characterized organizations, some organizational role-occupancy is voluntary, while some is non-voluntary; of those that are voluntary, some are formally contracted,

     

205

while others are not. As should now be clear, states are organizations, on my characterization (and democratic states include ‘citizens’ as role-bearers). Likewise, many entities in ‘civil society’ are organizations—such as clubs, churches, associations, committees, charities, and so on. Organizations are ubiquitous—and so, therefore, are seeming conflicts between organizational-role obligations and moral obligations. I now turn to consider those conflicts

3. Role-Versus-Morality Conflicts In the previous section, I noted that it will often be reasonable for organizational role-occupants to perform their role as instructed. The reasonableness of doing this is generated by two broad features of organizational roles: (1) the difficulty (sometimes, impossibility) of knowing exactly what others in the organization are doing (produced by organizations’ large size and division of labour), and (2) the reasonableness of believing that the organization overall functions in ways that are morally benign. Of course, features (1) and (2) do not always hold. When they do not, role-occupants may face a conflict between the role’s requirements and independent moral norms. How might such conflicts arise? First, consider a case in which feature (1) fails to hold: a case in which, despite an organization’s size and divided labour, a role-occupant comes to know the details of how others are performing their roles—and, specifically, comes to know that these other role-occupants are not acting in accordance with the organization’s overall benign purpose. For a concrete example, consider the revelation in 2018 that Australia’s largest banks had engaged in widespread and pervasive dishonest practices. These practices included charging fees when no service had been provided (including to deceased customers), lying to customers, forging customers’ signatures, impersonating customers, falsely witnessing documents, transferring customers’ funds to advisors’ personal bank accounts, underpaying interest on term deposits, and other such problematic behaviours (Royal Commission 2018; Royal Commission 2019). These practices were embedded in problematic organizational cultures—yet they nonetheless required actions on the part of particular role-occupants. If one role-occupant knows that another is engaged in such practices, then the moral question is raised of whether the first role-occupant can, or should, continue in their role as normal. This is not just a question of performing actions outside the organization, such as publicly ‘blowing the whistle.’ It’s also a question of acting within the organization, for example using their role to act within the organization to challenge the problematic conduct. Both kinds of reaction are potentially demandable by morality, yet both potentially conflict with demands (explicit or implicit) that issue from the organization’s rules and command relations. Thus, we have a (seeming) role-versus-morality conflict.

206



A second way in which role-versus-morality conflicts might arise is via a failure of feature (2): a case in which it’s not reasonable for a role-occupant to believe that the organization overall functions in ways that are morally benign. Unlike a failure of feature (1)—which was a failure of other role-occupants—a failure of feature (2) is failure at the holistic level of the organization itself. Such failures are not always identifiable with or reducible to failures at the individual level (French 1984; Copp 2007; Pettit 2007). Again, the Australian banking scandal demonstrates this possibility. The Australian government tasked a Royal Commission with investigating potential misconduct in the banking sector and providing recommendations for how the government should respond. In its report, the Royal Commission attacked the ‘culture’ of ‘dishonesty and greed’ in Australia’s largest banking and finance corporations (Royal Commission 2018, 73; 2019, 138). The Royal Commission resisted the idea that this misconduct was a matter of ‘a few bad apples,’ instead insisting that the banks themselves (as distinct from individual role-occupants) receive some share of the responsibility (2018, 87–88). It’s easy to imagine how an individual role-occupant might come to be aware of such a culture—and, thus, come to have a reasonable basis to doubt the organization’s moral credentials—even without any knowledge of which precise behaviours by which precise other role-occupants contributed to this culture. To see this, consider the example I gave in Section 1: a financial advisor might face the demand to satisfy criteria for probation, where this requires convincing their advisees to adopt high-risk investments (which are in the long-run financial interests of the bank), which in turn requires giving advice that is not in advisees’ best interests. Such demands arise out of an organization’s rules and command relations, including internal norms around profit maximization and the instrumental treatment of customers. Yet due to the organization’s large size and division of labour, such demands may not come from any identifiable role-bearer. Indeed, it’s possible that such problematic organization-level demands result from a combination of individual-level decisions that are themselves entirely benign (Pettit 2007; List and Pettit 2011; Hess 2014). Failures of features (1) and (2) are not the only mechanisms by which the demands of roles can come to conflict with the demands of morality, though they are illustrative of the general idea. In general, a role-versus-morality conflict is any tension between, first, the demands imposed on a role-occupant in virtue of her role, where these include informal internal norms, and, second, the demands that would ideally be imposed on the role-occupant, were the whole system concordant with morality’s demands, or the demands the role-occupant would abide by, were she being morally responsive to the system’s moral failings. (In the final section, I will address cases where morality mandates the abolition of the organization.) It’s important to see that role-versus-morality conflicts do not have an easy resolution. As emphasized in Section 2, it is often perfectly reasonable for a roleoccupant to go along with the demands imposed on her. Those demands cannot

     

207

simply be thrown out the window as soon as a failure of features (1) or (2) presents itself. After all, there are good moral reasons why we normally toe the organizational line: we are cognitively limited individuals who usually cannot know what others in our organization are doing, and we often have good reason to believe our organizations are acting for the good. These facts create morally powerful presumptions of obedience, which it would often be hubristic to override.³ As Cristina Bicchieri and Ryan Muldoon put it when discussing social (rather than organizational) norms: Most of the time, we are not aware of our expectation [that others expect us to comply with a norm], and compliance may look like a habit, thoughtless and automatic, or it may be driven by feelings of anxiety at the thought of what would happen if one transgresses the norm. Upholding a norm is not a matter of conscious cost/benefit calculations; rather, people tend to repeat patterns of behavior that they have learned and, on average, work well in a variety of situations. (Bicchieri and Muldoon 2014)

Thoughtless and automatic habits are difficult to override, and it’s not obvious that we have obligations to do so if they have served us well in the past. Moreover, even if one could override the presumption that favours organizational obedience, it’s far from obvious that an individual should ‘do the right thing’ in circumstances where others (whether individual or organizational) are doing the wrong thing. Such high-mindedness often makes one a target of resentment, discipline, and accusations of virtue signalling—exactly the opposite of what will change the problematic conduct or culture within the organization. We are often justified in trusting the system, and the legitimacy of this trust produces presumptively decisive reasons for abidance across a variety of different situations, even in cases where wrongful behaviour is being asked of us (as long as wrongful behaviour is not being asked of us too often). Role-versus-morality conflicts produce a normative (as opposed to merely psychological) tension.

4. Dissolving the Conflict: Fundamental, Imposed, and Derivative Demands I propose we begin resolving the tension by asking not about any individual roleoccupant, but rather about the organization as a whole. Specifically, we begin by ³ Here there are analogies to arguments that there are good reasons to obey the law, even in cases where the law is wrong (e.g. Raz 1986). This analogy becomes even more apt once one sees that citizenship is a role, the demands of which are largely obligations to obey laws. I lack space here to fully defend an analogy between the obligation to obey the state’s laws and the obligation to obey an organization’s demands; my suggestion is just that the justification for these obligations take roughly the same shape.

208



asking: what is the purpose of this organization? The answer to this question is not determined by the goals that are explicitly endorsed by leaders, executives, directors, managers, or shareholders. Instead, the answer is determined by looking at what’s essential to this organization: what purpose is such that the organization currently has that purpose and the organization would change its personal identity if it lost that purpose? That is the organization’s fundamental purpose. To see how this works, consider a legislature. Suppose the fundamental purpose of that legislature is to write laws that reflect the general will of constituents. This is just to say that: if this particular legislature stopped doing this, then there would be a rupture in the identity of the organization; we’d now be dealing with a new and different entity. Based on this fundamental purpose of the legislature, we can deduce the fundamental purpose of each individual role-occupant in that legislature: very roughly, the fundamental purpose of each individual legislator is to contribute to writing laws, in such a way that those laws reflect the interests or will of that legislator’s particular constituents (supposing a system with district-based representation). In this way, the fundamental purpose of the organization (as determined by its structure) determines the fundamental demands on each individual role-occupant.⁴ But a role’s fundamental demands are not directly action-guiding. The actionguiding demands of any role are more specific than the fundamental demands. The action-guiding demands have two faces. The first face is composed of the demands that an individual confronts as a result of the instructions, commands, hopes, expectations, and so on that others place on her in virtue of her role. These ‘others’ include role-occupants within the organization, the organization itself, and even outsiders who interact with the role-occupant qua role-occupant. Call these the imposed demands, since they are imposed by others on a role-occupant, in virtue of her role. In the case of a legislator, one example of an imposed demand is the demand to obey the party’s whip in a particular context. The imposed demands constitute the ‘role’ side of role-versus-morality conflicts. Yet the imposed demands are only one face of a role’s action-guidingness. The second face is composed of the demands that an individual faces in virtue of how her role’s fundamental purpose interacts with the context. Call these the derived demands, since they are derived from the role’s fundamental purpose (in interaction with a particular context). The derived demands might not be imposed on the role-occupant by anyone, not even by herself. No one may even have considered what the derived demands might be. The derived demands are, in this way, ⁴ One might doubt that any legislature has such a democratic fundamental purpose: after all, isn’t the UK Parliament now identical to the English Parliament before it was democratic? This is a question I cannot resolve here; I use the example of a democratic fundamental purpose simply to get across the idea. Plausibly, the precise fundamental purpose of each organization will be a highly contestable matter.

     

209

demands that can depart from any (explicit or implicit) demands that anyone imposes on the role-occupant. In the case of a legislator, one example of a derived demand is the demand to disobey the party’s whip in a context where the legislator knows the whip is enforcing a view that is both wrong and contrary to the party’s manifesto. The derived demands are action-guiding applications of the fundamental demands in a context. In light of these distinctions, my proposal for resolving the role-versus-morality tension is simple. When the imposed demands depart from the derived demands (which is to say, when the imposed demands are poor applications of the fundamental demands), then organizational role-occupants have obligations— grounded in the fundamental demands of the role—to use the fundamental demands to challenge the imposed demands. These obligations are themselves derived demands: they are demands that derive from the fundamental demands, in a context in which the role-occupant faces imposed demands that go against the proper application of the fundamental demands. Specifically, they are demands to act because of the fundamental demands, while challenging the imposed demands. By looking to the fundamental demands, we can reconcile the action-guiding demands with morality. We simply need find the derived action-guiding demands, rather than focusing only on the imposed action-guiding demands. To see how this plays out, consider the examples from Section 1. The easiest of these is the democratic citizen, whose imposed demands require that she pay taxes that fund a clearly unjust war. Meanwhile, her fundamental demand is to do what she can to contribute to the just governance of her state—by voting, petitioning, protesting, and so on. What about her derived demands? If we were in a context where her state was pursuing justice on all fronts, then her derived demand might simply be to pay her taxes and otherwise obey the law. However, given a context in which her state is pursuing a clearly unjust war, her derived demand is to protest, petition, and so on against the war. This is her fulfilling her (fundamental) role obligation; it isn’t her resisting her (fundamental) role obligation. Her derived role obligation is to use her fundamental role obligation to challenge her imposed role obligation. Now, she may not be required (by her role’s fundamental demands) to withhold taxation: exactly what ‘challenge’ she must make to the imposed demand will vary, depending on the values at stake. A trickier example is the chair of a condominium owner assembly. This person has the (fundamental) role of organizing, chairing, and contributing to meetings of all owners of apartments within a particular building, so that the owners can together deliberate about, and decide on, various matters of building governance. Let’s suppose the assembly is debating whether to dispose of common rubbish in a cheap yet environmentally unsound way, or in a more expensive yet environmentally sound way. Most owners are in favour of the cheap option. It’s the chair’s turn to speak in the debate. What should she say? This example differs from the citizen example, insofar as the chair’s fundamental role makes no mention of

210



moral values, such as justice or environmentalism: her fundamental role does not (it seems) require her to contribute to the building being run in an environmentally sustainable way. We might think her fundamental role demand is to contribute to meetings in a way that is likely to lead to consensus amongst owners, or in a way that is likely to make most owners satisfied. These fundamental role demands don’t seem to produce morality-respecting derived demands. However, crucially, when assessing the fundamental purpose of an organization, we will often need to look at the human purpose for which the organization was set up.⁵ The organization might have been set up to govern a territory, as in the case of states, or to provide satisfactory financial products and services, as in the case of banks, or to responsibly govern an apartment building, in the case of condominium owner assemblies. Such goals are often an important part of an organization’s fundamental purpose. Once such goals are included in the organization’s fundamental purpose, the chair’s derived role obligation will largely be determined by what it means for the assembly to responsibly govern an apartment building—where this produces the chair’s fundamental and derived role obligations. In a particular context, the assembly’s fundamental purpose may well give the chair the derived obligation to speak up in favour of the environmental option for rubbish disposal, on pain of frustrating the assembly’s fundamental purpose of governing responsibly. Another tricky example is a financial advisor in a bank. The advisor’s fundamental obligation is to provide sound financial advice to customers. This fundamental obligation is produced by the organization’s fundamental purpose (providing satisfactory financial products and services). The advisor’s imposed demand is to convince customers to make high-risk investments. In an ideal world, the advisor’s derived obligation would be to give advice that will satisfy the financial goals of the particular customer. This derived obligation is ultimately grounded in the organization’s fundamental purpose of providing satisfactory financial products and services. However, in a context with morally problematic imposed demands, the advisor’s derived demand is to use her fundamental role to challenge the imposed demands. This example differs from the citizenship and condominium examples, insofar as the advisor might have few avenues for using her fundamental demand to challenge the imposed demand. Yet even here, there are informal avenues for dissent. These include raising the issue with superiors or equals in the structure, or (in more extreme cases) publicly blowing the whistle on the incentive structures that

⁵ Often, but not always: sometimes, an organization’s fundamental purpose cannot be ascertained by asking what it was ‘set up’ to do, for example since the organization arose in an unintended way out of earlier organizations. It is difficult to say much in general about what determines an organization’s personal identity and, therefore, its fundamental purpose. This difficulty is one attraction of tying an organization’s fundamental purpose to its personal identity: organizations’ fundamental purposes are plausibly as varied, and varied for the same reasons and in the same ways, as the determinants of their personal identities. On the latter, see Rust 2019.

     

211

reward giving unhelpful advice. Exactly which actions are required will vary with the values at stake, just as in the other examples. One objection presents itself immediately. What if the fundamental purpose of an organization is morally dubious, and therefore produces morally dubious fundamental and derived role demands? Consider venture capitalist organizations. One might be tempted to say that the fundamental purpose of these organizations just is maximizing profit: this is what they aim to do and what they were set up for. They don’t have more substantive or morally respectable goals (such as ‘provide satisfactory financial products and services’). In these cases, the fundamental and derived role obligations will trace back to the organization’s fundamental purpose of maximizing profit. In some contexts, that purpose will produce a derived demand to act corruptly. And that derived demand can be defeated (one might think) only by a morality that is wholly ‘external’ to the organization. The role-versus-morality tension is not dissolved. An initial reply runs as follows. The fundamental purpose ‘making profit’ won’t produce the dilemma as outlined in Section 3. That dilemma relied on role-bearers having good reasons to obey their imposed obligations, as a matter of a general presumption or habit. Only when a role-bearer has good reasons for a general presumption in favour of the imposed obligations does the role-bearer face a genuinely normative role-versus-morality dilemma. And perhaps purely profitdriven organizations simply cannot produce good reasons for a general presumption in favour of abiding by imposed obligations. The initial reply is too quick. The moral legitimacy of the purpose ‘make profit’ was famously endorsed by Milton Friedman (1970) from within a utilitarian perspective. There is reasonable disagreement over the cogency of Friedman’s argument. Thus, it is not unreasonable for at least some role-bearers, in at least some profit-driven organizations, to presumptively abide by their imposed obligations. And such a presumption might further be justified via other moral theories. For example, by adapting the Aristotelian notion of eudaimonia, one might think that the truly flourishing venture capitalist corporation just is the one with the highest profits. If profit maximization is the fundamental purpose of a venture capitalist organization, and if that purpose might be (reasonably thought to be) morally benign, then a role-bearer might reasonably hold a general presumption in favour of performing their imposed obligations. But that benign organizational purpose might, in certain contexts, nonetheless produce role obligations (both fundamental and derived) that are clearly immoral—because maximizing profits will sometimes require role-bearers to act in ways that are clearly immoral. The worry, then, is that some role-bearers will face a genuinely normative roleversus-morality tension—but their organization’s fundamental purpose will not contain the resources to support the ‘morality’ side of that tension. Instead, the fundamental demand of any given role-occupant will be ‘do what you can to most

212



efficiently contribute to the organization’s making profit.’ Meanwhile, the derived demand of any given role-occupant will require the role-occupant to do whatever profit maximization demands from them in a particular context. In such organizations, the objector might continue, the imposed demands will often perfectly track the derived demands—where the latter are properly understood as demanding whatever profit maximization demands in a context. This is because the imposed demands—as at least when they come from directors, managers, and shareholders— will be in the service of profit. But then (the objector concludes) the idea of an organization’s ‘fundamental purpose’ won’t have captured, or responded to, the role-versus-morality tension. Instead, we will have vindicated the derived role obligations of lying to customers, forging their signatures, twisting their arms, and so on. In response to this line of reasoning, let’s grant for the sake of argument that some organizations have ‘maximizing profit’ as their fundamental purpose (perhaps venture capitalist organizations are the only ones of which this is true).⁶ For these organizations, we can—and should—demarcate the conditions under which that fundamental purpose can truly produce the role-versus-morality dilemma.⁷ Even for someone like Friedman, the legitimacy of the profit motive does not extend to corruption, lying, forgery, theft, and so on. The organization’s profit motive produces a reasonable presumption (held by role-bearers) in favour of obedience with imposed obligations, only if the organization’s profit motive is embedded within a broader system that prohibits the violation of certain basic rights. Because this broader system is a necessary condition for the normativity of the organization’s profit-focused fundamental purpose, we can see the rolebearer’s fundamental obligations as including the requirements of that broader system. Thus, the role-bearer’s fundamental obligations include a prohibition on corruption, lying, forgery, theft, and so on, if those actions are prohibited by the system that gives the organization’s fundamental purpose its normativityproducing legitimacy. Importantly, these prohibitions are still genuinely role obligations, because they operate on role-bearers only via the organization’s fundamental purpose. The constraints apply to the role-bearer insofar as she operates within an organization with a certain fundamental purpose, where that purpose is legitimate only within certain constraints. There is another, compatible, method of responding to this objection. This method of response starts by noticing that maximizing profit has rationality as a necessary condition. You can’t aim to maximize profit if you’re not (to some extent) rational. By ‘rational,’ I simply mean that an entity is capable of taking effective means to its ends, capable of ensuring that its beliefs and desires are more-or-less mutually consistent, and that it by-and-large aims to exercise those capabilities. ⁶ I thank Alex Barber for this example. ⁷ I thank Alex Barber and Sean Cordell for help with formulating this point.

     

213

Once organizations’ rationality is on the table, we can notice that organizations (even for-profit corporations) have the ability to use their rationality in a way that respects persons. More generally, corporations have the capacity to respect moral side-constraints on their pursuit of profit. This capacity can be gleaned from the fact that corporations at the very least pay lip service to the ideals of human rights, stakeholders’ interests, law-abidance, and so on. Of course, corporations might pay such lip service simply out of self-interest—as a way of, ultimately, pursuing profit. But this doesn’t matter: the fact that corporations can acknowledge such considerations demonstrates that they are capable of attending to them. This capacity produces side-constraints on the pursuit of profit, even if those constraints are not mandated by the social system in which the organization is embedded. Paradigmatically, consider the Kantian account of moral agency. As Kendy Hess (2018a) has argued, corporations satisfy three different Kantian conceptions of moral agency: they can ‘act on universalizable principles and treat humanity as an end in itself,’ they can ‘give such principles to themselves, treat their own “humanity” as an end itself, and act out of respect for the law,’ and they can ‘draw on empathically generated information and insights to inflect their performance’ (2018a, 67; see similarly Wringe 2014). The fact that organizations satisfy the conditions of moral agency produces a requirement to respect moral side-constraints—even if the organization’s purpose just is the completely unbridled pursuit of profit, and even if it is somehow reasonable for role-bearers to presumptively abide by their imposed obligations. Thus, even if an organization has the fundamental purpose of making profit, it also faces the fundamental demand to respect moral side-constraints (where that demand is grounded in its capacity to do so). This demand comes from morality, if not from the social system in which the organization acts. The organization’s fundamental demand to respect side-constraints grounds a related fundamental demand on role-occupants: specifically, role-occupants bear fundamental role obligations to perform their roles in ways that abide by side-constraints. This fundamental demand of the role-occupant (which is derived from the fundamental demand of the organization) then produces a derived demand to challenge certain imposed demands: that is, the fundamental demand on the role-occupant produces a (fundamental) role obligation to alter (imposed) role obligations, in any organization that is capable of attending to moral considerations. Crucially, I am suggesting that all of these demands on role-occupants are genuinely role obligations—they are obligations internal to the organization. This is because the moral demand to abide by hard constraints applies fundamentally to the organization itself, and only derivatively to the role-bearers. There is no role-versusmorality conflict, since the role itself requires morality (in all organizations that are moral agents). Now, arguably, not all organizations are moral agents: not all organizations are capable of attending to moral considerations. Some organizations might be

214



‘amoral,’ that is, instrumentally rational but incapable of any kind of moral reasoning (Hindriks 2018). In these organizations, there is no way of using the organization’s fundamental purpose or fundamental obligations to resolve roleversus-morality conflicts. This is because these organizations’ fundamental purposes (and the role-occupants’ fundamental demands, which derive from that purpose) may produce derived demands that conflict with morality, while lacking the resources to produce any derived demands that accord with morality. Perhaps some criminal organizations are like this. In such organizations, my proposal for resolving role-versus-morality conflicts will not work. But this doesn’t mean that role-occupants should simply obey such organizations’ edicts. Instead, if an organization is not a moral agent—and therefore is genuinely incapable of producing (fundamental and derived) role demands that accord with morality—then we should doubt whether roleoccupants have any reason to perform their role at all. Of course, the organization’s goals produce a kind of normativity that’s internal to the organization’s perspective. But that normativity doesn’t produce genuine obligations for members. In these cases, then, the role-versus-morality conflict should be resolved in favour of morality; the role should not be obeyed. This is different from the cases I focused on above, in which the organization has a morally benign or morally conditioned fundamental purpose (or faces the fundamental demand to be moral), such that the fundamental role obligation accords with morality, such that the role-versus-morality conflict dissolves altogether.

5. Conclusion I have suggested that we sometimes have ‘role obligations to alter role obligations’: obligations to use our role’s fundamental demands as a justification for acting in accordance with derived demands, where the derived demands require us to push back against imposed demands. My proposal can be seen as a variation of Michael O. Hardimon’s claim that ‘[t]here are different ways of being a flight attendant, oncologist, or police officer. Part of what it is to become a good flight attendant, oncologist, or police officer is to find a way of carrying out the responsibilities of these roles which suits one’s particularities’ (1994, 355). I would contest only the last part of Hardimon’s statement. My suggestion has been that the good flight attendant, oncologist, or police officer finds ways of carrying out the responsibilities that accord with that role’s place in its organization’s pursuit of its fundamental purpose—where that purpose will almost inevitably be infused with moral requirements. To be sure, the derived demands that I have proposed—role obligations to alter role obligations—are not the only way to push back against morally problematic imposed demands. A role-occupant might also (or instead) attempt to act upon

     

215

her organization ‘from the outside,’ to induce it to change its ways. Such actions are not properly construed as role performances, so any obligations to perform such actions are not role obligations. I have focused here on the ways in which our organizationally embedded roles themselves can contain the normative resources for pushing back in those cases where our organizations demand the morally dubious. By attending to the fundamental purpose of our organizations—which include the conditions for its moral legitimacy and the fact of its moral agency— we can infer the fundamental demand of our role. This fundamental demand often (though not always) contains intra-organizational normative resources for challenging imposed demands that conflict with moral requirements.

References Bicchieri, Christina and Ryan Muldoon. 2014. ‘Social Norms.’ Stanford Encyclopedia of Philosophy. Collins, Stephanie. 2019. Group Duties: Their Existence and Their Implications for Individuals. Oxford: Oxford University Press. Collins, Stephanie. 2023. Organizations as Wrongdoers: From Ontology to Morality. Oxford: Oxford University Press. Copp, David. 2007. ‘The Collective Moral Autonomy Thesis.’ Journal of Social Philosophy 38 (3): 369–388. French, Peter. 1984. Collective and Corporate Responsibility. New York: Columbia University Press. Friedman, Milton. 1970. ‘The Moral Responsibility of Business is to Increase Its Profits.’ New York Times Magazine. Hardimon, Michael O. 1994. ‘Role Obligations.’ Journal of Philosophy 91 (7): 333–363. Harré, Rom. 1979. Being Social. Oxford: Blackwell. Herzog, Lisa. 2018. Reclaiming the System: Moral Responsibility, Divided Labour, and the Role of Organisations in Society. Oxford: Oxford University Press. Hess, Kendy. 2014. ‘The Free Will of Corporations (and Other Collectives).’ Philosophical Studies 168 (1): 241–260. Hess, Kendy M. 2018a. ‘Does the Machine Need a Ghost? Corporate Agents as Nonconscious Kantian Moral Agents.’ Journal of the American Philosophical Association 4 (1): 67–86. Hess, Kendy M. 2018b. ‘The Peculiar Unity of Corporate Agents.’ In Kendy M. Hess, Violetta Igneski, and Tracy Isaacs (eds.), Collectivity: Ontology, Ethics, and Social Justice. Washington, DC: Rowman and Littlefield, pp. 35–60. Hindriks, Frank. 2018. ‘Collective Agency: Moral and Amoral.’ Dialectica 72 (1): 2–23. Hodgson, Geoffrey M. 2007. ‘Institutions and Individuals: Interaction and Evolution.’ Organisation Studies 28: 97–116.

216



List, Christian and Philip Pettit. 2011. Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press. Pettit, Philip. 2007. ‘Responsibility Incorporated.’ Ethics 117: 141–201. Raz, Joseph. 1986. The Morality of Freedom. Oxford: Oxford University Press. Ritchie, Katherine. 2013. ‘What Are Groups?’ Philosophical Studies 166 (2): 157–172. Rovane, Carol. 1998. Bounds of Agency: An Essay in Revisionary Metaphysics. Princeton: Princeton University Press. Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry. 2018. Interim Report. Volume 1. Available at . Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry. 2019. Final Report. Volume 1. Available at . Rust, Joshua. 2019. ‘Institutional Identity.’ Journal of Social Ontology 5 (1): 13–34. Thomasson, Amie. 2019. ‘The Ontology of Social Groups.’ Synthese 196: 4829–4845. Tollefsen, Deborah. 2015. Groups as Agents. Cambridge: Polity. Tuomela, Raimo. 2013. Social Ontology: Collective Intentionality and Group Agents. Oxford: Oxford University Press. Turner, Jonathan. 1997. The Institutional Order. London: Longman. Weber, Max. 1968. Economy and Society: An Outline of Interpretive Sociology. Ed. Guenther Roth. New York: Bedminster Press. Wringe, Bill. 2014. ‘May I Treat a Collective as a Mere Means?’ American Philosophical Quarterly 51 (3): 273–284.

11 Are Obligations of Friendship Role Obligations? Diane Jeske

1. Introduction A good portion of our self-conceptions involves understanding ourselves as occupants of what are often thought of as social roles. Thus, I think of myself as a daughter, as a sister, as a professional philosopher, as a professor, as a teacher, as an American. But I also think of myself under various descriptions which are such that I do not think of my satisfying them as my occupying a role. For example, I am an animal lover, in particular I am a cat lover, I am a reader of novels, I am someone who does regular cardio exercise, I am someone who enjoys horror films, etc. One way in which my occupation of a social role is often thought of as differing from those latter descriptions of me is that I have various obligations—so-called ‘role obligations’—in virtue of my occupation of the role. As a daughter I have an obligation to care for my mother as she ages, as a professional philosopher I have obligations to referee articles for journals, as a teacher I have obligations to promote my students’ intellectual development, as an American citizen I have obligations to vote and to pay taxes in the US. There are many questions we can ask about role obligations, such as: are they genuine obligations, that is, do they constitute or ground justificatory reasons for action? If they are genuine obligations, how do they differ, if at all, from obligations that are not tied to a particular role? Are they derived from non-role obligations or are they independent of the latter? My interest in this chapter, however, is with the question as to whether certain obligations, namely those of friendship, ought to be understood as role obligations. In addition to being a sister, a philosopher, a novel reader, and a cat lover, I am also a friend. Does the fact that I am a friend mean that I am an occupant of the social role of friend, or is ‘being a friend’ a description that applies to me but does not involve the occupation of a social role? One thing, I think, is quite clear: in virtue of being a friend to Tracy, I have obligations to her. But that is not sufficient for concluding that ‘friend’ is a social role: after all, I have an obligation to save the person drowning in the river in front of me, but ‘person on the riverbank’ hardly seems to count as a social role. Does it matter to our understanding of the Diane Jeske, Are Obligations of Friendship Role Obligations? In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Diane Jeske 2023. DOI: 10.1093/oso/9780192843562.003.0011

218



obligations of friendship whether we conceive of friend as a social role, and, thus, of the obligations of friendship as role obligations? Answering these questions involves determining what a role is, what a friend is, and how to understand role obligations. There is surprisingly little attention in the philosophical literature on particular roles and their attendant obligations apart from discussions of professional roles such as physician and lawyer (see e.g. Bowie 1985 and Luban 1983). So, figuring out how to understand roles and their attendant obligations is not a straightforward task. There is a larger literature on friendship and our special obligations to our friends (see Jeske 2022), and, while there is a lot of common ground in this literature with respect to how to understand the nature of friendship, there is far less agreement on how to understand our obligations of friendship. So, to answer the question as to whether obligations of friendship are role obligations, we need to pull apart the concepts involved—friendship, obligation, role, and role obligation—and then try to put the pieces together. In section 2, I will try to tease apart some different conceptions of how to understand a social role, concluding that it is not a univocal concept and that the best we can do when asking whether to be a friend is to occupy a social role is to ask certain other questions in order to assess the ways in which being a friend is like certain supposedly paradigmatic social roles. After a discussion of obligation, reasons, and role obligations in section 3, I will then present my own account of friendship and its attendant obligations in section 4. In the final substantive section of the chapter, section 5, I will use that account of friendship to compare and contrast it with paradigmatic cases of voluntarily accepted social roles, in particular the role of spouse or domestic partner, and with paradigmatic cases of non-voluntary social roles, in particular familial roles such as mother and daughter. I will argue that the ways in which friendship differs from these paradigms are more significant than the ways in which it is similar to them, and, thus, however we understand a social role, we ought not to understand being a friend as a social role nor ought we to understand the obligations of friendship as role obligations. There may be social expectations and norms concerning friendship, but friendship itself is neither defined nor constrained by those expectations and norms.¹ In fact, to conceive of it as a social role would be to diminish the ways in which it adds value to our lives and binds us to certain other people. The significance of friendship is in part a matter of how it is constructed by the parties involved and provides a space for understanding oneself in relation to another independent of everyone else and, very importantly, apart from societal expectations and norms. ¹ I should say that friendship is not necessarily constrained by these expectations and norms. Friends are likely to bring societal expectations and norms into their relationships, and so may use those expectations and norms to understand and regulate their interactions with one another.

      ? 219

2. What is a Social Role? As I said in the introduction, we need a way to distinguish between those descriptions of a person which count as descriptions of the individual as an occupant of a social role and those descriptions that do not. Michael O. Hardimon says that he ‘shall follow the common practice of using the term “role” to refer to constellations of institutionally specified rights and duties organized around an institutionally specified social function’ (Hardimon 1994, 334). Similarly, Tim Dare defines a role as ‘a position in a social network constituted by a distinctive set of normative statuses . . . that attach to a role-occupant by virtue of her occupation of that role’ (Dare 2016, 704; see also Andre 1991, 73).² These explanations of what counts as a role depend upon our having an understanding of what an ‘institution’ or a ‘social network’ is. Hardimon’s limiting his consideration of roles to institutional roles renders his account too narrow for my purposes in this chapter. Hardimon claims that institutions involve rules defining offices and positions within the institution, such as CEO and professor, which can be occupied by different people at different times: institutions, he says, are ‘on-going, self-reproducing structures’ (Hardimon 1994, 335). Hardimon denies that friendships (or friendship as such) are institutions in this sense, because a particular friendship essentially involves the two people who are friends. While Elaine may be CEO of CatCo now, it may also be that Jerry was the previous CEO of CatCo, and that George will be CEO after Elaine. But while I can have friends other than Tracy so that many people can satisfy the description ‘Diane’s friend,’ none of those people can occupy Tracy’s role in my friendship with Tracy. Each friendship is unique and is determined by the particular parties to it, while CatCo remains the same corporation throughout its changes in CEOs.³ But immediately after offering his characterization of an institution, Hardimon complicates matters by admitting that families are not clearly institutions— individual families, it would seem, are constituted by their members and their relationships to one another.⁴ Yet he wants to view familial roles as institutional roles because families can be understood as ‘institutionally-defined groups’ (Hardimon 1994, 336): familial roles are defined by institutions even if the family

² These definitions are obviously circular, and thus problematic, as I will discuss later in this section. ³ Consider, however, our concept of a ‘best friend.’ At any given time, only one person can occupy the role of ‘Diane’s best friend,’ just as only one person at any given time can be CEO of CatCo. But different people can occupy the role of my best friend at different times, again, just as CEOs can succeed each other at CatCo. This, I believe, complicates Hardimon’s claims (see n. 4.). ⁴ I am simply following Hardimon on this point. But are families constituted by their members? If my spouse and I have a child, have we created a new family or added a new member to an already existent family? If a single mother of an infant weds after the biological father’s death, does her new husband join her family and take on the role that the biological father previously filled, or is a new family created? I am not sure how to answer any of these questions, so I will follow Hardimon on this point.

220



itself is not an institution. He doesn’t clarify what he means by saying that families are ‘institutionally-defined.’ Does he mean that there is some larger institution that involves rules that govern families, so that families play the part in that larger institution that CEOs play in CatCo? If so, what is that institution? It is true that legal norms govern certain familial relationships such as those between spouses/ domestic partners and those between parents and dependent children, but such norms do not govern all such relationships: legal norms do not seem to govern sibling relationships, for example.⁵ And even with respect to those that the law governs, the legal norms do not extend to all of the obligations that we take to be attendant upon the familial roles involved; for example, a parent could meet all of her legal obligations to her dependent child without doing all that we think that parents are obligated to do for their children. So, it seems that Hardimon does not clearly differentiate friendships (which he says are not institutions but, rather, personal relationships) from families, given that neither friendships nor families have roles that can be occupied by different people over time. So, if families involve institutional roles, why claim that friendships do not?⁶ I don’t disagree with Hardimon’s claim that we should not understand ‘friend’ as a social role, but, as I will make clear later, I don’t think that his arguments for that denial are persuasive, and nor do they get at why it is important to deny that obligations of friendship are role obligations. Dare claims that, while some roles ‘are the product of deliberate institutional design guided by desired functional outcomes,’ others are instead the product of ‘social expectation in turn produced by widespread beliefs about the function’ of that role (Dare 2016, 705). So CEO of CatCo is a position intentionally created by some person or persons during their intentional creation of the corporation: they created the role of CEO to serve various functions that CatCo needs to have served as it aims to make a profit by marketing and selling products for humans and their cat companions. Friendship in general did not come into being in anything like the way that CatCo would have (if CatCo were a real corporation). But there are certainly social expectations regarding how friends are to treat each other and also about what friendships can do for us (i.e. about their ‘function’): they provide us with companionship, support, enjoyment, etc. However, the question is whether friendship is somehow ‘produced by’ those expectations, or whether, instead, the expectations arise from the nature of friendship. In the former case, the expectations would have an authority that they would not in the latter. If the ⁵ Or, at least, it governs them in small ways, such as when the law determines ‘next of kin’ relations in legal proceedings regarding issues such as inheritance and custody of dependent children. ⁶ Simmons also questions whether Hardimon can sustain the distinctions he draws between institutional roles, membership in a group, being in a personal relationship, and being involved in a social practice such as promising, to take Hardimon’s example (Simmons 1996, 28 n. 2). Further, Hardimon himself admits that to be a friend is to occupy a social role and that obligations of friendship are role obligations (Hardimon 1994, 336). So Hardimon’s focus on institutions is in the end unhelpful in attempting to get a broader understanding of social roles.

      ? 221 nature of friendship is determined by social expectations, then social expectations cannot be incorrect about what counts as a friendship: if people expect a friendship to provide comfort to the parties involved, then it is a trivial truth that friendships provide comfort to those party to them. But if expectations arise from reflection upon the nature of friendship, then there is plenty of room for those expectations to be misguided in all sorts of ways, particularly with regard to particular friendships. Both Dare and Hardimon claim that roles are defined by rights and duties. Hardimon understands roles, as we saw above, as ‘constellations of rights and duties’ and Dare says that roles are ‘constituted by a distinctive set of normative statuses’ (Dare 2016, 704). According to these definitions, roles are nothing over and above the relevant rights and responsibilities. As we will see in section 4, if roles were nothing but rights and responsibilities, then being a friend would certainly not be a role: friendship is a relationship between two people that grounds the duties that the two have toward one another. In other words, there is something that there is to be a friend that is conceptually independent of the duties that friends have to one another. However, it is not at all clear that either Hardimon or Dare really intends to reduce roles to clusters of rights and duties. After all, consider Hardimon’s claim that a role obligation is ‘a moral requirement, which attaches to an institutional role, whose content is fixed by the function of the role, and whose normative force flows from the role’ (1994, 334). Similarly, Dare claims that role obligations ‘attach to a role-occupant by virtue of her occupation of that role’ (Dare 2016, 704). These remarks suggest that there is a role to be occupied that is defined independently of the assumption of role obligations, and that one acquires the obligations as a result of occupying the conceptually prior role. It may be that both Dare and Hardimon intend to understand ‘normative statuses’ as including norms about what it is to occupy a role in addition to norms about what occupants of such a role ought to do (although, then, obligations do not ‘flow from’ the role, but are, if not entirely, still partly constitutive of the role). Otherwise, as definitions, Dare and Hardimon’s claims become circular: to say that one acquires role obligations in virtue of coming to occupy a role is really just to say that one acquires role obligations in virtue of acquiring role obligations. Thus, the occupation of the role (even if such occupation is understood in terms of normative status) is something distinct from the obligations resulting from it. I am not going to attempt to offer an account of what constitutes a social role. I think that the discussion in this section reveals that doing so is not a straightforward task, and that our concept of a role may very well not be a univocal notion. Thus, first offering an account of what it is to occupy a social role and then determining if ‘being a friend’ is a social role does not look to be a promising way of determining whether obligations of friendship are role obligations. So, I am going to adopt an alternative strategy. After examining the concepts of an

222



obligation and of friendship, I am going to compare and contrast friendship with examples of what are taken to be paradigms of roles and of their attendant obligations, namely being a daughter or a mother, and being a spouse or domestic partner. In comparing friendship to these two supposedly paradigm cases, we can get a better sense of what is at stake in understanding obligations of friendship as role obligations or not.

3. Obligations and Role Obligations I am not going to engage in a full conceptual analysis of the concept of obligation and then show how obligation, as a normative concept, differs from other normative concepts such as duty and right. My intentions here are much more modest. I am simply going to present a couple of different ways in which we use the notion of an obligation, so that we have some important distinctions available in trying to decide whether the obligations of friendship are role obligations. First, there seems to be one notion of an obligation where to say of a person that she has an obligation is to say of her that she has a kind of reason to behave in a particular sort of way. Consider a case in which I have promised Richard that I will bake a cake for his birthday. Most of us would say that, in virtue of my having made that promise to Richard, I now have an obligation to bake him a birthday cake and that that obligation either gives to me or constitutes for me a reason for me to bake that cake. Further, most would say of that reason that it is a moral reason.⁷ Whether or not we call it ‘moral,’ however, it clearly is a justifying reason, that is, it constitutes a consideration in support of my baking the birthday cake. All such reasons are prima facie: they provide a consideration in favor of acting, but not necessarily a decisive one. Thus, to say that I have a prima facie duty to do X is not to say that I ought to do X, because ought-judgments are best understood as judgments about our all-things-considered reasons (see Beran 1972, 207–208). Given that obligations are best understood as prima facie justifying reasons, they can be overridden or outweighed by competing considerations. Thus, the claim that I have an obligation to do X does not imply that I ought, all things considered, to do X or that X is the right thing for me to do. Justifying reasons need to be distinguished from motivating reasons for action.⁸ Motivating reasons are psychological states that move us to action. They do not

⁷ It is not always clear what a particular philosopher means when they say of a reason that it is a moral reason, and I think that the terminology of moral vs non-moral is often more misleading than helpful. So I am not going to say any more here about what it means to say of a reason that it is a moral reason. I just note that that is how most would classify it. ⁸ Francis Hutcheson called these latter ‘exciting reasons’ (1991 [1742], 308) because they excite us to action.

      ? 223 provide considerations in support of acting in a particular way.⁹ Questions about motivating reasons, are, thus, empirical questions about what caused a person to act in the way that she did. It is most likely the case that people are at least sometimes motivated by their understanding of themselves as occupying a role. Hardimon says that someone identifies with a role when she occupies the role, recognizes that she occupies the role, and conceives of herself ‘as someone for whom the norms of the role¹⁰ function as reasons’ (Hardimon 1994, 358; see Sciaraffa 2011 for further discussion of role identification). If I believe that the norms of my role justify action, and I want to act on such justification,¹¹ then identification will act as at least partly constitutive of my motive for acting in accordance with (my understanding of) the norms of my role. But of course recognition of the fact that the role that I occupy has certain regulating norms might motivate me because I want to remain in the role—for status, for a livelihood, for power and authority, or any number of other benefits—and I judge that I will be removed from the role if I do not act in such a way. People might also be motivated by recognition of their occupation of a role to subvert what they understand to be the traditional norms of the role, as when a rebellious teenager acts contrary to the ways in which a son is expected to respect his parents. Again, why people do what they do, that is, what their motivating reasons are, is an empirical matter that we cannot resolve via use of any of our purely philosophical tools. However, it is important to note that in acting in response to what they understand as the norms of their role, people often take as their guide the expectations of other persons with respect to the behavior of persons in those roles. In fact, when we talk about role obligations, one thing that we might be talking about is the expectations of other persons. For example, in trying to act as a professor ought to act, I will be guided, at least to a certain extent, by what I think others—my colleagues, other members of my profession, the administration of my university—understand as required of or forbidden to philosophy professors at research institutions. In fact, we might understand our talk about role obligations as reducible to, that is, as just another way of talking about, what people expect or demand of persons who occupy the relevant role, including those expectations explicitly stated in, for example, an employment or marriage contract. So, we can ⁹ At least not insofar as they are merely motivating reasons. Of course, some philosophers, such as those who follow Hutcheson and Hume, claim that any psychological feature that constitutes a motive to action also constitutes a justifying reason for action: desires are motives, and they provide justification for action. Nonetheless, the concept of a motivating reason is distinct from the concept of a justifying reason. ¹⁰ I take it that a norm of a role is just a statement of the obligations attendant upon the role. So, of course, people may disagree about the content of such norms. Thus, two people who ‘identify’ a role may be motivated to act in quite different ways. ¹¹ I have added in a relevant desire here in order to satisfy any Humeans about motivation. Those, like myself, who reject the Humean view that cognitive states alone cannot motivate might regard that judgment alone as sufficient to constitute a motivating reason for action.

224



say of a Nazi death camp commandant that it is an obligation of his role that he kill Jews as efficiently as possible: all that this means is that certain others, in particular, those who created the position to serve a particular function and those who are his superiors in the chain of command, expect and demand efficient killing of Jews from him. In other cases, such as with respect to the role of ‘wife,’ those obligations, in the sense of expectations, will not be so explicitly stated and may even be ambiguous or conflicting. In either case, we can call these ‘presumed obligations,’ insofar as they are what people, including, often, the person in the role, presume to be the obligations of a role occupant.¹² But if this is what we are talking about when we talk about role obligations, then it is perfectly consistent to say that one has a role obligation to do P but no reason to do P, and, in particular, no moral reason to do P or even a strong moral reason not to do P. And this is true even if we are talking about normative expectations rather than predictive expectations, that is, about what other people think that we ought to do as opposed to what they simply predict that we will do. My family might expect me to vote in a particular way, because members of our family have always voted in that way, but the mere fact of their expecting me to do so does not give me even a prima facie reason to do so. Of course, failing to meet expectations can have negative consequences that I have reason to avoid (see Andre 1991, 75), but meeting expectations can also have negative consequences, as the case of the Nazi death camp commandant shows. I might have reason to change other people’s expectations rather than to conform to them. The role of ‘wife,’ for example, has historically been laden with sexist and patriarchal expectations, and feminists have fought long and hard to dissociate the role from such expectations. (See Isaacs’s discussion of ‘knowing one’s place’, this volume.) So presumed obligations are empirical psychological facts about persons that have, in and of themselves, no genuine normative significance. Thus, from now on, I will set aside any identification of role obligations with presumed obligations, insofar as doing so would render role obligations normatively inert, that is, as not normatively binding on us. Many roles are assumed voluntarily in very explicit acts such as promising or contracting. And insofar as promises or contracts ground or constitute justifying reasons in the form of obligations, such explicitly assumed roles will come equipped with obligations in the sense of justifying reasons. But such obligations are the result of the voluntary assumption of the role, that is, of the voluntary assumption of those obligations, not the result of the mere occupation of that role. Not all obligations that are assumed voluntarily are assumed via explicit and discrete acts such as the making of a promise or the signing of a contract. As we will see with friendship, we can do a series of actions over time, no one of which in

¹² The term ‘presumed obligation’ arose in the context of discussions with Tracy Isaacs.

      ? 225 itself provides grounds for obligations, but, added together, ground justifying obligations, potentially wide-ranging and very strong obligations. In any case, in talking about role obligations we need to always be clear as to whether we are talking about obligations that are the result of some voluntary act on the part of the agent in assuming the role, or talking about obligations that are somehow the result of the mere occupation of the role, regardless of whether one has voluntarily assumed it. In section 5 I will argue that any ‘obligations’ attached to nonvoluntarily assumed roles are mere presumed obligations, that is, not genuine normative requirements.¹³ In asking whether obligations of friendship are role obligations, then, I will put aside questions about motivating reasons in what follows. This is because, as I have said, such questions are empirical psychological questions that cannot be answered via philosophical inquiry alone.¹⁴ I will also put aside presumed obligations insofar as they, like motivating reasons, are not in and of themselves of normative import. In asking whether the obligations of friendship are role obligations, I take myself from here on to be asking whether those justifying reasons arising from friendship are best understood as justifying reasons that one has in virtue of occupying the social role of ‘friend.’

4. What is Friendship? There are clearly societal expectations with respect to how friends are to be treated and valued, both by each other and by those outside of the relationship, particularly with respect to our treatment and valuation of friendship and friends as compared to romantic or sexual relationships. We are implicitly told, it seems, that our primary commitment is, or ought to be, to a romantic/sexual partner (and then, eventually, to the children we may—and are often expected to—have with that partner). ‘Mere’ friends¹⁵ then take a secondary role in our lives, with a major part of their function being to support us in our primary commitments to partner and/or children.¹⁶ Those who place friends above romantic partners are viewed with disapproval, and those who structure their loves around friends rather than around a spouse and/or children are often pitied, it being assumed that they were ¹³ See Jeske 1998 for a fuller defense of this claim, and for my arguments against Hardimon’s view. ¹⁴ Of course, people are often motivated by their understanding of themselves as friends, but people are also often motivated by their understanding of themselves as a person on a river bank near a drowning person. Such motivation, then, does not seem relevant to whether we should understand obligations of friendship as role obligations. ¹⁵ Of course, ideally, one’s romantic partner, and eventually, one’s adult children, will also be friends. So when I speak of ‘mere’ friends I do not mean the ‘mere’ to be in any way denigrating, but only to indicate the lack of a sexual or biological element to the relationship. ¹⁶ One needs only to do a Google search of ‘friends or relationship’ to see how often those in romantic relationships are warned of the dangers of not moving their friends to a secondary role in their lives.

226



unable to attain the more valuable goals of romance and parenthood and so settled for friendship. This is one story about what we can think of as the social role of ‘friend,’ if by a social role we mean to be referring to a societal conception of a function, its value, and its place in an entire network of personal relationships. And, of course, people have both predictive and normative expectations of how friends will and ought to behave. But, as Aristotle pointed out millennia ago, we also have a concept of what a ‘true’ or ‘genuine’ friend is, and it is that concept that philosophers of friendship have been trying to analyze ever since.¹⁷ I am going to present my own understanding of what a true or genuine friend is, how a true or genuine friend differs from some other understandings of a friend, and why even if the latter can helpfully be understood as social roles, the former cannot. Genuine friendship is best understood as a complex relationship between two¹⁸ people¹⁹ involving some sufficient combination of the following elements: (1) a mutual special concern for each other’s well-being considered as an end (i.e. not merely as a means to something else such as one’s own pleasure or improved social status), (2) special concern that is stronger than that felt for any other person merely qua person, (3) a history of interaction with each other that displays in some way that mutual special concern, (4) some sort of positive attitude such as love or affection (as opposed to other sorts of positive attitudes such as respect or approval), (5) a desire to be a part of each other’s lives in some way, and (6) a special knowledge of each other. I say a sufficient combination of those elements because I don’t think that (1)–(6) provide us with necessary and sufficient conditions for what constitutes friendship. While it is true that the vast majority of relationships that are friendships will meet all six of these conditions, it

¹⁷ This concept of a genuine friend has some paradigmatic representations in film and literature; for example, think of Ellis and Andy in The Shawshank Redemption and Huck and Jim in The Adventures of Huckleberry Finn. Both of these are cases of relationships that subvert social expectations and lie outside of established social networks. This kind of example will be important for my later argument. The concept of a genuine friend as opposed to other types of ‘friends’ has, of course, its paradigmatic statement in Aristotle’s Nicomachean Ethics VIII and IX, where he distinguished friendships based on virtuous character from those based on pleasure or utility. ¹⁸ It might be suggested that friendship need not be a relationship between two people, because there can be ‘friend groups,’ composed of several people, and which can lose some members and gain new ones. However, I think that such cases really involve overlapping friendship relations between pairs of members of the group. In fact, I could be a member of such a group and be friends with some but not all other members. Friendship is not a transitive relationship, i.e. A may be friends with B and B with C, and yet A and C not be friends. In such friendship groups, A and C may be members because of their separate relationships with B. ¹⁹ I say ‘people’ here only because human persons are my focus. I in fact believe that human persons can be friends with members of other species, and that many of us are friends with our companion animals. Importantly, this friendship we have with our companion animals is, I believe, a friendship in the same sense that we have friendships with other human persons, not some derivative or lesser type of friendship. That many people would reject such claims I take to be more evidence of the way that obligations of friendship are not presumed obligations or attached to a socially defined role.

      ? 227 might yet be the case that a relationship could count as a genuine friendship even if it did not. (See Jeske 2008 for a fuller discussion of this issue.) Genuine friendship is constituted by intimacy (I take (1)–(6) to constitute intimacy), where intimacy is a kind of caring closeness.²⁰ This is why the term ‘special’ occurs more than once in describing the elements of friendship: friends are bound to each other in unique ways. They feel and care about each other, know each other, interact with each other, and play a role in each other’s lives that creates a metaphorical space in which only they can move. This is part of the reason why the loss of a genuine friend leaves such a permanent hole in one’s life: one can no longer enter that metaphorical space and be the person one can be only in relation to one’s friend, because nobody else has precisely the understanding of one that that particular friend has.²¹ And that understanding of oneself that the friend has, has helped to shape one’s own self-understanding, so that one’s self-understanding involves the friend as an important constituent (see Cocking and Kennett 1998). Genuine friendships involve, by their very nature, closeness between two people, and these close personal relationships create a private world for the parties to them, not in the sense of shutting others out or of withdrawing from the rest of the world, but in the sense of allowing each a unique way of being that is often felt to be one’s true or best way of being, at least when the friendship is functioning well.²² We say that our best friends know who we ‘really’ are, that we can be ourselves with them. But that self that we are with them is in part a creation of our friendship. It is quite clear that friends and friendships serve many functions in our lives. Human beings need to feel love and affection and to return such love and affection. Without that sense of mattering to at least some other persons, it is very difficult for most of us to find meaning or purpose in our lives as a whole. Our other projects diminish for us when we have nobody close to us with whom we can share our progress, our achievements, and our failures. Friends provide us with support and solace in difficult times, and they share our joy in the good times. Most prosaically but also perhaps most importantly, we just enjoy spending time with our friends, either in person, via Zoom in times of global pandemics, or in the exchange of texts or phone calls. Friends make us laugh, and even if they do not have the same interests we do, they take an interest in our interests because they

²⁰ There are, of course, types of relationships that can be said, in ordinary language, to be intimate but not caring, such as that between captor and long-term captive or between a spouse and his or her abusive partner. So I am simply here stipulating that I am using ‘intimacy’ to refer to a subset of what we would normally understand as close relationships, i.e. I am using it to refer to the ones that are caring. ²¹ This is part of the reason why the loss of a parent who has become one’s friend is so very painful: parents and children know each other in ways that no one else possibly could and their relationship extends over the entire course of one of the friends’ lives. ²² Badly functioning friendships can often make us feel like our worst selves, and that is one consideration that often makes it rational to terminate a friendship.

228



are our interests. With friends, we can do or say what we would never say to anyone else, because we are confident of being accepted—we do not need to selfcensor or be scrupulously polite in the way that we do in any wider arena. We can take off our public and even semi-public faces, and just breathe as who we are. In the tradition of Aristotle, some philosophers view friendship as a place in which friends ought to improve each other morally (see Thomas 1989). Others, such as Dean Cocking and Jeanette Kennett (2000), claim that there is an important distinction between being a good person and being a good friend: sometimes a morally good person will not act as a good friend would act and sometimes a good friend will act in a way that is not morally acceptable. I agree with Cocking and Kennett that the notion of friendship is not itself an inherently moral notion, although I do think that friendship is intrinsically valuable and is thus an important constituent of a good or a worthwhile life. I also think, however, that morally serious friends will aim to improve each other insofar as they care about one another, because they will thereby care about each other’s moral character. It will also be harder for morally bad people to sustain the type of care and concern involved in friendship, but certainly not impossible. I have no doubt that many top Nazis, for example, felt genuine love and concern for their friends and families. While friendship itself is not a moral notion, it does ground obligations in the sense of reasons for those party to the relationship. There are various theories about the grounds of such relationships (see Keller 2013). Elsewhere, I have argued for what Simon Keller (2013) calls ‘the relationship view,’ the view that our reasons to care for our friends are grounded by the special relationship of intimacy in which we stand to them (see Jeske 2008). It is the fact that my friend and I are intimate with each other that gives rise to reasons for us to care for each other. Thus, my friend and I have reasons to care for each other that we do not have to care for just any person qua person. Standardly, then, each of us has reason to be partial to our friends, insofar as we have more reasons to care for them than we have to care for just anyone, although, of course, the degree of need displayed by other people vis-à-vis our friends is also a relevant factor. Importantly, then, as I said in section 2, friendship is not a social role if one means by ‘social role’ simply a cluster of duties and responsibilities or rights. Friendship grounds our special obligations to our friends and so is not constituted by those obligations. Given the fact that our obligations to care for our friends are grounded in the relationship in which we stand to them, it is the nature of the relationship, including the nature of the friends themselves, which determines the content of those obligations. Caring for someone is often a complex affair, requiring knowledge of what the other needs or wants, how one can contribute to the other’s wellbeing, the extent to which the other wants our aid, and when it is best to simply stand by, ready to pick up the pieces after the friend tries to do without one’s

      ? 229 assistance. The particular care called for by a particular friend will often be a function of the way in which we figure into that person’s life: the ways in which we know her, the ways in which we interact, and the ways in which we express our love and concern. It is important, then, to note that friendships are as diverse as the people who enter into them. As I said, I don’t think that we can give necessary and sufficient conditions for what constitutes a relationship of friendship, and in large part this is a function of the diverse nature of persons: the ways in which people can be intimate with each other in a caring manner will vary with the natures of the persons involved. (In other words, each friendship will involve a unique combination of some or all of the elements (1)–(6) that I laid out above) And the way a given person is close in a caring way with one friend will always be different from the way that she is close in a caring way with another. Friendships involve complex negotiation over the course of their existence, as the friends adjust and adapt to one another, changing each other in the process. We negotiate who we are with each other and how we are going to care for one another and express that concern. What I hope has emerged from this portrait of the nature of friendship is the way in which it creates or constitutes a private space for the parties to it. Importantly, genuine friendship is a space in which friends can be themselves, shorn of social expectations and norms. In fact, as I have argued elsewhere (see Jeske 2021), social conventions such as the norms of politeness, are often shed in friendship, and their reinstatement often indicates a growing distance between the friends. Common features of friendship, such as private jokes and casual glances that can be read only by each other, are testament, even if of a mundane nature, to friendship as a private space. Thus, I think that thinking of friendship as a social role, however we understand that latter concept, would be to undermine what makes genuine friendship so special and important in our lives, namely its character as a space defined by two people who are, in turn, defined by it. In fact, given the nature of certain friendships, what friends owe to each other is to subvert social norms rather than to abide by them.²³ In order to more clearly see why this is the case, I want to contrast friendship with two other types of relationships which supposedly involve paradigmatic social roles, that is, the spousal or domestic partnership relationship, and the relationship between mother and daughter. In seeing how friendship differs from these and how its associated obligations differ from the obligations generated, or not generated, by those roles, we will see why thinking of obligations of friendship as role obligations is misguided.

²³ Huck and Jim are a good example of this. See Cohen 2020 for some real-life examples of people subverting social expectations regarding what ‘mere’ friends owe one another.

230



5. Is Friendship a Social Role? I am not going to consider social roles that are clearly defined by organizations, roles such as ‘professor,’ ‘mayor,’ ‘lead actor,’ or ‘private first class.’ These are standardly what we think of as jobs into which we are hired, promoted, or elected, and are defined in terms of functions, responsibilities, and privileges by organizational rules, by-laws, or contract (although, particularly at higher levels, the specified functions, etc., may be very general and/or vague and will thereby require interpretation). It is quite clear that being a friend is nothing like these roles. So I am going to begin by considering, first, the role of spouse or domestic partner, and, second, the roles of mother and daughter. The role of spouse or domestic partner²⁴ has one clear difference from being a friend: the former is a legal status with various legal rights, privileges, and responsibilities, while the latter is not.²⁵ Spouses acquire their status via either an explicit, discrete contractual act, or via a group of actions which one ought to know commits one in a certain way. Spouse simpliciter is a legally defined category, requiring an explicit contractual act (the act of marrying) on the part of those who would acquire the status. But there is also the status of common law spouse, which is not entered via an explicit contractual act, but, rather, by engaging in various acts (living together, presenting in ways commonly understood as ways in which a married couple present themselves) and meeting certain conditions (being of legal age to marry and of sound mind). Insofar as some states have laws stating the conditions to be met to qualify as a common law spouse, then one has accepted that role when one voluntarily meets those conditions and knows or ought to have known that doing so constitutes a commitment of a certain kind. So, the genuine obligations attached to the role of spouse are not acquired by mere occupation of a role, but, rather, by voluntary contract or a voluntary commitment. These are certainly legal obligations, but they are also moral obligations insofar as we have obligations to be faithful to our agreements. The legal system is clearly an institution, and it has determinate rules regarding spousal obligations to one another, both during a marriage and in its aftermath. Thus, ‘spouse’ as a social role is, as Dare says, ‘the product of deliberate institutional design guided by desired functional outcomes,’ even if such outcomes are not easily stated or are contested. And ‘spouse’ is also a social role created by ‘social expectation in turn produced by widespread beliefs about the function of ’ marriage (Dare 2016, 705). These expectations and beliefs are subject to change, as shifts in attitude to gay marriage show, and such changes can then impact the legal situation. Over time, also, as I mentioned before, sufficient criticism of social ²⁴ From now on I will speak only of spouses as shorthand for both spouses and domestic partners. ²⁵ Friends can, of course, enter into various legal relations with one another via contracts but doing so has nothing inherently to do with being a friend.

      ? 231 expectations can alter the legal situation even as cultural attitudes lag behind. For example, certain alterations in the legal status of wives regarding property and bodily integrity did not entirely dislodge sexist attitudes about female submission and compliance with husbandly demands, which remained prevalent at least in certain areas of the US and other Western nations through the latter part of the twentieth century. So presumed obligations may not match genuine obligations, insofar as normative social expectations do not match contractually acquired obligations. But the best spousal relationships have an aspect that goes beyond both their legal aspects and the social norms and expectations involving it, namely the way in which the best spousal relationships involve friendship. However, the role of spouse is defined by legal norms and social expectations—the best marriage must be a friendship, that is, a relationship of caring closeness or intimacy. This aspect of any marriage, as with any friendship, lies beyond the reach of social and legal demands. The obligations that spouses owe to one another as friends predate the marriage and continue to exist after legal contracts are entered into. It is sometimes the internalized social norms and expectations regarding spousal roles that play a part in undermining the friendship that once existed, thereby leaving nothing but a marriage, which the parties often then decide to terminate. Some would say that civil marriage exists as an institution to protect the intimate relationship underlying it. But then the question arises why we do not regard any such intimate relationship as deserving of the same protections and privileges (see Brake 2012). Marriage has always been an institution deeply embedded in social understandings of gender roles, inheritance, alliances between families, appropriate rearing of children, and unifying of socio-economic class. As a legal institution it has been shaped by social norms and expectations, which have in turn been shaped by the legal institution. But behind any good marriage are two friends (often, but not always, with ‘benefits’) who sometimes struggle as they feel compelled to conform what was previously a private space into one that conforms with social expectations governing spousal roles. Here I think that it is quite clear that the obligations—in the sense of justifying reasons—of friendship need to be prioritized over the expectations of other persons or presumed obligations. However, the way in which our society encourages marriage and romance to be placed above friendship confuses the situation, often leading to a denigration of what really ought to matter about any good marriage. Now let’s consider the roles of mother and daughter. What it is for one person to be another’s daughter is complicated, because there are different senses of what it is to be a mother (see Jeske 2017). There is of course the biological notion of mother, where a mother is someone who has contributed an ovum that is fertilized and then proceeds to gestate the resulting fetus in her womb until she gives birth to it. We now have various alterations of that story: a woman who contributes an ovum but who does not carry the fetus in her womb, a woman who carries

232



someone else’s ovum in her womb, etc. There is also what we can think of as the custodial notion of a mother, where a mother is a woman who is the primary caregiver for a dependent child. A custodial mother may be legally recognized as such, whether or not she is a child’s biological mother, or she may be such de facto but yet not have the legal rights and responsibilities accorded to legal custodial mothers. We can think of a woman or girl satisfying any number of descriptions, then, in order to qualify as someone’s daughter. The relationship between mother and daughter has, for our purposes, one clear difference from the relationship between spouses. While both spouses acquire their roles voluntarily, the mother‒daughter relationship is not symmetrical in that way: mothers at least often acquire their roles voluntarily while daughters almost never do. To become a biological mother, a woman often chooses to have sex or to be artificially inseminated. (Of course, cases of rape or jurisdictions in which abortion is not readily available complicate this picture.) To become a custodial mother, one either adopts or chooses to foster a child, takes over the care of a child from another, or becomes a biological mother and then keeps the child. But one becomes a biological and or custodial daughter via the choices of others, choices over which one usually has no authority. One cannot opt out of being someone’s daughter, just as one had no choice in becoming such. Both mothers and daughters have presumed obligations, that is, people have normative expectations about how mothers and daughters are to treat one another. Mothers have legal obligations regarding their daughters, and also obligations arising from their voluntary assumption of the role of mother, insofar as becoming a mother is a commitment to care for a child. But the role of daughter is not a contractually based one. It does not seem that the mere fact of a woman’s having given birth to me or having custodial responsibility for me grounds any justificatory reasons for me to care for her. Of course, we might think that if the woman has given me a wonderful life or raised me in a way that goes beyond minimal requirements, then I ought to be grateful to her. But obligations of gratitude, if they exist, are not attached to a role as such, and can also arise outside of contexts involving the occupation of a role. However, daughters often do have moral obligations to their mothers, and that is the result of the friendship that often forms between an adult daughter and her mother.²⁶ I was extremely close to my mother: there is no doubt that we had a very well-functioning friendship. I have no doubt that my mother was both my biological mother and my custodial mother when I was a child. However, occupation of those social roles of daughter was not the grounds of my most stringent obligations to my mother; rather, the grounds of those duties was the friendship between us. There is no doubt that her dedication to her custodial role played an ²⁶ Putting aside any derivative moral obligations present, such as those arising from promises or from the promotion of intrinsic value.

      ? 233 important causal role in generating our later intimacy and my gratitude for such played a role in my love and affection for her, and her biological relationship to me also played a causal role in generating her dedication to caring for me in my dependency and later. But we came to be the very best of friends, and our interactions, as we both grew older, while certainly colored by our long mother‒ daughter relationship, also generated for us that private space that I have suggested is so characteristic of friendship. This friendship between us was not a matter of social expectations or of the need to have people serve various functions—it was a matter of the caring closeness that we had developed over the years, in part facilitated by our occupation of what we can think of as social roles protected by certain legal and social norms. (See Jeske 2017 for a fuller defense of the foregoing view about filial duties.) What I have been arguing via these examples of spousal and parental relationships is that whatever social roles we take people to occupy, where those social roles are understood in terms of social and/or legal norms and expectations, they can also be friends. The friendship may lead to occupation of the role of spouse, or the occupation of the roles of mother and daughter may lead to the friendship. But the friendship itself is a creation of the two people party to it, and provides them with a space separate from all social norms and expectations. In fact, it’s being separate in such a way is what allows it to play the part it does in so many of our lives, namely, allowing us to be ourselves, or at least that self which is at play within the friendship.²⁷ Because of the fact that friendship is a realm constituted by the personal bonds between the two friends, it provides a space that can underlie or coexist with the networks created by social roles and their attendant expectations and norms. It is from this space that critiques of social roles can occur—as I said in the last section, friends may have obligations (justifying reasons) to subvert social expectations with respect to how they treat each other, and, thus, to change social expectations regarding the social roles they occupy alongside their friendship. To think of friendship as a social role and of its obligations as role obligations, then, would be to undermine the intimate character of friendship and of friendship’s potential for countering and changing social expectations about how we ought to live and structure our personal relationships.

6. Conclusion I have argued that, given the way that social roles are tied either to social norms and expectations, or to intentionally created institutions with delineated rights ²⁷ We are often different with different friends and they know us in different ways. This is because, of course, part of who we are in friendship is determined by the other party to the relationship.

234



and responsibilities, friendship ought not to be understood as involving social roles. Thus, the obligations of friendship are not role obligations. In fact, as I have stressed, the very unique and valuable part that friendship plays in our lives depends on its being prior to and beyond social norms and expectations in important ways. Social roles such as spouse or domestic partner, or mother and daughter, can coexist with the people occupying the social roles also being friends, where the friendship grounds obligations in the sense of justifying reasons that are only contingently related to the two persons’ occupation of the relevant social roles. It is the friendship between the spouses or between the mother and the daughter that is the personal bond between them, which may be facilitated or hindered by the simultaneous occupation of the social roles. But just because ‘friend’ is not a social role, it does not follow that we ought not to use social institutions such as the law to protect and maintain friendships. What we always need to keep in mind, however, is that what we are protecting is an intimate space that lies beyond social norms and demands.²⁸

References Andre, Judith. 1991. ‘Role Morality as a Complex Instance of Ordinary Morality.’ American Philosophical Quarterly 28 (1): 73–80. Aristotle. 1985. Nicomachean Ethics. Terence Irwin (trans.), Indianapolis: Hackett Publishing Co. Beran, Harry. 1972. ‘Ought, Obligation, and Duty.’ Australasian Journal of Philosophy 50 (3): 207–221. Bowie, Norman E. 1985. ‘ “Role” as a Moral Concept in Health Care.’ The Journal of Medicine and Philosophy 7 (1): 57–64. Brake, Elizabeth. 2012. Minimizing Marriage: Marriage, Morality, and the Law. New York: Oxford University Press. Cocking, Dean, and Kennett, Jeanette. 1998. ‘Friendship and the Self.’ Ethics 108 (3): 502–527. Cocking, Dean, and Kennett, Jeanette. 2000. ‘Friendship and Moral Danger.’ Journal of Philosophy 97 (5): 278–296. Cohen, Rhaina. 2020. ‘What if Friendship, Not Marriage, Was at the Center of Life?’ The Atlantic. . Dare, Tim. 2016. ‘Robust Role-Obligations: How Do Roles Make a Moral Difference?’ Journal of Value Inquiry 50: 703–719.

²⁸ I would like to thank Richard Fumerton and Tracy Isaacs for helpful discussion of this chapter and of related issues.

      ? 235 Hardimon, Michael O. 1994. ‘Role Obligations.’ Journal of Philosophy 91 (7): 333–363. Hutcheson, Francis. 1991 [1742]. An Essay on the Nature and Conduct of the Passions and Affections. With Illustrations on the Moral Sense. In British Moralists 1650–1800, pp. 305–321. D. D. Raphael (ed.), Indianapolis: Hackett Publishing Co. Isaacs, Tracy. This volume. ‘The Part We Play: Social Group Membership as a Role’. Jeske, Diane. 1998. ‘Families, Friends, and Special Obligations.’ Canadian Journal of Philosophy 28 (4): 527–555. Jeske, Diane. 2008. Rationality and Moral Theory: How Intimacy Generates Reasons. New York: Routledge. Jeske, Diane. 2017. ‘Filial Duties.’ In Palgrave Handbook of the Philosophy of Aging, pp. 365–83. G. Scarre (ed.), New York: Palgrave Macmillan. Jeske, Diane. 2021. ‘Polite Strangers and Rude Friends.’ In The Philosophy of (Im)politeness. C. Xie (ed.), pp.75–92. Cham: Springer Publishing. Jeske, Diane (ed.). 2022. The Routledge Handbook of Philosophy and Friendship. New York: Routledge. Keller, Simon. 2013. Partiality. Princeton: Princeton University Press. Luban, David (ed.). 1983. The Good Lawyer: Lawyer’s Roles and Lawyer’s Ethics. Totowa, NJ: Rowman and Allanheld. Sciaraffa, Stefan. 2011. ‘Identification, Meaning, and the Normativity of Social Roles.’ European Journal of Philosophy 19 (1): 107–128. Simmons, Andrew J. 1996. ‘External Justifications and Institutional Roles.’ Journal of Philosophy 93 (1): 28–36. Thomas, Laurence. 1989. Living Morally: A Psychology of Moral Character. Philadelphia: Temple University Press.

12 My Job and Its Requirements Thomas H. Smith

1. Introduction My topic is the ethics and metaphysics of occupations, such as teacher, waiter, and priest. My interest in occupations derives from a broader interest in the brand of moral naturalism defended by Judith Jarvis Thomson in her last book Normativity (2008). In Thomson’s chapter on what she calls ‘directives’ (ch. XII), which are judgements that some thing x should, must, or ought to ϕ, she maintains that ‘the concept “defect” lies at the heart of the concepts “should” . . . “must” [and] “ought”’ (230). Her account of directives, which I construe as judgements about requirements, relies on her account of ‘goodness-fixing kinds’; these are kinds that set standards by which their instances are evaluable (ch. II). For Thomson, simplifying somewhat, some x ought to ϕ just if x falls under some goodness-fixing kind K such that, for any instance y of K, if y does not ϕ then y is a defective K (212).¹ Thomson gives plumber, mathematician, and stenographer as examples of goodness-fixing kinds, and in her chapter on what it is for something to be a virtue in a kind K (ch. V), treats the kinds carving knife, seeing eye dog, and stenographer as being on a par (70). She does however make no mention of occupational kinds in her chapter on directives. In that chapter, she moves directly from requirements concerning the kinds toaster, pancreas, terrier, and seeing eye dog to ones concerning the kind human being. Her silence on requirements concerning the kinds plumber, mathematician, and stenographer suggests that either (i) she simply does not know what to say about these, or (ii) she thinks that there are no such requirements, or (iii) she thinks that such requirements can be fully accounted for in terms of requirements concerning the kind human being.

¹ I have simplified Thomson’s account in two ways. First, for actions that a human being ought to perform she adds this condition: that x knows what will probably happen if she ϕs and what will probably happen if she does not (216). (The discussion of non-actions is incomplete, but it seems that Thomson would leave the issue of which capacities and traits a human ought to have to ‘moral theory’ (218) as opposed to the kind of moral metaphysics that she is engaged in). Second, for non-humans, she adds a condition regarding goodness-fixing kinds that are sub-kinds of other such kinds (211–214). See my n. 7 for some discussion of this condition.

Thomas H. Smith, My Job and Its Requirements In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Thomas H. Smith 2023. DOI: 10.1093/oso/9780192843562.003.0012

    

237

Thomson’s silence is unfortunate, because, in the context of her moral naturalism, occupational kinds are of great interest and deserving of more attention. One reason why is that on the one hand, the kind teacher is in some ways similar to a functional artefactual kind such as toaster. It is practically analytic that teachers teach and toasters toast. And functional kinds like toaster and pancreas are, for Thomson, paradigm cases of goodness-fixing kinds. On the other hand, teachers are very different from toasters: toasters are subject to very few standards. Teachers are human beings and so subject to very many potentially conflicting ones. It would be useful to know what implications these similarities and differences might have. Another reason is this: one of the main attractions of Thomson’s moral naturalism, which I should say owes a good deal to Geach (1956) and von Wright (1963), and has much in common with Kraut (2007, 2011), is that it combines a hard-nosed demystifying agenda with a non-reductive pluralism about value and normativity. On the one hand, there is no mystery about the standards by which a toaster or pancreas is assessable; while human beings are more complicated, the idea that the same is true of them is refreshing. On the other hand, there are many varieties of goodness, in von Wright’s phrase, good toasters, good pancreases, and so on. (Contrast those Procrustean theories that rest everything on good wills, good states of affairs, or good worlds.²) It would be in keeping with both the demystifying agenda and the pluralism to take seriously the standards set by a kind such as plumber. After all, there is no real mystery about them. Notice, too, the many varieties of occupational role and, more generally, social role. Perhaps enquiry into occupational roles could give us the foundations for a more wide-ranging enquiry into less clearly defined social roles such as parent, friend, leader, and citizen. And perhaps this would enable the kind human being to bear less (or even any) of the theoretical load. The project of accounting for the standards to which we are subject by direct appeal to some non-mysterious conception of what it is to be a good or defective instance of human being is hugely ambitious. It is easy to despair of making much progress with it. The project of accounting for some standards, the ones that we fall under by virtue of our occupational roles, looks much more tractable. For one thing, the societal purposes that occupational roles serve are pretty clearly defined and understood (more so than is the case with parent, friend, leader, and citizen).³ Moreover,

² Indeed, for Thomson, act, event, fact, state of affairs, possible world, and, I would guess, will are not goodness-fixing kinds (25–26). I might add that another attraction of her approach is that it seems to give the right kind of theoretical role to our own attitudes. It doesn’t simply account for value and requirement in terms of what some person, or community might prefer, prescribe, commend, or promote, but instead by appeal to goodness-fixing kinds. But since some of these kinds, and certainly kinds such as teacher and school, will be in some way socially defined and constructed, normativity, in some of its forms, will be partly grounded in us. ³ ‘Bullshit jobs’ may be an exception. See my n. 6.

238



while all social roles are in some sense embedded in and defined by social institutions, because of the present state of social metaphysics we have a reasonably good grasp of the kinds of institution that embed occupational roles: they are school, restaurant, church, etc., corporate bodies with governance structures, powers, policies, etc.

2. Functions (1) Hearts are for pumping blood to enable it to circulate (2) Paring knives are for cutting foodstuffs Sentences (1) and (2) plausibly have a definitional status. They tell you something about what it is to be a heart or paring knife. Jeremy David Fix (2019, sec. 3) makes something like this point when he says that a claim attributes an essential property to a kind when it is explanatorily fundamental. (1) and (2) are fundamental in this way. If someone asks for an explanation of them, there is little to say, other than the same thing again, with added emphasis, to indicate that this is bedrock: (3) That’s just what a heart/paring knife is (4) That’s just what hearts/paring knives are for Likewise, (5) A teacher is for imparting knowledge and cultivating skills and epistemic virtues (6) A waiter is for taking orders, serving and attending to patrons’ needs have a fundamental, and hence a definitional status. Upon recognizing this, one mistake that, say, a waiter might make is that of thinking that (7) Since I am a waiter, I am, by definition, for taking orders, serving, and attending to patrons’ needs This is patently untrue, but perhaps easily conflated with, or inferred from (8) I am a waiter, and a waiter is, by definition, for taking orders, serving, and attending to patrons’ needs

    

239

which a waiter might truly think.⁴ Consider now: (9) Since I am a waiter, I am for taking orders, serving, and attending to patrons’ needs This too is a mistake, if a waiter thinks it. For no one is a means to the end of taking orders, serving, etc. It is not just that no one is for such a demeaning purpose. It is a mistake to understand ‘ϕ-er’ as used to identify someone by their occupation, as meaning thing for ϕ-ing. A vacuum cleaner is a thing for cleaning. A contract cleaner is a person who habitually cleans and has the credentials (which may just be reputational or recognitive) to clean. Something similar applies to any use of ‘ϕ-er’ that identifies someone by their occupation. Indeed, a contract cleaner is in some ways a more perfectly actualized instance of its occupational kind than a vacuum cleaner is of its artefactual kind. For not every vacuum cleaner cleans; a vacuum cleaner might never be used. But one is not a cleaner if one does not habitually clean (although one may be a qualified cleaner). One might think that no one is for anything (other than perhaps whatever it is that a human being is for, eudaimonia, or whatnot), but I think this is an overstatement. As we shall shortly see, members of the House of Habsburg were for ruling. That’s what they were born and bred to do. This second mistake is easier to make than the first. For (1) and (2) entail respectively (10) For any x, if x is a heart, then x is for pumping blood to enable it to circulate (11) For any x, if x is a paring knife, then x is for cutting foodstuffs So one might think that the like entailment holds for (5) and (6). However, (5) and (6) do not even entail the corresponding existential generalizations: (12) For some x, x is a teacher, and x is for imparting knowledge and cultivating skills and epistemic virtues (13) For some x, x is a waiter, and x is for taking orders, serving, and attending to patrons’ needs Should we be puzzled by this? In one way, no. Claims (1) to (6) are generic statements, and a generic such as ⁴ Compare: it is false that the number of planets is necessarily >5. But one might conflate this thought with, or infer it from, the following truth: The number of planets is 8, which is necessarily >5. See (Quine 1976, sec. 1).

240



(14) A horse has four legs is not falsified by exceptions. Still, horses generally have four legs, so we might be puzzled as to why every heart is for pumping blood, and no teacher is for imparting knowledge. Part of the answer is that when a generic has a definitional status as in (1), (2), (5), and (6)—and arguably also in (14)—the property it predicates is attributed to the kind, with no direct implications as to whether, or in what way, it holds of instances of it. (1), (2), (5), and (6) tell us something about what it is to be (as it may be) a heart, paring knife, teacher, or waiter, and we cannot from this infer anything about particular hearts, paring knives, teachers and waiters. Contrast: (15) Teachers are underpaid which is not definitional. This tells you something generally true of teachers. Still, one might wonder, why are occupational kinds like teacher and waiter functional kinds (essentially for a purpose) when their instances are not functional objects (not things for a purpose)? I think the answer is that no particular teacher owes her existence to the function that teachers perform, although teachers in general do. Contrast hearts and paring knives. Any heart exists in part because hearts are for pumping blood. For it is the product of a process by which hearts have been designed by natural selection to pump blood. (Roughly, a heart is designed to pump blood in that hearts that did this effectively were selected for.) Likewise, any paring knife exists in part because paring knives are for cutting foodstuffs. It is not just that it will have been designed and manufactured to do this; the process of design and manufacture has itself been designed, by a kind of cultural selection, to better produce instruments for cutting foodstuffs. (Roughly, effective processes of design and manufacture flourished while ineffective ones died out.) There are teachers because teachers are for imparting knowledge and cultivating minds and skills. For there are teachers because of processes, of training, recruiting, and retaining teachers that have been designed in order that teachers can effectively impart knowledge and cultivate minds and skills. But it is not true of any teacher that she exists, even in part, because teachers are for imparting knowledge and cultivating minds and skills. If you are a teacher, you came into existence in the normal way that humans do. There will be habits, skills, and policies of yours that exist because of what teachers are for. But your very existence is not explained by this function. Contrast schools (not school buildings, but the organized institutions that they support). Every school exists because of processes designed in order that schools effectively educate children.

    

241

One might suppose that my reasoning here relies on the mundane point that one is a human before one becomes a teacher, or that one is essentially human and contingently a teacher. But here is why that is not the central point. Some embryos are for research. (Whether this is right is another matter.) Any such embryo exists because of a cloning process that is designed to facilitate embryological research (cloning is not IVF, which some people use to have a baby). But any such embryo is an embryo before it is research material. And it is contingently research material. The example shows that being first of a natural kind and later, contingently, of a functional kind can suffice for being a functional object. But it may not show that being first a human and later, and contingently, of a functional kind can likewise suffice. For whether embryos are humans is moot. So here is another example that does show that. The House of Habsburg produced human beings that were for ruling, for each of them resulted from processes of breeding designed in order that the Habsburgs rule. But any such ruler is a human before it is a ruler and is contingently a ruler. To sum up this section, teacher is a functional kind, but teachers are not functional objects. If you are a practising teacher, then it is likely that, to some extent, you perform a function and serve a purpose, that of imparting knowledge and cultivating minds and skills. This is what teachers, generically, are for, and it is what your school is for. But it is not what you are for.

3. Jobs ‘Job’ like ‘role’ is multiply ambiguous. ‘My job’ can signify my occupation, that is, the type of work I do (16) or the type of worker I am (17). (16) My job is to teach (17) I am a teacher; that’s my job It can also signify a particular position I hold, a position within some corporate body, such as a school, restaurant, or church. (A self-employed person will not have a position in the same way, but they may have a position within a community, trades association or similar.) Examples of this use include (18) Karen now has Phil’s old job (19) The job has been advertised (20) Because of the restructuring, that job is being eliminated/split into two/ combined with another/redefined

242



But we also use ‘job’ to signify a duty, or collection of duties, that one has by virtue of one’s position and/or occupation. I think this is the most natural way of understanding sentences such as (21) My job is to prosecute criminals (22) The job I have is to visit and value your property to ensure a maximum achievable sale (23) It’s my job to cut the meat and sell it to customers (24) I have the job of delivering the bad news when we let someone go (25) My job involves supporting vulnerable people Are there duties to be certain ways as well as duties to do certain things? I am not sure. But it is certainly the case that you can be subject to requirements by virtue of your position and/or occupation to be certain ways. A Catholic priest should be male and unmarried. A driving instructor should be able to read a number plate at twenty metres. A sumo rikishi must be a male who has finished nine years of schooling; they must be at least 167 centimetres tall and weigh at least 67 kilograms, etc. We sometimes use ‘job’ to mean a requirement to be a certain way, as in (26) It’s my job to be attentive in case a patron needs something (27) It’s my job to know who the local gangs are and what they are up to But a more common way of putting it is (28) In my job I must be physically fit (29) In my job I must stay abreast of the latest developments in my field Henceforth I shall mostly speak, not of the duty but the requirement sense of ‘job’. ‘Job’ can also mean function, as when we say things like (30) A heart has the job of pumping blood to enable it to circulate This can create confusion as we also say things like (31) A teacher has the job of cultivating their pupils’ knowledge, skills, and epistemic virtues And I think (30) entails not only (32) but also (33). (32) For any x, if x is a heart, then x’s job is to pump blood to enable it to circulate

    

243

(33) For any x, if x is a heart, then x is supposed to/should/ought to pump blood to enable it to circulate Likewise, (31), on the face of it, entails not only (34) but also (35). (34) For any x, if x is a teacher, then x’s job is to cultivate her pupils’ knowledge, skills, and epistemic virtues (35) For any x, if x is a teacher, then x is supposed to/should/ought to cultivate their pupils’ knowledge, skills, and epistemic virtues (30) and (31) are true in both the function and the requirement sense. We can infer from the generic (30), read in either sense, to the universal (32) read in that same sense. But we can only infer from the generic (31) to the universal (34) if both are read in the requirement sense. No particular teacher is for what teachers in general are for. In one way this difference between what we can say about hearts and what we can say about teachers does not matter as (31) supports (35) regardless (although it is a nice question whether (31) does so on both readings). Still, any teacher who misses the difference between the function and requirement senses might mistakenly think that she is for something. How does the holding of a position and/or the having of an occupation generate requirements? Of course, one typically acquires contractual obligations to one’s employer when one fills a position. More broadly, one typically acquires obligations to fulfil expectations one knowingly induces in others (such as clients and colleagues) by presenting oneself to them as having the position and/or occupation that one has.⁵ But it is likely that the requirements that one is under by virtue of one’s position and/or occupation are not exhausted by those explicable in these ways. A police officer has no contractual obligation to remain physically fit and, given the high levels of obesity in many police forces, few people can seriously expect police officers to be physically fit. Still, one might think that a police officer ought to try to remain physically fit because of their occupation. I think it is simply that to enter a position and/or an occupation is, if one is sincere, willingly to accept and take on a job, in the requirement sense. You are subject to the relevant requirements because you willingly took them on. (There may be exceptions around deception, duress and the taking on of immoral or pointless requirements.⁶, ⁷) This is so even if one had no good options when one ⁵ Here I presuppose a version of the principle of fidelity defended by Scanlon (1998, ch. 7). ⁶ So some bullshit jobs do not generate requirements as the work serves no worthwhile purpose, and perhaps an immoral one. Of course, there may be other, e.g. contractual and financial reasons to do what your job requires. ‘Bullshit jobs’ is the coinage of Graeber (2018). I thank an editor for asking me to think about them. ⁷ Thomson’s view of non-human kinds is that if K is a sub-kind of K+, and any K that does not ϕ is a defective K, and any K+ that does ϕ is a defective K+, then the more encompassing kind trumps its subkind in that for any x that falls under both K and K+, it ought not to ϕ. She devotes some pages (211–214) to defending such a clause, to deal with an imagined case in which some terriers are sold

244



took a position, and has serious misgivings about it, or one’s occupancy of it. ‘This is my job’, one thinks. ‘It’s what I’m supposed to do.’ Another confusing matter is that your position is likely to have a function, even if you do not. Probably, it will have been created in order that some purpose is served, which means it is for the serving of that purpose (unless it has since been implicitly or explicitly redefined in order that it serve some other function or purpose). That is why managers say things like (36) The position is there to ensure that asset defects are managed effectively (37) We have created this position in order that car owners are deterred from parking illegally The purpose or function of your position need not be your purpose or function, or even your job, in the requirement sense. Your job may be to ticket cars that are parked illegally, not to deter anyone from parking them illegally. For many positions, their requirements include a requirement to present yourself in a certain way. In some cases, this requirement is to present yourself as being for this or that. (Think of the ‘Here to help’ badges that sales assistants wear.) This can be a cause of spiritual damage. Many of us want to be good ϕ-ers, and to be seen as conscientious and professional ϕ-ers, but when this involves presenting oneself as a functional object the requirement can seem to threaten one’s authenticity. For some, no doubt, the adoption of a work persona is enjoyable and enriches their lives. We shall return to self-presentation later when we discuss Jean-Paul Sartre’s café waiter. In order not to find the requirements of one’s working life overwhelming, I recommend keeping the distinctions made in this section clearly in mind. In short, you have a job (requirement) by virtue of your job (position) and/or job (occupation). Your job (position) is likely to have a job (function) which need not be your job (function) or job (duty). Probably, you have no job (function), but it may be part of your job (requirement) to present yourself to others as having a job (function).

4. Requirements In the last section, we saw that you ought to be and do the things that it is your job to be and do (subject to exceptions around deception, duress, immorality, etc.) by with their vocal cords cut, as ‘Quiet Dogs’. She wants to ensure the result that a terrier ought to be able to bark, even if it is also of the kind Quiet Dog. One might wonder why Thomson bothers to deal with this odd case. I speculate that she was thinking about how to block an inference such as the one from ‘If a financier isn’t greedy, she is a defective financier’ and ‘You are a financier’ to ‘You ought to be greedy’. Her rule about sub-kinds would block the inference because even if any non-greedy financier is a defective financier, any greedy human being is a defective human being.

    

245

virtue of your position and/or occupation. In this section I say a little more to clarify and explain the nature of these requirements. Consider again (33) For any x, if x is a heart, then x is supposed to/should/ought to pump blood to enable it to circulate (35) For any x, if x is a teacher, then x is supposed to/should/ought to cultivate their pupils’ knowledge, skills, and epistemic virtues First, these ‘should’s and ‘ought’s need not imply that x has reason to do anything. Nor need they imply that x has an ability to comply with the ‘ought’. Nor need they imply that x is at fault if x does not comply. These points are obvious when x is a heart or paring knife. But they hold even when x is an agent. To see this, consider someone constitutionally unsuited to their job and incapable of effectively discharging some core responsibility. This might be the result of a hiring error or labour shortage. Or the person may have been suited to their occupation, but become ill-suited, due to some crisis, illness, or accident. If you are a heavy jockey, a partially sighted driving instructor, a waiter prone to myoclonic twitches or jerks, a teacher who cannot make themselves understood because of their accent, or an atheist priest, then you cannot effectively do something that, by virtue of your position, you are supposed to do. You are unable to do what you ought to do; you may not be at fault for this, and may not have any reason to do anything in particular about it.⁸ Second, these ‘should’s and ‘ought’s are requirements on x. The claim is not that (38) It ought to be the case that, for any x, if x is a teacher, then x cultivates their pupils’ knowledge, etc. Although this is true too. In (35) the ‘ought’ is predicated directly of x; it states a requirement on x. (38) says of something sentence-like that its truth is somehow a requirement. (38) expresses a standard, or norm, by which the world may be evaluated. (33) and (35) set a standard, or norm, by which a particular is evaluable as (as it may be) an excellent, adequate, or defective heart or teacher. Third, it would be too quick to say that a heart en route from its donor to its transplant recipient is, while it is being transported down the motorway, defective, as it is not pumping blood. Likewise, if a teacher does not comply with a

⁸ You may have reason to leave the position, but if you are not supported in making this transition, which is likely to involve suffering a period of unemployment, and some re-skilling, then you may not have this reason.

246



requirement imposed by her position and/or occupation, she is thereby a defective teacher only if under favourable conditions she does or would comply. Fourth, some requirements may be qualified. If a heavy jockey is sufficiently fearless, fit, and skilful they need not be a defective jockey. One might even think that there are, or could be, heavy jockeys for whom their weight is not even a defect, for they manage to use it to their advantage. More generally, it is plausible that some requirements are such that if some K violates them this can but need not suffice for its being a defective K. Perhaps, for some of these requirements, the violation must suffice for the thing in question’s having a defect, but this defect can be all-things-considered compensated for by the thing’s other characteristics (call these pro tanto requirements). Perhaps, for others, the violation can but need not suffice even for the thing in question’s having a defect (call these prima facie requirements).⁹ No doubt there is more to be said, but it suffices to note that some requirements on us humans are likely to be qualified in something like these ways. While (33) and (35) wear the fact that they express conditional requirements on their linguistic sleeves, and while they may also express qualified requirements (as just conceded), they do not express requirements that are relative to some kind. As is well known, in ‘good teacher’, the evaluative adjective is attributive, that is, semantically inseparable from the noun that it modifies. A good teacher is not someone that is (a) good and (b) a teacher (Geach 1956). Such expressions relativize an evaluation to a kind. There is no natural way of likewise relativizing an ‘ought’ to a kind. Granted, we do say things, like, ‘As a teacher, she ought . . .’. But I can only hear this as meaning something like Because she is a teacher, she ought . . .¹⁰ Or She ought, using the powers she has by virtue of being a teacher, . . .

⁹ We normally talk of prima facie and pro tanto ‘ought’s when these are reason-entailing ‘ought’s, but I do not see why we could not apply these qualifications to the requirements under consideration here. ¹⁰ This is the sense in which Thomson thinks we can say things like ‘As a teacher, she ought . . . ’. It is very easy to misread Thomson on this point, because she says that ‘ . . . what a thing ought to do is what it ought to do qua being of this or that kind’ (210). But she does not here relativize the requirement to the kind. Just before, she glosses ‘qua’ as a modifier of ‘ought’, as ‘in virtue of ’. And just after she glosses the claim quoted above as: Alternatively put: it is not true of A that it ought to V unless A is of a kind such that everything of that kind ought to V—where it is the fact that everything of that kind ought to V that makes it the case that A ought to. (210) It is also clear from her chs. X and XI that for her ‘ought’ is never relative.

    

247

Or perhaps She ought, while on duty, . . . We can say that something is good with respect to its being a teacher or toaster, leaving open that it may be bad in other respects. But we cannot likewise say that someone ought to ϕ with respect to their being a teacher, leaving open that they ought not to ϕ with respect to their being, say, a parent. If we want to reconcile an ought to ϕ because she is a teacher with an ought not to ϕ because she is a parent, the only way to do so is by qualifying one or both as pro tanto or prima facie or something similar. And if we want to reason from some so qualified ‘ought’s about a particular case to a single unqualified ‘ought’ (and perhaps we should not want this in every case) there is no substitute for the sensitive use of the relevant facts, principles, hunches, and value-judgements. Of course, one can release oneself from the grip of an ‘ought’ that is grounded in one’s position and/or occupation by changing one’s job. But this may not be straightforward, and, until one makes the change, the ‘ought’ is binding on one, and always will have been binding on one. Consider now this passage from Gilbert Harman: Sometimes ‘ought’ or ‘should’ is used to express an evaluation of something in terms of some associated function, need, role, normal case or ideal. A heart ought to pump at a regular rate; if it does not, there is something wrong with it. A tree should have strong roots; if it does, we say it has good roots. A paring knife ought to cut well. A teacher ought to help students acquire an interest in learning and an ability to learn. ‘Ideally, a plain yogurt should have some astringency as well as a sweet/sour character . . . ’ (Consumer Reports, August 1983, p. 386). There is something wrong with Tess that leads her to torment little Eddie like that. A judgment of this sort made about a particular agent, as in the last example, need not imply that the agent has reasons to act in a certain way. We can feel there is something wrong with Tess without feeling that it is wrong of her that she acts in that way. (Perhaps Tess is afflicted with a compulsion of some sort.) However, a judgment of this sort is not just an evaluation of a situation, like the judgment, ‘It is wrong that children work under such conditions.’ It is an evaluation of an agent. (Harman 1986, 135)

This reprises many of the points just made. Harman makes the point about this ‘ought’ not implying reasons, fault, or ability (this is implicit in his use of the supposition of Tess’s compulsion). He makes the point that it applies to an agent, not a situation. He does not attempt to relativize ‘ought’. Harman makes a small

248



error when he characterizes these uses of ‘ought’ and ‘should’ as evaluative.¹¹ For one does not evaluate anything simply by saying that it ought to do something or be some way. Statements of this sort express requirements, standards, or norms by which we might evaluate things, but they do not, in and of themselves evaluate anything. Even Harman nods. There is a more important error that is at least suggested by Harman’s ‘wrong with’ formulation, and the fact that, like ‘ought’, it isn’t, by him, and isn’t, naturally, relativized to an occupational kind.¹² If you are a heavy jockey or atheist priest, it is not just that you need not be at fault and that it need not be wrong of you that you are that way. You need not be faulty, and there need be nothing wrong with you. You are a defective jockey or priest. You are not doing something that you ought to be doing, or you are not some way that you ought to be. But the fault is not in you. There may be some other thing that is defective, the panel that hired you, or, more generally the societal mechanisms for the division of labour. Perhaps these are too insensitive to individuals’ skills and needs. Or perhaps it is just your and your employers’ luck that is defective. Why is it that if you are a bad teacher, it does not follow that there is something wrong with you? Is it because, as we argued in the previous section, you are not a functional object? I’m not sure, for you do fall under a functional kind and this does ground non-relativized ‘ought’ judgements. You are not employed as a teacher in the same way as a knife may be employed as a paperweight. The standard by which we evaluate you is external to the kind human being, but internal to the kind teacher, and you are a teacher. In this sense the standard is internal to you. The standard by which we evaluate a knife employed as a paperweight is external to it, because a knife so employed is not a paperweight.¹³ This shows up in a linguistic datum marked by von Wright: if a knife so employed is effective, it is not a good paperweight but good as a paperweight (von Wright 1963, ch. II, sec. 1). But if you, so employed, are effective, you are not just good as a teacher, you are a good teacher. I think the answer is just that teacher is not a fundamental kind. Whatever your job, what you are is a human being. By this I do not merely mean that it is what ¹¹ Harman’s ‘good roots’ is evaluative, as are his uses of ‘something wrong with’. But his ‘should’s and ‘ought’s are not. Of course, in the right context, you can conversationally implicate an evaluation by stating a requirement, e.g., by throwing down a blunt paring knife while saying ‘A paring knife ought to cut well.’ ¹² If I say ‘there is something wrong with my teacher’ I might conversationally implicate that the defect is a defect in a teacher, but I do not say this. ¹³ I borrow this use of ‘internal’ and ‘external’ from Fix (2019). But Fix identifies the idea of a standard that is internal to a kind under which a thing falls with the idea of a standard to which that thing is ‘by nature’ subject (15). As the example of a teacher shows, these are distinct ideas. (I press this point a few paragraphs later.) Or rather, they are the same only if you suppose that fundamental kinds are the only kinds. Perhaps Fix does suppose this, although he does not say so in terms. All his examples of kinds are of fundamental kinds.

    

249

you necessarily are: each of us is necessarily an animal as well as a human being. I also mean that, as Wiggins says, human being . . . [carries] us to a certain sort of conclusion . . . Where no such conclusory answer is provided to the question ‘What is it?’, it will be indeterminate what we are thinking about. (2016, 45)

If I refer to something as an animal, it will be indeterminate what we are thinking about. (A pike? A butterfly?) And so, you can still press the question what is it? If I refer to something as a human being it is not likewise indeterminate what we are thinking about. Of course, you can ask for more information. And I can add that the human being in question is a teacher, or waiter. But to specify or give details is not to determine what was previously indeterminate. You already knew what we were thinking about. The additional specification told you something more about it, namely what it does (for a living), and not what it is. As we might say, human being is our determinate form. Artefacts and organs differ from teachers in this respect. A knife is of the fundamental kind knife. This is its determinate form. A heart is of the fundamental kind—it has the determinate form—heart. This holds even of a knife (or whatever) that is used as, and even good as, a paperweight. A defective heart or knife is bad as what it is. It is by nature bad. We can think of it as not fully what it is, not fully realized or actualized. Certainly, we can infer that there is something wrong with it. If you are a defective teacher, you are not thereby bad as what you are.¹⁴ You are not thereby by nature bad. You are not thereby not fully what you are, not fully realized or actualized. So, we cannot infer that there is something wrong with you. Many of us are not well suited to every aspect of our jobs. We may try to improve ourselves, and to compensate for our defects as teachers, or waiters, by excelling in other aspects of the job. We might persuade ourselves that we are violating only a prima facie or pro tanto requirement and are all-things-considered non-defective. But when these defects are things we cannot change, and perhaps especially when they are brought on mid-career by a diagnosis, a crisis, an accident, or some such, we are apt to punish ourselves, like an atheist priest, with the thought what is wrong with me? The answer is often ‘nothing’. One can violate a requirement as fundamental as that of being able to ϕ if a ϕ-er, and have nothing wrong with you, although you are not as you ought to be.

¹⁴ Granted, you are bad at what you do (or at least bad at something that you do). But we should not think of what a person does (their occupation) as their principle of activity in the AristotelianLeibnizian sense borrowed by Wiggins (2016). It is not that by which we can determine which particles of matter are parts of you, and which are not, as you trace a path through space and time. Your principle of activity is that of living.

250



5. Positions As we have seen, a teacher is unlike the teaching position she holds in that the latter will typically be a functional object, existing for a purpose and designed and created as such. To serve this purpose, a position must be, broadly speaking, used by a human being, and, normally, as it was designed to be used. Plausibly teaching position is a fundamental kind: for any teaching position, a teaching position is what it is. (We saw that by contrast, for any teacher, a human being is what it is.) In all these ways, positions are akin to artefacts. Strikingly, they appear to have much in common with another kind of socially constructed or defined artefact, corporate bodies. For a corporate body generally • is a socially designed and constructed object (hence a kind of human artefact) • has human constituents • persists through actual and counterfactual replacement of human constituents • can share human constituents with other corporate bodies • is of a kind that has a function, for example (39) A University is for facilitating education and scholarship in communion • is thereby evaluable. So too on the face of it, a position generally • is a socially designed constructed object (hence a kind of human artefact) • has a human constituent (occupant) • persists through actual and counterfactual replacement of constituents (occupants) • can share human constituents (occupants) with other positions • is of a kind that has a function, for example (40) A professor is for teaching, research, and administration • is thereby evaluable. If our theory of positions is that they are essentially bite-size corporations, then it displays the virtues of simplicity and parsimony: it introduces no fundamentally new concepts or kinds. (We must reconcile this with the fact that recourse to the familiar can be a theoretical vice.)

    

251

This conception of positions suggests a further thought: positions are parts of corporate bodies. Think of a company flowchart (an ‘org chart’). What does it depict? Socially constructed entities, one might think, and the relations of, for example, accountability and authority that hold between them. So, there are the faculties, divisions, and departments. These are plausibly thought of as parts of the corporate. But positions may appear on the flowchart also, for example, the position of HR manager. We might think of these as parts—perhaps the simple, partless parts—of the corporate body too. In short, we might think that what the flowchart shows is the mereological structure of the corporate. Functional objects enter into part‒whole relations. A bicycle has, as parts, wheels, a saddle, derailleur, and so on. A functional part of a functional whole will typically have, as its purpose, some contribution to the well-functioning of the whole, some specifiable way of enabling or helping it to fulfil its purpose, and it will be evaluable as such. Given this, we might now suppose that a position, such as that of professor is a functional part of a functional whole, university, and has as its purpose some specifiable contribution to the well-functioning of that whole, and that it is evaluable as such. This view must accommodate the fact that, like other corporates, universities have human constituents. One way to do this would be to distinguish parthood, which holds (at times) between artificial bodies (the university, the departments, the positions) and occupancy which holds (at times) between each of these things and human beings. So, I currently occupy a lecturing position, but I also cooccupy, with others, a university, faculty, etc. But we need not settle these details here. One might now suppose that the requirements on you, if you are a teacher, waiter, etc., derive from the requirements on your position, which are set by its function within the corporate. The details of the derivation need not be settled here. But we might suppose that they are more complicated than the simple inheritance, by you, of the evaluative and normative properties of your position. If there is a problem with the professor it does not follow that this problem is with you, if you are the professor. Perhaps you are part of the solution. Perhaps you are required to get together with colleagues and managers and seek a resolution: a sabbatical, a job-share, a change of role. But there is no reason to suppose that there will be simple inheritance, by you, of the professor’s fault. The canvassed position‒occupant dualism has emancipatory potential then, the potential to liberate you from an attribution of fault to your position. But I don’t think the dualist view can bear the weight here given to it. If positions are objects at all, they do not bear the relevant properties. The analogy between corporations and positions is much weaker than I initially made it seem. We have very many terms for corporate bodies. ‘The Supreme Court of Justice’, ‘The University of Manchester’, ‘Microsoft’, and so on. And it is not just the big beasts. ‘The Hebden Bridge Co-op Store’, ‘The Swindon

252



Moonrakers’ (a pub quiz team), ‘Handforth Parish Council’, and so on. We frequently use these terms referentially and, on the face of it, predicate a wide range of properties of their referents: locational properties, legal properties (ownership of assets, contractual obligations), psychological properties (knowledge, plans), actions (announcements, strategies), powers and normative properties (duties, rights, and responsibilities). We carry on as if these predications are not systematically false. And these predications are not straightforwardly ‘paraphraseable away’ as sentences about human beings. Of course, if a corporation does something or is some way, this is at least in part because human beings have done things and are, or have been, certain ways. But for the corporation to do what it did is not simply for the human beings to do what they did; we can’t offer a statement of the latter as a paraphrase of the former. A human being’s making of an announcement, for example, does not by itself suffice for the corporation that it speaks for making an announcement. If a human being makes the announcement and this makes it the case that the corporation makes the announcement, then this obviously suffices. But we can’t give this as a reductive truth-condition as it contains the very thought that we are trying to paraphrase away, namely that the corporation made an announcement. Of course, more could be said. But taking the evidence at its face value, it seems that corporations exist, and we have an interest in predicating all sorts of interesting things of them. Contrast positions. We have very few designated terms for positions. To see this, we should first be clear about which uses of expressions do, and which do not, refer to positions. ‘The U.S. presidency’ refers to a position. ‘The U.S. president’ does not. As Russell (1905) taught us, (41) The U.S. president is napping quantifies existentially, being equivalent to (42) Some x is (uniquely) U.S. president and x is napping. There are uses of ‘The U.S. president’ that have readings on which they generalize beyond any particular incumbent, such as (43) The U.S. president has the power to make treaties (44) The U.S. president has the responsibility either to sign or return with objections every bill passed by Congress But these are naturally understood as quantifying universally, thus

    

253

(45) For any time t and person x, if x is, uniquely, U.S. president at t, then x . . . etc.¹⁵ While others, which cannot be read as universal quantifications as they admit of exceptions, such as (46) The president pardons a turkey at Thanksgiving are naturally understood as generics akin to ‘The lily flowers in June’. I cannot find any uses of ‘The U.S. president’ that cannot be understood in one of these ways. Likewise for other expressions that identify a human being by their occupancy of a position. Once we understand this, we should be struck by how few designated terms for positions we have. We might have used expressions like ‘The Apple CEO’ to pick out a position that persists through the actual and counterfactual replacement of its constituents, much as we use ‘Apple’ to pick out a corporation that persists through the actual and counterfactual replacement of its constituents. But we do not. We might have coined a marker, like the suffix ‘-ship’ or ‘-y’ that we could systematically deploy to generate such a term. But we did not. We do have a device for systematically turning some temporally non-rigid designators of humans into temporally rigid designators of their positions: it is ‘The position (office) of . . . ’ locution.¹⁶ But even this only works for kinds of position that only one person has at a time. I am a lecturer of philosophy in the University of Manchester, and not the only one, so one cannot use this device to generate a term for my position. For positions such as mine, we must resort to expressions like ‘Tom’s job’, or ‘Phil’s old job’. I think we can easily explain why we do not have many terms for positions. We do not have many things we want to say about them. In this too the contrast with corporations is striking. We do not predicate locational properties, legal properties (ownership of assets, contractual obligations), psychological properties (knowledge, plans), actions (announcements, strategies), powers or normative and evaluative properties (duties, rights, and responsibilities) of them. We might say (47) That’s a great position (48) The position ought to be eliminated/split into two/combined with another/redefined

¹⁵ Where a power (or similar) was acquired at a certain time, we need only change this to ‘For any time t since 2001 . . . etc.’ For some sentences we might want to add a universal quantifier ranging over (sufficiently nearby) worlds. Arguably, as ‘the president’ is ontologically committing and universal quantifications can be vacuously true, we should add a clause to the effect that for some t there is at least one y who is uniquely president at t. ¹⁶ I borrow ‘rigid’ and ‘non-rigid’ from (Kripke 1980) of course.

254



But I’m not sure that (47) evaluates the position; it seems to evaluate someone’s having it. It is perhaps elliptical for ‘That a great position for [depending on the context] the right person/you/someone . . . ’ etc. Nor does (48) predicate a requirement of the position. It can only be understood as saying that it ought to be the case that the position is eliminated/split into two . . . etc. One might suppose that the U.S. presidency has powers, duties, and rights that it transfers to, or bestows upon, anyone who occupies that office. I doubt it. Biden is president and has the powers, duties, and rights of any president because, after a popular vote, he was certified as president by the Electoral College and Congress. And ‘He is president and has the powers etc. of any president because he was certified as such’ is, in turn, true because of certain social facts: roughly, because of the general or collective acceptance by U.S. citizens of U.S. constitutional law. Do we deepen our understanding by inserting into this explanation the idle wheel ‘because he holds the position of the president’? I don’t think so. The presidency is not an agent; it doesn’t do anything. So it is hard to see how it could have powers etc., and how it could transfer anything to anyone. To explain the president’s powers, duties, and rights by appeal to some such act of transfer is, perversely, to explain the familiar with the unfamiliar. Our main interest in positions is in (i) tracking their persistence over time, especially with regard to the actual and counterfactual replacement of their occupants and (ii) the networks or structures made up of them, the relations between them, and their human occupants. And even this is not that interesting. This explains why we have very little to say about them, why we have few designated terms for them and why the few that we do have, or which are easy enough to construct, are for esteemed or powerful positions; for these will generally be of more interest. We can pursue this interest without referring to, or quantifying over positions at all. Instead of talking about someone getting the presidency we can talk about them becoming president. Any talk of the history of the presidency can be recast as talk of the history of those people who were (or tried to be) president, with a particular focus on those times at which they were (or were trying to be) president. Any talk of the dignity of the presidency can be paraphrased as talk of the dignity that we both attribute to and expect of anyone who is president, because they are president. Instead of the function and purpose of the presidency, we can talk of the function and the purpose of someone being president. And so on. What should we conclude about the metaphysics of positions? I am not sure. Here are three views that hold some appeal to me. First, one might conclude that there are no such things. John Hyman notes that The phrases ‘take a bath’ and ‘meet one’s death’ are examples of a fairly common construction in English, which uses a verb with a so-called eventive object. One

    

255

can take a bath and have a chat; one can make a choice or a promise; and so on. . . . The construction . . . is merely a syntactic expedient, and the verb does not express a genuine relation. (Hyman 2015, 56)

To take a bath just is to bathe; to meet one’s death just is to die, and so on. We might try to take the same line on the verb phrases that we use to ‘relate’ us to our ‘positions’; that is, to talk of winning, getting, entering, having, holding, filling, occupying, inhabiting, keeping, leaving, and exiting an office or position. These verbs have the same whiff of semantic vacuity as ‘take’, ‘meet’, ‘have’, and ‘make’ in Hyman’s examples. And we can paraphrase them away: to quit the presidency just is to cease to be president etc. A bit of semantic evidence that may support this view is this. . . . when Socrates took a bath, the taking of it was the bath he took; and when he met his death, their meeting was his death . . . (Hyman 2015, 56)

Call those verbs that are syntactic expedients and which do not express genuine relations light verbs. And call a predicate that we have recourse to if we dispense with one made up of a light verb plus object, a basic predicate. ‘Bathed’ and ‘died’ are basic with respect to ‘took a bath’, ‘met his death’. Likewise, ‘is president’ is basic with respect to ‘holds the presidency’. Hyman’s observation is this: nominalizations of his light verb phrases (‘the taking of the bath’) and nominalizations of the corresponding basic predicates (‘the bath’) are co-referential. Likewise: when Biden won the presidency, the winning of it was his becoming president; and now that he holds that office, the holding of it is his being president.

A second view is that positions just are artefacts of, or abstractions from, our interest in them, such that a position exists only if we have taken sufficient interest in it (there is a broad brush parallel here with intuitionism, or, more broadly, constructivism in the philosophy of mathematics). This would explain why we have a principle of identity for things like the U.S. presidency or the Waynflete Chair of Metaphysical Philosophy but none for more lowly positions. I was one of four new departmental hires when I was first employed as a lecturer. As I recall, two people had recently left the department. Did I get a newly created position, or one of their positions? Might one of the old positions have been divided into two, of which I got one? Or did those positions go out of existence to be replaced by four new ones? We might sometimes talk as if there are answers to such questions, for, say, budgeting purposes. But really, they look like non-questions. The concept of a position does not provide any clear criterion as to how to answer them. But the positions in which we take sufficient interest do appear to have a principle of identity. No entity without identity.

256



A third view would be that positions exist as absences or gaps, in something like the way that argument places ‘exist’ in concepts, properties, and relations. Katherine Ritchie (2020) and Kit Fine (2020) offer views that suggest something of the sort. On the view as I would express it, a corporate entity has a form or structure that is unsaturated or gappy; and human beings fill or occupy the gaps. We may think of the corporate entity itself, then, as comprising the human beings in their positions: it is the completion of its incomplete structure, by those human beings. Whichever of these three views is right, or nearly right, and even if none of them is anywhere near right, positions are not the sorts of things that act or which are subject to requirements. To sum up, positions do not fulfil their emancipatory promise. I conclude with a discussion of how the idea of a job’s requirements might oppress one, and how to release oneself from its grip.

6. Some Sources of, and Emancipatory Strategies for, Workplace Alienation and Anxiety Recall Sartre’s café waiter: His movements are animated and intent, a bit too precise, a bit too quick; he approaches the customers with a bit too much animation; he leans forward a bit too attentively, his voice and his eyes expressing an interest in the customer’s order that is a bit too solicitous. Finally, here he is, on his way back . . . carrying his tray with the recklessness characteristic of a tightrope walker, holding it in a constantly unstable and constantly disrupted equilibrium, which he constantly restores with a light movement of his arm and hand. His behaviour throughout strikes us as an act. (Sartre 2018, 102–103)¹⁷

Initially, it seems that Sartre’s beef is that this guy is not for real. He is trying to ‘imitate [an] automaton’ and ‘concentrates on his successive movements as if they were mechanisms’. ‘Indeed, his facial expression and even his voice seem to be mechanical.’ Sartre notes that the waiter ‘cannot immediately be a café waiter in the sense in which this inkwell is an inkwell, in which the glass is a glass’. As we have put it here, waiter can never be what someone fundamentally is, as inkwell is what an inkwell fundamentally is. Nor is waiting what anyone is for, as the containment of ink is what an inkwell is for. But it is as if the waiter has mistaken his job, or perhaps the function of his job (position), for his function. By acting like a robot waiter, he is trying to be a functional artefact. The thought that one’s job makes one, or requires one to be, a functional object is one source of workplace alienation and anxiety. One route to it is ¹⁷ Intriguingly, Sartre alludes to ‘the duties of my station’ (2018, 104).

    

257

over-investment. By identifying too strongly with one’s position and/or occupation, hoping to find one’s true self in it, one risks turning oneself into a tragi-comical simulation of an artefact, for whom the mask eats into the face. As Sartre continues to watch, he has a new thought. The man is ‘playing at being a café waiter’ and that ‘[he] plays with his condition in order to actualize it’. We might read this as neo-Hobbesian social ontology. By personation, by acting as, say, a university, a human being can see to it that this university acts, confers a degree for example. Likewise, one might think, by personating or presenting himself as (‘playing at being’) a café waiter, a human can bring a waiter into existence (‘actualize it’). On this way of understanding things, the waiter is distinct from the man (‘if I represent him . . . I am not him’), just as the human at the ceremony is not the university. Consider: He knows full well what it ‘means’: the obligation to get up at five o’clock, to sweep the floor of the premises before the rooms are opened, to get the coffee machine going, etc. He knows the rights that come with it: the right to a tip, trade union rights, etc. But all these concepts and judgements refer to something transcendent. These are abstract possibilities, rights and duties, accorded to a ‘legal subject’. And it is precisely this person that I have to be and that I am not. (2018, 104)

There is the waiter, then, an abstract subject of rights and duties, on the one hand, which ‘I am not’, but which, by some contingency, ‘I have to be’.¹⁸ It is the waiter, not the man that is fundamentally a thing with a function, with associated rights and duties. The thought that one is distinct from one’s position, but, as its occupant, somehow contained, trapped even, within it is a second source of workplace alienation and anxiety. (For some it may follow dialectically from the rejection of the previous thought.) One is there, hidden and unobserved, behind the mask; a ghost in a socially constructed machine, driven by its purposes. Or, to change the analogy, one is like the matter—the aluminium, say—that makes up the functional part of the bicycle, formless in itself but given form and purpose by one’s occupancy of an object with a function. This is an unhappily detached and depersonalized (even psychotic) thought. There is another reading of Sartre’s talk of ‘playing at being a waiter’, or perhaps another strand in his discussion, according to which the waiter is playing in order

¹⁸ Sartre’s translator Richmond says (lx) that avoir à être (to have to be) ‘is used by Sartre to indicate the future-orientated, dynamic and responsible aspects of the for-itself: rather than simply “being” something (e.g. myself ) or some way, I have it to be. Although it can be quite simply translated into English as “to have to be”, the reader needs to be careful in some instances not to read the phrase in the sense that involves the idea of obligation. If the for-itself “has X to be” it is not obliged to be X, but chooses itself as being or aspiring to be X.’ So while a job may bring obligations, one chooses to take or to keep a job, one is not duty-bound to do either.

258



to fulfil a self-presentational requirement. One might think of this too as a further step in a dialectic. He is playing at being a café waiter . . . This obligation is imposed in the same way on all shopkeepers: their condition is entirely ceremonial, and the public demands them to actualize it as a ceremony; there is the dance of the grocer, the tailor, the auctioneer, through which they try to persuade their customers that they are nothing more than a grocer, an auctioneer, a tailor. A grocer who dreams is offensive to the buyer, because he is no longer completely a grocer. Etiquette requires him to contain himself in his grocer’s function . . . (2018, 103)

The thought that one is required to present oneself as something one is not, a functional object, is a third source of workplace alienation and anxiety. But there is an emancipatory reading of Sartre’s talk of play. One can read it as psychoanalysis (‘The child plays with his body to explore it, to take stock of it; the café waiter plays with his condition in order to actualize it’). The waiter has a conception of a ‘waiter’ that is a thing for waiting, as a knife is a thing for cutting, a waiter for whom the kind waiter is what he is. He knows that he is not this thing, but also that he might appear to others as this thing. Perhaps it is the very conception we were just discussing, on which the waiter exists because the waiter personates him. And he plays with this conception. We might think of this play as a strategy for alleviating alienation and anxiety, by exploring what Goffman (1959) was to call ‘role distance’. One can toy with a question such as Where do I stop, and where does he start? And as ‘play is a type of research and investigation’, such play can help one to find a non-anxious way to think about difficult issues. In addition, by ‘amusing himself ’, perhaps by fulfilling the demands of the job to an exaggerated and absurd extent, the waiter can ironize his condition. He can both meet the expectations, indeed exceed them, while holding on to a sense of who he is (the performer, not the role). This can perhaps go too far; a playful attitude can become a flippant one. If one starts to think that it is a joke that one has requirements by virtue of one’s position, this may lead to problems. Von Wright has this to say about the goodness of artefacts (‘instrumental goodness’): An attribution of instrumental goodness of its kind to some thing is . . . secondary in the sense that it logically presupposes a judgment of goodness for some purpose. (von Wright 1963, ch. II. sec. 2)

So, for example, if x is a good paring knife this is because x is a paring knife and is well suited to cutting foodstuffs. There is another way in which the judgement that x is a good paring knife is ‘secondary’. It is of secondary importance. What one cares about, if one has the purpose of cutting foodstuffs, is whether one can get

    

259

hold of something that serves that purpose. Whether it is of a kind that is essentially associated with that purpose, as the kind paring knife is, is of secondary concern. (For von Wright this is largely just a morphological matter.) Von Wright has this to say about ‘the goodness of ability or capacity or skill’ (‘technical goodness’): The good K is a K who is good at the proper activity of Ks . . . an attribution of technical goodness of its kind to some being is a secondary valuation. Its basis the primary valuation is a judgment to the effect that this being is good at something. (von Wright, 1963, ch. II. sec. 9. I have elided a paragraph break)

So, for example, if x is a good teacher this is because x is a teacher and good at teaching. There is another way in which the judgement that x is a good teacher is a ‘secondary valuation’. Whether or not someone is good at teaching is more important than whether they are of a kind that is essentially associated with that activity, as the kind teacher is. (For von Wright, this is largely just a matter of their credentials; I’d add that it is a matter of whether they habitually teach.) Another emancipatory strategy is to refuse to move from primary to secondary evaluations. If I peer review your class, it is not you, nor even you relative to the kind professor, that I am primarily evaluating, it is your teaching. I try to think of my work as a lecturer like this: what contributes, or should contribute, to my university fulfilling its purpose, is neither me nor my position; it is my teaching (research etc.), the habitual activity by virtue of which—together with my credentials—I am a lecturer. We need, not a dualism of positions and occupants, but one of ϕ-ers and ϕ-ing. A final route to emancipation is found in a famous passage of Marx’s in The German Ideology: . . . in communist society, where nobody has one exclusive sphere of activity but each can become accomplished in any branch he wishes, society regulates the general production and thus makes it possible for me to do one thing today and another tomorrow, to hunt in the morning, fish in the afternoon, rear cattle in the evening, criticize after dinner, just as I have a mind, without ever becoming hunter, fisherman, shepherd or critic.¹⁹

We can think of what Marx envisages as a sweeping away of all secondary valuations; a world in which we can evaluate teaching but not teachers, for there are none. One thing that would be welcome about this—whether this be in ¹⁹ I am grateful to Gabriel Wollner for alerting me to the brief discussion of this passage in (Cohen 2000, ch. V, sec. 7), from where I take the Marx quotation (132). Cohen takes Marx’s point primarily to be, not about the value of doing what one wants, or doing a number of different sorts of things, but the value of human relations not mediated by roles and institutions.

260



communist society or a gig economy—is that of never having to trudge through the moments of the fraught Sartrean dialectic that I have itemized.²⁰

References Cohen, Gerald Allan. 2000. Karl Marx’s Theory of History (2nd edition). Oxford: Oxford University Press. Fine, Kit. 2020. ‘The Identity of Social Groups’. Metaphysics 3 (1): 81–91. Fix, Jeremy David. 2019. ‘Two Sorts of Constitutivism’. Analytic Philosophy 62 (1): 1–20. Geach, Peter Thomas. 1956. ‘Good and Evil’. Analysis 17 (2): 33–42. Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Anchor Books. Graeber, David. 2018. Bullshit Jobs: A Theory. London: Penguin. Harman, Gilbert. 1986. Change in View. Cambridge. Mass.: MIT Press. Hyman, John. 2015. Action, Knowledge, and Will. Oxford: Oxford University Press. Kraut, Richard. 2007. What is Good and Why. Cambridge, Mass.: Harvard University Press. Kraut, Richard. 2011. Against Absolute Goodness. Oxford: Oxford University Press. Kripke, Saul A. 1980. Naming and Necessity. Cambridge, Mass.: Harvard University Press. Quine, Willard Van Orman. 1976. ‘Three Grades of Modal Involvement’. In The Ways of Paradox (2nd edition). Cambridge, Mass.: Harvard University Press, pp. 158–176. Ritchie, Katherine. 2020. ‘Social Structures and the Ontology of Social Groups’. Philosophy and Phenomenological Research 100 (2): 402–424. Russell, Bertrand. 1905. ‘On Denoting’. Mind 14 (56): 479–493. Sartre, Jean-Paul. 2018. Being and Nothingness. Trans. S. Richmond. London and New York: Routledge. Scanlon, Thomas Michael. 1998. What We Owe to Each Other. Cambridge, Mass.: Harvard University Press. Thomson, Judith Jarvis. 2008. Normativity. Chicago and La Salle, Ill.: Open Court. von Wright, Georg Henrik. 1963. The Varieties of Goodness. London: Routledge and Kegan Paul. . Wiggins, David. 2016. ‘Substance’. In his Continuants. Oxford: Oxford University Press, pp. 41–71. ²⁰ Thanks to the directors of the AHRC-funded Role Ethics Network for inviting this contribution, and to all network members for inspiring discussions. Thanks to the editors for their infinite patience and insightful comments on the penultimate draft. Special thanks to Suddhasatwa Guharoy for his comments on this draft.

PART FOUR

WELLBEING, SELFHOOD, AND ROLES

13 Three Relations Between Roles and the Good Samuel Clark

1. My Interests, Plan, and Method I’ll begin with where I am coming from, by comparison with Michael Hardimon’s title for his classic paper ‘Role Obligations’ (Hardimon 1994). First, roles: before joining the research network which eventually gave rise to this volume, I had not thought about roles as such. I had thought and written about what I now realize can usefully be collected together as particular roles: parent, child, soldier, employee, craftsman (Clark 2010, 2012, 2013, 2017). This chapter is therefore partly exploration of territory which is new to me, partly redeployment of work towards my book Good Lives (Clark 2021), all in service of my own concerns about human selfhood, reflexivity, good, and practical rationality, as they relate to roles. Second, obligations: I agree with Bernard Williams (2005) in thinking that both contemporary Western liberal morality and contemporary anglophone moral philosophy are over-focused on obligations, which are only one part of the ethical landscape, and shouldn’t be allowed to take over more of it. I am interested instead in the good life, that is, what it is for someone’s life to go well for her, also known as well-being, welfare, prudential value, etc. So, on the one hand, I join Hardimon and this volume in their focus on roles; but on the other, I depart from his almost complete attention to the obligations associated with roles. My plan is to explore relations between roles and the human good by investigating three ways they might fit together. Roles could be: 1. Tools for self-shaping towards a good understood as desire-fulfilment, and thus as independent of the means used to pursue it. 2. Parts of good-making practices which transform individuals by creating goods and initiating those individuals into them. 3. Methods for self-discovery which begin and direct the achievement of the good understood as self-realization.

Samuel Clark, Three Relations Between Roles and the Good In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Samuel Clark 2023. DOI: 10.1093/oso/9780192843562.003.0013

264



My aim is to show that roles are methods of self-discovery in some real cases, and that this is evidence that the good is self-realization (not desire-fulfilment, not just created by institutions, and not wholly up to us). A remark on method: lots of my work, including this, is about and draws on autobiographies—here, specifically martial autobiographies, and even more specifically James Salter’s autobiography Burning the Days (Salter 2014), backed up and contrasted by autobiographies by Tim O’Brien (2003) and Siegfried Sassoon (1937). I am not going to say much in defence of that method in general—I am just going to bring my particular texts and examples in as I need them, do what I want to do with them, and hope that they show their value in practice. I do want to say, though, that an underappreciated part of the work of philosophy is not argument in the sense of movement from premises to conclusion, but showing: displaying a distinctive way of grasping some phenomena as vivid and coherent. Autobiographies are good at doing that for the particular phenomena I am interested in, and the first-personal character of this chapter draws on that strength.

2. Relation 1: Tools for Self-shaping On this first account, roles are social technology for shaping ourselves towards a good life understood as the fulfilment of desires which are independent of those roles. I will unpack that thought by example (not yet a martial autobiography example, for reasons which will become clearer). Cycling club: I want to improve my general fitness, get some fresh air, and meet some new people. So I join my local cycling club, Lancaster and South Lakes CTC.¹

Taking on this role of CTC member recruits a variety of causally effective objects and processes to help fulfil my desires: infrastructure like a club-house and a tool library; collections of information and expertise; mechanisms of habit-formation and routine such as regular Thursday rides; the motivational forces of friendly competition and public commitment; and so on. If taking on this role works, it helps me change myself in accordance with my reflexive desires—I get fitter, as I wanted, by recruiting the world’s and other people’s causal powers to act on me. But taking on the role doesn’t typically change my desires: they’re given prior to and independently of the cycling club; the cycling club is just a contingent means towards fulfilling them. I could instead join park runs, or just go cycling on my own. The cycling club is a tool, to be judged by its effectiveness compared to alternative tools.

¹ A branch of the Cyclists’ Touring Club, a national charitable organization with 68,000 members, founded in 1878, which aims to promote and support cycling and cyclists. I give this detail just to show that the CTC is definitely an institution which supports roles.

      

265

Making explicit the assumed account of the good and my action towards it here: good is desire-fulfilment; rational action, including taking on roles, is instrumental. Now consider another example: Army: I want to improve my general fitness, get some fresh air, and meet some new people. So, I join the army.

Here’s James Salter arriving at West Point in July 1942: It was the hard school, the forge. To enter you passed, that first day, into an inferno. Demands, many of them incomprehensible, rained down. Always at rigid attention, hair freshly cropped, chin withdrawn and trembling, barked at by unseen voices, we stood or ran like insects from one place to another, two or three times to the Cadet Store, returning with piles of clothing and equipment. Some had the courage to quit immediately, others slowly failed. Someone’s roommate, on the third trip to the store, hadn’t come back but had simply gone on and out the gate a mile away. That afternoon we were formed up in new uniforms and marched to Trophy Point to be sworn in. (Salter 2014, 47)

This is one version of a familiar account: military basic training is typically described autobiographically in terms like ‘forge’, ‘inferno’, ‘destruct-testing’. This is recognizable to anyone who has read a martial autobiography, or seen Full Metal Jacket, or An Officer and a Gentleman, or Tigerland, or any of a number of other films in their genre.

3. Transformative Roles Why is it funny or odd to think of pursuing my desire to get some fresh air etc. by joining the army? Partly because it’s disproportionate to my weak-sounding desires: it’s a hammer to crack a nut. But it also sounds perverse because of a significant distinction between non-transformative and transformative roles. Here’s an example of a non-transformative role: I have played Father Christmas at my children’s school Christmas fair (this is actually true, unlike cycling club or army). I put on the costume to pursue some desires I happened to have, to amuse children and to support my wife in her role as PTA chair; I took the costume off again when they were fulfilled; the red suit went back in the PTA cupboard; I was unchanged. But compare the role of soldier: to join the army isn’t just a way of trying to fulfil desires one already has, it’s opening oneself to being remade—to being turned into someone with different desires, and therefore with a different good. And that’s why it’s odd to use the army, unlike using the cycling club, as a merely contingent tool for fulfilling prior and independent desires. An objector interrupts: couldn’t joining the cycling club be transformative? Couldn’t playing Father Christmas be transformative at the right moment?

266



Perhaps it’s when I realize that I really want to work with children, and change careers. I reply: yes, that’s true, because being transformative is a relation between a role and a particular person, not an intrinsic property of the role. Playing Father Christmas was not in fact transformative-for-me, but it might have been for someone else, or for me at another time. That said, some roles are typically transformative, that is, transformative-in-fact for a wide variety of people, and soldier is one of them, by design. Indeed, soldier is an important sub-type of a class of typically transformative roles: ascetic roles. Being a soldier is a form of (what I label) regular asceticism: one joins a totalizing rule-governed community, in which one lives in public, without individual property, and works collectively for some good which the community shares, maintains, and expresses. The other common example of regular asceticism is being a monk or a nun, and this isn’t a coincidence: early Christian monasticism, at least, was modelled on soldiering. One of the earliest monasteries, as distinguished from loose temporary groupings of hermits, was founded in the Syrian desert round the fourth century  by St Pachomius. He had been a Roman legionary, and knew how to organize people to live in a hostile environment: an army camp under a formal, codified rule and roster of duties, enforced by whoever enacts the role of ‘abbot’ (Rousseau 1999). I will use regular ascetic roles as a continuing example throughout the rest of this chapter. The existence of transformative roles—including, but not limited to, regular ascetic roles—is a problem for relation 1. At least, it limits that relation’s application to a subclass of roles, where I wanted a general account. My relation 2 offers a solution to that problem.

4. Relation 2: Good-making Practices At least about this regular ascetic kind of role, it’s tempting to say that they are parts of collective practices which create and sustain the good for those within them.² That means two things. First, these roles give individuals who take them on goods they wouldn’t otherwise have: to be initiated into a transformative role is to have one’s goods radically changed. Here is Salter again, later in his West Point career: I was undergoing a conversion, from a self divided and consciously inferior . . . to one that was unified and . . . right . . . There were images of the struggle in the air on every side, the fighter pilots back from missions deep into Europe, rendezvous

² This account clearly owes a great deal to Alasdair MacIntyre (especially 2007), but I do not intend to be doing any interpretation of MacIntyre or attributing any views to him here. I am just borrowing some ideas for my own purposes.

      

267

times still written in ink on the backs of their hands, gunners with shawls of bullets over their shoulders, grinning and risky[.] I saw them, I saw myself, in the rattle and thunder of takeoff, the world of warm cots, cigarettes, stand-downs, everything that had mattered falling away . . . More than anything I felt the desire to be rid of the undistinguished past, to belong to nothing and to no one beyond the war. At the same time I longed for the opposite, country, family, God, perhaps not in that order. In death I would have them or be done with the need; I would be at last the other I yearned to be. (Salter 2014, 70–71)

This constitutive account further means, second, that the institution creates those goods: these things are only good because the institution and its roles exist. A regular ascetic transformative institution isn’t just a mechanism for changing people’s good; it’s a creator of goods with an associated mechanism for initiating people into them. Other institutions have the same properties: they turn what would otherwise be bare states and possibilities and movements into goods, and into actions aiming at them, by placing them in a structure which makes sense of them as parts of a whole. Rational action, then, is not merely instrumental, but historical: what makes a movement an action at all, and therefore subject to standards of rationality and irrationality, is its place in a history which explains it. So, on this account, the institution is necessary to the good into which the individual who takes on one of its roles is initiated. It isn’t just a means to recognizing and having some good, it constitutes that good and rational action towards it.

5. Experiences of Recognition and Resistance Even if relation 2 solves the problem of transformative roles for relation 1, it has its own problems. Consider Salter’s conversion again. In his first year at West Point he is nearly the worst cadet in the school, continually bumping up against and ineffectually resisting the institution. He experiences it as completely wrong for him, an ugly, oppressive, shaming intrusion. Plenty of other recruits have had the same experience—Tim O’Brien (2003) describes it vividly in his If I Die in a Combat Zone, for example. Unlike O’Brien, Salter does eventually find that West Point fits him, or perhaps he finds a way to make himself fit it, by adopting a Romantic ideal of the company commander which he finds in a German military science text: This youthful but experienced figure was nothing less than a living example to each of his men. Alone, half obscured by those he commanded, similar to them but without their faults, self-disciplined, modest, cheerful, he was at the same time both master and servant, each of admirable character. His real authority was not based on shoulder straps or rank but on a model life which granted the right to demand anything from others. (Salter 2014, 68)

268



Siegfried Sassoon, for just one other example, had a similar ideal at the beginning of the First World War (Sassoon 1937). For Salter, this figure isn’t just a hero to admire or to emulate, as perhaps it was for Sassoon. It’s a delighted selfrecognition: I knew this hypothetical figure. I had seen him as a schoolboy, latent among the sixth formers, and at times had caught a glimpse of him at West Point. Stroke by stroke, the description of him was like a portrait emerging. I was almost afraid to recognize the face. In it was no self-importance; that had been thrown away, we are beyond that, stripped of it. When I read that among the desired traits of the leader was a sense of humor that marked a balanced and indomitable outlook, when I realized that every quality was one in which I instinctively had faith, I felt an overwhelming happiness, like seeing a card you cannot believe you are lucky enough to have drawn, at this moment, in this game. I did not dare to believe it but I imagined, I thought, I somehow dreamed, the face was my own. (Salter 2014, 69)

This is Salter’s individual response as a singular human being, but it maps to some common human possibilities of response to regular ascetic and other transformation: experiences of deep unfittingness or of deep fittingness, or of both sequentially. I want to say: these are experiences of discovery, of selfrevelation—literally, revealing something about the subject of initiation to themselves through how they react to that initiation. Sometimes, regular ascetic institutions are explicitly recruited as means of self-revelation—Sassoon, for example, joined up to find out whether he was just the drone he appeared to be. But even when that’s not so—Salter joined up just to please his father—that is one of the things these roles do to us, whether we want them to or not. They’re a method of self-discovery by probing or testing for the fittingness or unfittingness response.

6. Relation 3: Self-discovery and Self-realization On this third relation of roles and the good, some roles are a method of selfdiscovery, that is a mechanism for gaining self-knowledge. Such a role can help the individual who tries it on to discover her unchosen, seedlike, initially opaque self, and thereby to discover her particular good, which is that self ’s development from latency to completion and expression: her self-realization. This is the view of the self, its good, and our knowledge—self-knowledge— about them that I defend in Good Lives (Clark 2021). My three main ideas here are self-knowledge, the self, and the human good, and I’ll now consider them individually.

      

269

Self-knowledge Recent philosophical debate under the heading ‘self-knowledge’ has been about an epistemic question, the nature of my knowledge of my current mental states. Is it unmediated? Transparent? Infallible? Incorrigible? Is it like or unlike my knowledge of other people’s current mental states? In contrast, I’m interested in substantial self-knowledge (Cassam 2014): answers to questions like who am I? What am I? What’s my character—am I brave, selfish, creative, lazy? How would I react in a crisis or under threat? What is in me that I need to express? What do I really want and really believe? What would make my life a success for me? Why did I do what I have done? Did I make the right choices? Why has my life gone the way it has—is it down to me or down to the world? I think, with Quassim Cassam, that most people who aren’t professional philosophers are interested in this kind of self-knowledge—answers that would satisfy the ‘Delphic demand’ to know yourself—rather than the philosopher’s epistemic question. I claim, then, that trying on a transformative role like soldier is a way of pursuing substantial self-knowledge: to join up or be drafted is a way of finding out what you’re made of. An objector interrupts: what we’re dealing with here is story-telling, not selfdiscovery. It’s making something, not responding to what’s already there, so talk of self-knowledge rather than self-creation or self-invention is misplaced. I reply: the reported phenomenology in martial autobiographies, and in other autobiographies of transformation, is of inward-looking response to something unexpected and demanding, not of outward-looking free invention. This is a common phenomenology across otherwise different and singular people. We can also see it, for just one example, in Edmund Gosse’s Father and Son (Gosse 1983). Objector: it’s not a story then, it’s a story in Salter or Gosse or whoever, telling it in their autobiographies. Who knows what it was then, if anything, but the later autobiographical telling is a retrospective invention along conventional, generic lines which also appear in plenty of novels and films, and which have a history and a cultural location. Reply: I take the objector’s view to be that the self is not an object, ‘out there’ in the world or deep ‘inside’ me or you, to be uncovered and represented in autobiographical narrative. She claims that the self is a self-interpretation in autobiographical form, and that the object of self-knowledge is created by narrating oneself into being, not revealed. Views of this broad type are held, for example, by Charles Taylor (1989), Marya Schechtman (1996), Mark Freeman (1993), and Jerome Bruner (2004), and in the existentialist tradition. On this picture, self-knowledge isn’t like other kinds of knowledge, or at least it’s not like our ordinary picture of knowledge: it’s not a representation of, or a deferential response to, or a way of catching or grasping, anything there in advance of investigation, which we might misrepresent or misinterpret. This picture will be

270



made clearer by two analogies. First, an error theory of self-knowledge. The Delphic demand asks us to uncover something which is just not there to be uncovered: a real self. To find the demand compelling—as so many of us do—is to make a mistake. Where an error theory of moral language claims that we talk about and search for a non-existent moral reality, independent of our moral practices, this error theory of self-knowledge claims that we talk about and search for a non-existent self, independent of our practices of self-interpretation and self-representation. Second, a response to scepticism with a Kantian flavour: if self-knowledge is taken as grasping something independently there, then the road to scepticism is left open, because we might be permanently opaque to ourselves, or such that our self-representations are unavoidably distorted beyond the possibility of knowing ourselves. Maybe our powers of understanding cannot reach to things as they are here. The response is then to argue that the objects of knowledge are constructed by those powers. What we know about ourselves is made not found. The transcendental argument could go: self-knowledge cannot be impossible; if the self were independent of our powers of understanding it might be; so, the self is not independent of, but is constructed by, those powers. I have two replies to this objection, distinguished by how the objector conceives the self. First, self as persona. Schechtman connects self-narration to our construction of a social self which allows us to live in our particular kind of society. We—me, and, I expect, you my readers—live in a commercial, literate, rule-of-law governed society, which understands some creatures, and some artificial things like universities, limited liability companies, and cycling clubs, as continuing entities with rights and duties. They are able to hold assets, to make contracts, to be held to account, to sue and be sued, to plan and invest for their future interests. That is, they are entities which have legal personality, a persona in the sense of the mask which an actor in classical Greek theatre wears in order to inhabit a dramatic role. So, my objector could mean that autobiographical narrative creates a persona. And this could be true: personae are perhaps sometimes or partly created by autobiographical narration, although they are also made by dialogue, by inhabiting institutional roles, by picking up local stereotypes and ready-mades, and by social assignment (as for example when an infant or a corporate body is a legal person for ownership purposes). But whether or not that is true, the persona isn’t the self we wanted self-knowledge about, the self for which life goes well or badly, because it makes sense to worry—and sometimes it is true—that my masks hide or distort the real me, and thereby do me harm. Self-knowledge then requires looking behind the persona, and self-realization requires taking it off. We can see examples of this especially in slave narratives (e.g. Douglass 1986) and in some women’s autobiographies (e.g. Lessing 1995). However, this doesn’t completely deal with the objection. Some of my objectors—Bruner, for example—should be satisfied with it, because by ‘self ’

      

271

they mean ‘self-concept’: how I represent myself to myself or think about myself. But other narrativists mean something deeper than ‘my persona is my autobiographical self-interpretation’. For Charles Taylor in particular, the consequential questions about living a human life well only come into play at all when we interpret ourselves into existence. There is no self to which standards of life’s going well or badly apply, prior to reflection on the self. I therefore need a second reply: pluralist realism about self-knowledge. I can accept for the sake of argument that self-interpretation creates the object of one particular kind of self-knowledge, a self-interpretation. Perhaps it is knowledge of what my life means to me, or of how I evaluate myself. But I want to defend the genuine experience of self-discovery, which reveals the self as unchosen, seedlike, and initially opaque. That is, as something there which we must uncover by active enquiry, not something created by interpretation. My account is pluralist about three things: about kinds of self-knowledge; about our means for gaining it; and about our resources for representing it. My account is realist in that it denies both the error theory and the Kantian account of self-interpretation. I argue that there is a self there prior to and independent of investigation, and that we can—with work—reveal it. My argument for that account is: the more distinct methods of self-discovery we can identify, the less plausible the single, totalizing narrative self-interpretation account will look. The Delphic demand aims at self-knowledge of many different kinds, and we can pursue it using many different means. I suggest, first, that a narrative self-interpretation is, at best, one among many kinds of self-knowledge, and not the most interesting to us. We can also gain knowledge of ourselves as unchosen, initially opaque things with distinctive potentials to be developed, expressed, and more or less skilfully handled. Second, that narrative self-interpretation is, at best, one of many ways we might gain self-knowledge, and not the most effective. In my book I argue for selfdiscovery by pleasure and by solitary asceticism; here I focus on self-discovery by initiation into a regular ascetic role. The force of that depends, of course, on how compelling I can make each of these distinct kinds and methods of self-knowledge, and I cannot do anything more towards that now. So, I leave this as a plan and promise, and move on.

Self ‘Self ’ is said in many different ways: it can mean separateness, immediate consciousness, mentality, essence, personhood, agency, the owner of one’s actions and states, self-conception, social identity, etc. I use ‘self ’ to mean that thing, whatever it is, which is the subject of a human life, for which that life goes well or badly, and which is the thing known by the person with self-knowledge.

272



I claim that the experience of self-discovery through taking on regular ascetic roles shows that we should understand that self as unchosen, seedlike, and initially opaque to itself. The self is unchosen in that we experience it as discovered not invented—and that’s an accurate experience, because we’re not self-creating. It is seedlike in that discovery can be a kind of waking, a germination: a beginning of growth from mere potential or capacity into expression and action. Or it can be a recognition of unfittingness—that these are not the circumstances in which I can self-realize, as O’Brien discovers. It is initially opaque in that self-discovery is work. It requires a gradual process of uncovering and expression, and to know oneself is an achievement which not everyone manages, and which, perhaps, no one completes. Objector: this is a metaphysically very costly account of the self. It seems to require some unified, unchanging, non-physical thing standing behind and owning all of the particular actions, experiences, and states of a human being. That is, it seems to be an appeal to a soul. Reply: it would be too costly, if I were claiming that the self is like that (i.e. if I had a ‘further fact’ view about the self, in Parfit’s (1984) term). But I am not: all I am claiming is that the self has these three features, and I can be agnostic about how they’re realized. I can be perfectly happy with a reductionist account of the self, on which there is nothing more to it than the various physical and/or mental parts which make it up. I can be perfectly happy with the idea that a self just is a living, embodied human being. I do not need anything too metaphysically costly to claim what I am claiming.

The Human Good In a slogan, ‘the good is the good of a self ’: what’s good for some thing depends on what it is, so we can move from an account of the self to an account of the good. For example, L. W. Sumner makes this move in his Welfare, Happiness, and Ethics (Sumner 1996): the self is essentially mental; so the good must be some mental state; and the obvious candidate is happiness. I make a parallel move, sharing Sumner’s structure but replacing his premise: the self is essentially seedlike, so the good must be its growth and flowering; that is, one’s self-realization. So, the relation of roles to the good is that taking on a transformative role is a method of self-discovery that makes deliberate movement towards one’s own good possible, because I cannot plan my route to a destination before knowing what that destination is. Roles are not just tools for achievement of what’s already out in the open, and not just creators of good. The good we can discover by taking on roles, and paying attention to our reactions to them, is self-realization: your life goes well for you when, and in the ways that, you develop and express your innate potential. That potential will

      

273

overlap with the potential of other human beings, just because we’re so closely related to one another, but it will also have distinctive features which are individual to you, and shared with only some—or at the limit, no—other humans.

7. Rational Action and Transformation What does this mean for rational action? I’ll address that question by comparison with L. A. Paul’s idea of transformative experience (Paul 2014), and with the pictures of practical rationality implied by my first two relations between roles and the good. ‘Transformative experience’ is a technical term for a familiar phenomenon. Consider such experiences as becoming a parent, fighting in a war, or becoming chronically ill. They have two things in common: First, they are epistemically transformative. Living through such a transformative experience provides a kind of knowledge only available by that first-personal acquaintance. Only a parent knows what it is like for her to have a child; only those who have fought know what combat is like for them (the claim is not that nothing can be known third-personally about these and other experiences, it is that not everything can be known that way). Second, they are personally transformative. I am someone else after becoming a parent; the soldier is someone else after her baptism of fire; the ill person is not the same as her former, healthy self. We are not ‘someone else’ in the sense of having a new body, a new history, or a new legal personhood: Sassoon still had his private income after his war; I still had my limp after becoming a parent. But the transformed person at least has new personality traits and projects, and perhaps further has a new self: they become such that they no longer fully identify with their earlier self. I no longer fully identify with my pre-parenthood self, but have a more complex, alienated, sometimes uncomfortable relation to him, of pride, amused affection, awkward recognition, incomprehension, embarrassment, shame. That is, my life as a person is disunified over time by the transformative experiences I live through. This is a repeated theme of, especially, martial autobiographies. For one example, Siegfried Sassoon’s Memoirs of a Fox-Hunting Man and Memoirs of a Military Officer (in Sassoon 1937) explain his life as a transition from innocence to experience, where the innocent pre-war fox-hunting Sassoon is an ironically presented ancestor to the author Sassoon who has lived through fighting in the trenches. Sassoon knows, of course, that he is the same human being as this naive young person he recalls; but he is not fully identified with that person. For Paul, the fact of transformative experience in human lives has deep consequences for practical rationality, and in particular for planning for the future. Each of us wants to make rational decisions about our own futures, where ‘rational’ means intelligent, good, correct, successful rather than unintelligent, bad, mistaken,

274



unsuccessful. But how? One popular and much-studied answer is: maximize expected utility. The utility (for you) of some state of the world is its value (for you). The expected utility of some possible future state is its percentage chance of occurring multiplied by its value. The expected utility of an action you might take is the sum of the expected utilities of all of its possible outcomes. The rational action to take is then that action, of those available to you, which has the highest expected utility. So, for example, the expected utility of buying a lottery ticket is: (chance of winning x (prize – cost of ticket)) – (chance of losing x cost of ticket) Given that the chance of winning is very small, the answer to this sum is negative; ‘don’t buy a ticket’ has a higher expected utility; and that is therefore the rational action of the two. A standard criticism of this way of making decisions is that we usually do not know the relevant probabilities. Lotteries are very unusual in that we do. This is true, but we’re not completely in the dark in many cases, and decision theorists have developed sophisticated techniques for dealing with decisions under uncertainty (Steele and Stefánsson 2020). Paul’s distinct criticism is that we sometimes cannot know the other element of the expected utility calculation: you cannot know the value of outcomes for you in decisions about whether to undergo a transformative experience. Or at least, you cannot know that value if it depends, even partly, on what it will be like for you to have that experience, and on what you will then want or care about. This is for two reasons: first, transformative experiences are epistemically transformative, so you cannot know what it will be like to have such an experience until it’s too late to decide not to have it. Second, transformative experiences are personally transformative, so you will be someone else afterwards, with different responses, desires, and cares. So you cannot, even in principle, work out whether undergoing a transformative experience will maximize your expected utility. Each of us wants to make rational decisions about our own futures—whether or not to have children, whether or not to join up, how to respond to becoming chronically ill—but we do not know how. This fact of transformative experience, and the fact of transformative roles which I have discussed here, both push in the same direction: against an instrumental account of practical rationality. We cannot plan for our own futures by taking our desires as given and working out what actions would best satisfy them, because some of the actions we could take, and some of the things which could happen to us, will change our desires in ways we cannot predict. They will change our selves. For roles in particular, the fact that some roles are transformative means that adopting such roles cannot just be a means to a preset goal. So, rational action is either not possible or more than just instrumental. My response to the transformation problem with relation 1 between roles and the good was to suggest relation 2: roles are parts of good-making practices which

      

275

transform individuals by creating goods and initiating those individuals into them. This is also a possible response to Paul’s criticism of instrumental rationality: perhaps we can plan for our futures by using existing roles to create our future goods. Soldier is typically a transformative role, which means that it would be a mistake to use it to pursue pre-existing desires and the good of satisfying them. But I could join up, not as a means of satisfying my current desires like cycling club, but as a way of giving myself a good by taking on a role in an institution which creates that good. That is, as a way of remaking myself into someone with that good. This new good isn’t dependent on what my desires now are, nor on what it would be like for me to experience that role, so my planning doesn’t fall to Paul’s problem. My response to relation 2 was to appeal to responses of fittingness and unfittingness provoked by taking on transformative roles, especially in Salter’s autobiography, but intending to generalize the point. Taking on a transformative role can be a method of self-discovery as well as a method of self-transformation. In trying such a role on, I wake features and potentials in myself that I didn’t know I had; reveal my unchosen, initially opaque, seedlike self; and thereby discover my individual good, which is my individual self-realization. Looking at transformative roles this way should also push us to rethink practical rationality as more substantive than the purely formal pictures I have associated with relations 1 and 2. Substantive accounts in general claim that some goals are intrinsically rational, in contrast with formal accounts, which claim that only means are rational or irrational, and that the goals they aim at are brute, arational givens, perhaps just desires. On this picture, as Hume says, it isn’t irrational to prefer my complete ruin to the least inconvenience to a stranger (Hume 1969, 463/book II part III section iii). One popular substantive account, the self-interest account, denies this: my own good is an intrinsically rational goal for me, and to fail to have it or to fail to act on it is in itself practically irrational. On my substantive account of practical rationality, self-discovery and the selfdevelopment it enables are not merely things we might want alongside our other desires: they are partly constitutive of practical rationality for us. It’s intrinsically rational to find out what I am and to develop it, not merely rational as a means to some independent good, nor merely as governing what means we should pick if we happen to have those goals. It’s always rational to discover and develop myself. This doesn’t mean that it’s not rational to do other things, but it does mean that we can criticize as irrational someone who doesn’t care about or pursue her own selfdiscovery and self-realization. It can therefore be rational to take on a transformative role: not only as a means to achieve a preset goal; not only as a way of finding out what I’ll be like and what goals I’ll have after my transformation—if I happen to have the second-order goal to find that out; and not only to give myself a new good constituted by the role in its institution. Transformative roles are ways of discovering and realizing ourselves, and to that extent taking them on is an instrinsic, not merely instrumental, requirement of practical rationality.

276



8. Conclusion On the picture I have developed here, what transformative roles do for us is offer mechanisms for self-realization via the possibility of self-discovery by finding a role individually fitting or unfitting. To try on the role of soldier is to probe myself for recognition of or resistance to it; to discover something about myself—a piece of substantial self-knowledge—through that response; and thereby to have an opportunity for self-realization. Perhaps being a soldier wakes something in me that needs to be expressed, and can be developed by staying a soldier, as happened to Salter; or perhaps it wakes resistance which reveals what I need to do instead, as happened to O’Brien. What being a soldier wakes in Salter is a certain kind of Romanticism: a need to be the passionate and doomed hero, without which his life wouldn’t have been a success for him, but which he’s self-aware enough about to satirize. He tells a story against himself of giving some disastrously terrible advice to a classmate out of that Romanticism, for example. This particular self-realization is the good for Salter, but not necessarily for anyone else, and is in tension—as elements of anyone’s good can be—with other equally real goods for Salter. I have explored three possible relations between roles and the good: they are tools for self-shaping towards an independent good understood as desirefulfilment; or they are parts of good-making practices which transform individuals by creating goods and initiating those individuals into them; or they are methods for self-discovery which begin and direct the achievement of good understood as self-realization. The point of that exploration was to show the importance of my relation 3 for our understanding of roles in human life, by showing that regular ascetic transformative roles like soldier are methods of self-discovery in some real cases; that this reveals the self as unchosen, seedlike, and initially opaque to itself; and that this is reason to think that the good is self-realization (not just desire-fulfilment, and not just created by institutions). The point was further to offer a substantive picture of practical rationality, opposed to purely formal accounts, on which it is intrinsically rational to pursue self-knowledge and self-realization.

References Bruner, Jerome. 2004. ‘Life as Narrative’. Social Research 71: 691–710. Cassam, Quassim. 2014. Self-Knowledge for Humans. Oxford: Oxford University Press. Clark, Samuel. 2010. ‘Love, Poetry, and the Good Life: Mill’s Autobiography and Perfectionist Ethics’. Inquiry 53: 565–578. Clark, Samuel. 2012. ‘Pleasure as Self-Discovery’. Ratio 25: 260–276.

      

277

Clark, Samuel. 2013. ‘Under the Mountain: Basic Training, Individuality, and Comradeship’. Res Publica 19: 67–79. Clark, Samuel. 2017. ‘Good Work’. Journal of Applied Philosophy 34: 61–73. Clark, Samuel. 2021. Good Lives: Autobiography, Self-Knowledge, Narrative, and Self-Realization. Oxford: Oxford University Press. Douglass, Frederick. 1986. Narrative of the Life of Frederick Douglass, an American Slave: Written by Himself. Houston A. Baker Jr. (ed.), London: Penguin. Freeman, Mark. 1993. Rewriting the Self: History, Memory, Narrative. London: Routledge. Gosse, Edmund. 1983. Father and Son. Peter Abbs (ed.), London: Penguin. Hardimon, Michael O. 1994. ‘Role Obligations’. Journal of Philosophy 91: 333–363. Hume, David. 1969 [1739]. A Treatise of Human Nature. Ernest C. Mossner (ed.), Harmondsworth: Penguin. Lessing, Doris. 1995. Under My Skin: Volume One of My Autobiography, to 1949. London: Flamingo. MacIntyre, Alasdair. 2007. After Virtue: An Essay in Moral Theory (3rd edn.). London: Bloomsbury. O’Brien, Tim. 2003. If I Die in a Combat Zone. London: Flamingo. Parfit, Derek. 1984. Reasons and Persons. Oxford: Clarendon Press. Paul, Laurie A. 2014. Transformative Experience. Oxford: Oxford University Press. Rousseau, Philip. 1999. Pachomius: The Making of a Community in Fourth-Century Egypt. Berkeley: University of California Press. Salter, James. 2014. Burning the Days: Recollection. London: Picador. Sassoon, Siegfried. 1937. The Complete Memoirs of George Sherston. London: Faber and Faber. Schechtman, Marya. 1996. The Constitution of Selves. Ithaca, NY: Cornell University Press. Steele, Katie and H. Orri Stefánsson. 2020. ‘Decision Theory’, The Stanford Encyclopedia of Philosophy. Edward N. Zalta (ed.), . Sumner, L. W. 1996. Welfare, Happiness, and Ethics. Oxford: Clarendon Press. Taylor, Charles. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge: Cambridge University Press. Williams, Bernard. 2005. Ethics and the Limits of Philosophy (with a commentary on the text by A. W. Moore). London: Routledge.

14 Participatory Wellbeing and Roles Alex Barber

1. Introduction Whatever else we say of them, the professional, familial, and other social roles we occupy have a mediating function: they shape the relationship holding between individuals and the larger groups to which they can belong. This is easiest to see when the group is an organization. One becomes part of an organization—one belongs to it, is a member of it—only by occupying one of the various roles within it, typically as an employee. But in addition to role occupancy conferring membership of the group, role performance is an individual member’s means of contributing to the agency of the group. This, once again, is clearest for professional roles within organizations: there is not much to the doings of a broadcasting organization, for example, beyond the coordinated performance, by role occupants, of the newsreader role, the head of estates role, and so forth. Similar points can be made regarding other groups and other roles, even if the situation is far less transparent. By becoming a friend one becomes part of a friendship group containing at least one other, and the friendship ceases to function—or at least becomes dormant—if neither party does any of the things that friends are meant to do. One belongs to the Roberts family only by having a familial role of some kind, by being a sister, father, daughter-in-law, or some such. And a family ceases to function—arguably ceases to exist save as a purely abstract entity—if its members fail in sufficient numbers to play their different parts, even though those parts are not specified in a set of agreed terms and conditions.¹ In the present chapter I draw on the mediating function of roles just outlined to analyse a specific phenomenon, what I call participatory wellbeing. This is the wellbeing that accrues to a person by virtue of their membership of and contribution to the workings of a group. Participatory wellbeing is something of a puzzle. While there is no denying that our level of wellbeing is sensitive to our participation in collective activity, it is surprisingly difficult to understand how that impact occurs. In this chapter I come at the topic by setting up a puzzle, what I call the puzzle of participatory wellbeing. Wellbeing is meant to be a property of ¹ For parallel observations about the mediating function of roles, see Miller (2010, 52–54), Ritchie (2018, 22), and Zheng (2018, 869). See also Section 4 below.

Alex Barber, Participatory Wellbeing and Roles In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Alex Barber 2023. DOI: 10.1093/oso/9780192843562.003.0014

   

279

individuals, and yet in reality wellbeing is enhanced—often at least, and perhaps even typically—through participation in collective endeavour. (As we will see, the difficulties this contrast generates have parallels with better-known difficulties facing the apportioning of ethical responsibility in contexts of collective agency.) Roles come in because the impact of participation on wellbeing is much more tractable if we view it through the lens of role occupancy and role performance. With that agenda in mind, I introduce the puzzle in Section 2, and in Section 3 consider some false starts, including the only discussion of it I am aware of, Gwen Bradford’s brief and (as she frankly admits) inconclusive treatment in her recent book on achievement. In Section 4 I lay the groundwork for a solution couched in terms of roles by elaborating on the comments made above about their having a mediating function. This serves as the foundation for the account to be given in Section 5 of why the puzzle of participatory wellbeing seemed so intractable in Section 3 and of the different dimensions along which participation in collective endeavours can affect individual wellbeing. A background ambition, here, is to expand our philosophical understanding of the link between human wellbeing and social roles. Plenty of valuable work has been done on how an individual’s sense of self—incorporating esteem, identity, and status—is affected by the roles they occupy (e.g. Goffman 1990 [1959]; Clark, this volume). Others have sought to adapt Aristotle’s notion of flourishing—a concept sitting at the heart of many discussions of wellbeing—to the ethics of roles (e.g. MacIntyre 1981). In this chapter I complement this existing work by using role occupancy and performance to chart the relatively unexplored isthmus connecting human wellbeing to human participation in collective projects.

2. Introducing the Puzzle of Participatory Wellbeing Philosophers diverge not only in their substantial accounts of wellbeing but in their statements of the target notion, the thing a theory of wellbeing is meant to be a theory of. Roger Crisp opts for ‘what is non-instrumentally or ultimately good for a person’, later glossing this as ‘what makes life good for the individual living that life’ (Crisp 2021). Shelly Kagan (1994) distinguishes how well a person’s life is going from how well-off they are, reserving ‘wellbeing’ for the latter. Anna Alexandrova (2017, 3–25) and Stephen Campbell (2016) are, like Kagan, alert to the possibility that philosophical debates about wellbeing are not picking out a single quality. It is noticeable, though, that all these variants, including Campbell’s rough and ready statement (‘Substantive theories of well-being purport to tell us what ultimately makes something good or bad for an individual and, more broadly, what makes a life go well or poorly for the one who is living it’, 2016, 402), have this in common: wellbeing is a property of individuals.

280



Let’s call this ‘Claim 1’, in anticipation of a Claim 2 with which it is in tension. Claim 1: Wellbeing is an attribute of individuals. Most philosophers of wellbeing will see Claim 1 as non-negotiable, an uncontroversial stipulation about the topic of enquiry. They may countenance an extension of wellbeing-talk to some non-human subjects, those that have whatever trait makes human individuals a seat of final value. Beyond that, few will go. ‘True’, they will say, ‘talk is rampant in the extra-philosophical world of the wellbeing of the planet, or of society, or of organizations, and so on, as per the examples below; but whatever is meant by this talk is not what we philosophers mean when we talk of wellbeing.’ The European project struggles on, despite the splitting headache caused by the Brexit process. Its wellbeing if not its survival ought to matter to the US. (Nougayrède 2017) The mycelial web . . . [is] important to the wellbeing of the beech, oak, and planted spruce forests of Europe. (Wohlleben 2017, 247)

The so-called ‘wellbeing’ of the EU or of forests, these philosophers will say, is either metaphorical or else has at most instrumental value: it can have an impact on (genuine) wellbeing but is not an instance of it. If it turns out that no coherent account of final value for an individual is possible, they will say, then so be it, but branching out to make it something that can be had by collectives would be to change the topic.² Pressure to acknowledge a collective dimension to wellbeing emerges, however, not simply from sloppy or figurative use of language by the folk, but from within philosophy. It emerges, in fact, from two entirely plausible claims, 2a and 2b (which together will entail Claim 2). 2a. The exercise of agency is a fundamental source of wellbeing. The thought here is not just that the products of agency enhance wellbeing (such as when a person bakes a tasty spanakopita) but that the agency as such does so ² This may have to do with how modern concern with wellbeing arose out of the egalitarian ambitions of eighteenth- and nineteenth-century utilitarians, who hoped to make ‘everybody . . . count for one, nobody for more than one’ (‘Bentham’s dictum’ according to J. S. Mill 1863, 91), even if the notion has floated free from those origins. If we look further back, we find more openness on this score. Aristotle in his Politics, for example, after linking eudaimonia to the exercise of virtue, also ‘treats the courage of a city, and the justice and wisdom of the city, [as having] the same force, and the same character, as the qualities which cause individuals who have them to be called just, wise, and temperate’ (1323b29 (1995 [c.350 ]; cf. Beggs 2003 and pace Cordell 2017). To the extent that cities’ having such virtues constitutes their having eudaimonia, eudaimonia is arguably distinct from the modern notion of wellbeing as it is usually understood.

   

281

(again as when a person bakes a tasty spanakopita). ‘Fundamental source’ is deliberately vague: it is meant to bypass questions over whether the exercise of agency is a constitutive element of wellbeing or just a very proximal cause of it, something that will vary from one substantial theory of wellbeing to the next. But no such theory can afford to ignore the wellbeing associated with the exercise of agency as such, as can be seen by running through a representative spread of them.³ This is the simple truth behind John Rawls’s remark that ‘the realization of self which comes from a skillful and devoted exercise of social duties . . . [is] one of the main forms of human good’ (1990, 73). In using ‘social duties’ as an instance of how exercising agency can be a source of wellbeing, Rawls incidentally highlights a further claim about agency: 2b. The exercise of agency is as often collective as it is individual. This, too, is beyond serious challenge on at least one reading, though we must distinguish between a stronger and a weaker reading. The stronger reading of 2b is that a collective can itself be an agent, in some robust sense that isn’t a simple shorthand for the collective’s individual members being agents. Many baulk at the notion of collective agents in this strong sense, though they would allow 2b on a weaker reading: that individuals act together and so do things together. On this weaker reading, talk of collectives as agents is just a useful façon de parler.⁴ 2a and 2b together give us the thought that is in tension with Claim 1: Claim 2: Wellbeing has a collective dimension. Claim 2 continues with the policy of using vague wording, at least for now; and its being ‘in tension with’ Claim 1 is likewise not meant to imply any outright contradiction. There is potential for contradiction, as will soon become clear. But the tension between 1 and 2 is best thought of as throwing up a puzzle that we can use, in the usual way in analytic philosophy, to map the logical terrain as we introduce precision.⁵ To take us forwards, it will be useful at this point to draw parallels between this tension and the difficulties that emerge for attributions of ethical responsibility (as

³ In the capability approach, agency, or the exercise of practical reason, is among the ‘central human capabilities’ (Nussbaum 2001, appendix). In the natural law tradition, ‘excellence in agency’ is among the goods that make action intelligible (Murphy 2001, 114). Even experientialist theories must hold that experiencing oneself as doing worthwhile and meaningful things enhances wellbeing if they are to be plausible (cf. the case in Mill 1863 against the hedonic calculus model of Bentham 1789). ⁴ Some (e.g. Tuomela and Miller 1988) have sought to reduce collective intentionality to individual intentionality rather than to jettison it. ⁵ The tension might be thought of as an instance of the ‘how-possible’ questions one often finds in philosophy (to borrow Quassim Cassam’s helpful label, 2007, ch.1). In this case, the question is one of how participatory wellbeing is possible.

282



opposed to wellbeing) when agency is collective rather than individual in nature. One immediate gain of drawing these parallels will be recognition that there are really two distinct puzzles in each domain. I will begin, therefore, with a reminder of the two puzzles in the domain of ethical responsibility, then introduce their equivalents in the domain of wellbeing. This will give the present discussion the sharper focus it needs. In discussions of ethical responsibility for acts that are collective in nature, one debate is over whether attributing ethical responsibility to collective bodies—IBM, Japan, the married couple, etc.—is legitimate. The contrary view is that praise and blame can ultimately only be attributed to individuals, except as a façon de parler. This debate deserves to be called a puzzle—we can call it the puzzle of collectiveagent responsibility—because there are strong grounds both for and against the suggestion that collectives are sometimes ethically responsible.⁶ But whatever we say about this first puzzle, there is a second—call it the puzzle of participatory responsibility—that centres on how individuals, acting as part of a collective, should or should not be deemed ethically responsible for something that neither they nor any other single individual performed, namely, the collective act. It is in the nature of a collective act that no single individual seems to perform it, yet it is only performed at all through the actions (or inactions, as it may be) of a number of individuals. There thus seem grounds both against and for holding individuals to account. It is extremely difficult to say anything principled about how responsibility should be attributed to these sub-agents, so to speak. Is it a matter of how critical each was to the event? How salient? How wilful? None of these seems to capture what we need.⁷ By analogy, and as a way of elucidating the ‘tension’ between Claim 1 and Claim 2, we can distinguish two puzzles about wellbeing in the context of collective agency. We can even give them parallel names: the puzzle of collectiveagent wellbeing and the puzzle of participatory wellbeing. For the first, while we may find it useful to say that a collective body—the EU, the London Symphony Orchestra, etc.—is a locus of wellbeing, that it is thriving or struggling, there are also strong reasons for not wanting to talk that way, save as a useful shorthand, and for insisting that individuals alone are the ultimate bearers of wellbeing. Most obviously, because collectives (unlike the people who make them up) are not sentient, they seem metaphysically ill-fitted to be subjects of ultimate concern. I am sympathetic to something like that reservation and so will barely mention the possibility of collectives as bearers of wellbeing beyond the end of the present paragraph.⁸ It has been worth the digression, though, for several reasons. One is simply that it is ⁶ In brief: collective entities often seem to behave as agents and are usefully treated as if accountable (for); yet collective agents don’t seem capable of the mental states needed for the exercise of ethically responsible practical reason (against). ⁷ Isaacs 2011 is especially lucid on the need to distinguish these two puzzles. ⁸ Eric Wiland’s welcome exploration of just this topic (2022) came to my attention too late to be incorporated into the present discussion.

   

283

useful to distinguish this first puzzle from the second, which will be the focus henceforth. Another is that what we say in response to this first puzzle has ramifications for what we say about the second (which is why I will mention this first puzzle one last time in Section 3). And finally, the potential for outright contradiction between Claims 1 and 2 heralded earlier emerges in the context of this first puzzle rather than the second. It emerges out of strong readings of 2a and 2b, which yield a commitment to the wellbeing of collectives that is in direct conflict with Claim 1. Thus, suppose we read 2b as implying that the EU is an agent in some robust sense not easily reduced to the agency of EU citizens and we clarify 2a as meaning that the exercise of agency is always a source of wellbeing to the agent. Contrary to 1, 2a and 2b now entail that collectives have wellbeing, and not just as a figure of speech or in the sense of the members of those collectives having wellbeing. To escape this contradiction, we would need to either reject Claim 1 (individualism about wellbeing), or treat collective agents as a species of individual, or resist one or other of the strong readings of 2a or of 2b. Let us now switch our attention instead to what I am calling the puzzle of participatory wellbeing. If an individual participates in a successful collective act or series of acts—a musician in a wildly successful orchestral concert series, say— then it is tempting to suppose that this success should redound upon their individual wellbeing, despite its being a collective act, just as it should redound upon the wellbeing of the other individuals involved. As with the parallel puzzle of participatory ethical responsibility, the puzzle emerges here only once we try to say anything principled about how collective success should so redound. We will see plenty of false starts in the next section, but to give the flavour, consider the absurdity of splitting the wellbeing that emerges from this exercise of collective agency equally between all the participants. That faces two immediate objections: first, not all participants participate equally, which seems pertinent; and second, it is difficult to the point of impossible to specify a non-arbitrary cut-off for who is or isn’t a participant. The unused reserve string player who attended all the rehearsals? The box-office clerk? Addressing the puzzle of participatory wellbeing occupies the rest of the chapter. In Section 3 I describe the limited prospects for answers that make no use of the notion of a role; then, in Sections 4 and 5, I offer a solution that draws on the mediating function of social roles sketched in the first paragraph of the chapter.

3. Participatory Wellbeing and Some False Starts Our challenge is to give a principled account of participatory wellbeing, the contribution to an individual’s wellbeing that accrues to a person from their

284



participation in a collective act of some kind.⁹ A prima facie case for thinking this contribution corresponds to something real can be drawn from examples, and the example of a musician performing within a musical ensemble can serve as our paradigm case. The musician’s wellbeing, like that of other individuals in the ensemble, seems to be enhanced by their participation in what turns out to be a successful performance by the ensemble, even though no musician is solely responsible for that success. A substantial account of participatory wellbeing will say how and why an individual’s wellbeing is enhanced through such participation. The label ‘puzzle of participatory wellbeing’ is justified thanks to the difficulty of coming up with a substantial account that isn’t obviously inadequate. To illustrate this, I begin in this section by taking the via negativa, presenting three simple but flawed proposals, and then reviewing Gwen Bradford’s short discussion of the problem, for she too has recognized the difficulty here and recognizes a similar set of dead ends. The first implausible suggestion is that participatory wellbeing should be thought of in terms of the benefits for the participant of the collective act’s yield or product. First flawed strategy: focus on the benefits of the product A participant’s wellbeing is enhanced through the coming into existence of the collective act’s product, which has value for her.

This misidentifies the value that needs explaining, a point already made with the example of spanakopita but which bears repeating using a collective-agency example. The value of a performance to an ensemble player is not the same as it is to a member of the audience. The exercise of agency as such, in producing music—or rather, in contributing to the production of music, which distinguishes this case from that of an individual making a spanakopita—has value for her over and above that of receiving welcome auditory stimulation. Even if the product, the auditory output, does have value to her, it is the value to her of the exercise of agency as such that we are trying to accommodate. This may push us towards a different, more sceptical thought: since the performance of Janáček’s Sinfonietta (or whatever) generates wellbeing for its performer, and since that performer was the orchestra rather than any individual member of the orchestra considered in isolation from the rest, we should reject the very notion of participatory wellbeing. This gives us: Second flawed strategy: reject the notion of participatory wellbeing The wellbeing that accrues from the exercise of agency accrues to the agent (which, in the case of collective acts is the collective itself) and not to a proper part of the agent (such as an individual within the collective). ⁹ Or costs to wellbeing in cases of collective failure, but for simplicity I will ignore these.

   

285

This account implies that we must either treat the collective itself as the subject of wellbeing or else give up on the thought that the exercise of agency as such promotes wellbeing when the agency at issue is collective in nature. Either way, there is no such thing as participatory wellbeing. This account is unattractive no matter which fork we follow. It is difficult to make sense of the idea that an orchestra can be a subject of wellbeing, save as a shorthand for its having instrumental wellbeing, most obviously for individual orchestra or audience members. It is arguably even more difficult than acknowledging collective agents as bearers of ethical responsibility (cf. Pettit 2007). But irrespective of whether we can make sense of collectives as subjects of wellbeing, still there seems to be a sense in which individuals can and do benefit by participating in collective achievements and, as we have seen, not simply through the product’s having value for them. These shortcomings could push us towards a third approach: try to identify the agential contribution made by each participant, isolating it somehow from its context as part of a collective act, and consider only this solo-ized activity, so to speak, when trying to ascertain the impact of participatory activity on each individual’s wellbeing. The viola player plays the viola part, the tympanist theirs, and these distinctive contributions are the sources of their respective participatory wellbeing. Third flawed strategy: isolate the individual contribution Isolate what the participant has done from what others have done and treat their doing that as the source of their enhanced participatory wellbeing.

While this is at least apparently targeting what needs to be targeted, unlike the other two strategies, it too seems to be missing something fundamental about collective action. The problem is not just the oddity of, to take our musical example, thinking of an ensemble performance as synchronized solo karaoke (an oddity that only increases if we switch to cases where to-and-fro responsiveness is wholly unavoidable, such as having a conversation). The deeper problem is that, in collective acts, what an individual can count as having done depends, in ways that have import for their participatory wellbeing, on what others do. If the cellists all arrive drunk and under-rehearsed, by the end of the evening the lead violinist will not have participated in a successful performance even if she herself plays exactly as she might have done on a good night. And this failure redounds upon her: her participatory wellbeing is diminished, we want to say, by the fact that what she participated in was a flop. Yet what the lead violinist did ‘taken in isolation’ was identical in both scenarios. Before turning to a more promising, roles-based approach, let us consider Bradford’s brief discussion of this problem in her book on achievement and the wellbeing associated with achievement (2015). One reason for doing so is to find an ally, someone else who thinks there is indeed a problem here. Another is to

286



compare her foray with those above. She does not claim to have solved the problem, but she does make several useful moves, including but not limited to the ones cast above as ‘flawed strategies’. According to Bradford, an achievement is a difficult process (an act or coordinated series of acts) that competently results in an outcome, the product of the achievement. We can illustrate this with the example of Joseph’s making his first successful dovetail joint. This is an achievement insofar as it is difficult and involves the exercise of competence in carpentry rather than luck. Bradford argues against the thought that being an achievement is dependent on the product’s having value—it adds elegance to a cabinet, say. Her grounds for this are that some achievements have no valuable product, and her prime examples here are of sporting achievements. The value of achievements, the contribution they make to the wellbeing of the achiever, does not turn on the value of their products (how could it if some achievements have zero-value products?) but in the exercise of competencies. This puts her in a broadly Aristotelian tradition that sees wellbeing as arising from the exercise of human capacities. Bradford spends the greater part of her book discussing cases where ‘we are the sole authors of our accomplishments’ but in her final chapter (2015, 174–177) she notes that many achievements ‘are the work of a group, where the members together engage in a difficult process that competently causes its product’ (174). Since achieving in this sense is value-generating, she is led to our now-familiar question (a familiarity helped by my having earlier borrowed her choice use of the word ‘redound’), which she couches in terms of a sporting example: How does the value of their achievement redound to the lives of each player . . . ? That is, how should we consider the value to be apportioned among the players? Does each share an equal amount in the win, or does the value vary, and if so, according to what factors? (174)

She runs through and rejects some options, including the three I have just considered but others too. We have already seen why she rejects what I called the first flawed strategy. She thinks some collective achievements have products of zero value, and defends this with a sporting example—the Houston Dynamo winning the semi-finals of the Major League Soccer championship against DC United. Since the outcome is valueless in itself, the value of the collective’s achieving a win cannot be understood in terms of the valuable product. While I disagree with this aspect of her definition of achievement, we arrive at the same place: the doing has value, irrespective of the value of the product.¹⁰

¹⁰ I agree with Bradford, that is, on the irrelevance of the product’s value to participatory wellbeing. I disagree with her definition of achievement, though, since being wrong about the value of the target of your activity seems an important way in which one can count as having failed to achieve much. Perhaps

   

287

She also considers the third flawed strategy above, which she expresses as the thought that ‘each contributor gains only the value he generates and for which he is solely responsible—only his own exercise of rationality and will’ (174). By ‘rationality and will’ here, she means human capacities the exercise of which can contribute to wellbeing on her account. The suggestion, then, is that an individual’s wellbeing is enhanced by their achievement (in her sense) as an individual, considered in isolation from the collective achievement. She dismisses this, on grounds compatible with those I offered earlier. It follows from the account I have given [of the wellbeing that comes with achievement] that each contributor does indeed garner this value [‘the value of the activity for which he is solely responsible’]. However, if this is the only value each contributor gains from the group achievement, that would mean winning the championship makes no difference in value, which doesn’t seem right at all. The question is how the value of the win is apportioned to the lives of the players. (174, emphasis added)

Bradford does not address the second flawed strategy (in which the only wellbeing generated by collective endeavour is the wellbeing of the collective). This could be because she fails to distinguish between the team (considered as a single entity) winning and the members of the teams (considered severally) winning, though it could be just that she takes it as given that individual contributors benefit.¹¹ She does, though, consider and reject several further options, all of which arise out of her worry over how to apportion the wellbeing generated by winning. I have already touched on the implausibility of these further options at the end of Section 2, but did not go beyond asking rhetorical questions, so it will be useful to press a little deeper by drawing on Bradford’s discussion. The first of these further options is that ‘the value is apportioned equally to all the contributors. Everyone who has a hand in the win earns an equal share of its value’ (174). Bradford’s objection to this is simply that it doesn’t seem plausible in some cases. In our music example, being the tenor seems like it might be more significant to wellbeing than being the person who operates the curtain, despite both tenor and curtain-operator contributing to the success of the performance. Bradford then goes on to consider a suite of different factors that could explain the mince pies I so skilfully and laboriously produced, like any mince pies, just weren’t worth the trouble. The examples she appeals to when arguing for the lack of any need for achievement to have a product—games, including sports—are, I suggest, best seen as an important but special case in discussions of achievement and wellbeing, a fact that Bernard Suits trades heavily on in his influential discussion of final value (2014, 182–196). They typically involve an essential element of simulation, rehearsal, or pretence, and difficulties in the analysis of their value shouldn’t drive our overall analysis. ¹¹ She writes: ‘It is not the achievement of any one player, but of all of them’ (p. 174), apparently ignoring the other option, namely, that it is none of them but rather it is the team as such that achieves the win.

288



divergence from strict equality. One is crucialness. Recognizing that someone on the bench throughout the final may have been crucial to the team’s having reached the final in the first place—it was nearly knocked out in the second round— encourages the thought that this matters at least as much as, say, salience on the day of the final. Against this, the person who opens the curtain is just as crucial as the tenor yet seems to warrant less in the way of credit. She next considers effort. The tenor, and certain players on a team perhaps (especially those not on the bench), put more in, so get more out. But that doesn’t explain the case of the lucky striker who took their only chance, and on the face of it contributes more to the outcome than the hard-working striker who took none of the many chances that came their way. This last example suggests a further possibility: that it is size of contribution to the process that matters. But again, that doesn’t seem to capture all cases (think of two players making the same size of contribution but where one is making their debut) and is in any case difficult to know how to quantify. Bradford ends her discussion with the thought that none of these considerations alone is sufficient, but that each of them matters. Effort, outcome, and crucialness all matter for how much of the value of a group achievement an individual contributor gains from her contribution. Precisely how much they matter and how they interact with each other, however, is a question I’ll leave to explore another time. (Bradford 2015, 177)

I agree with Bradford’s suspicion that participatory wellbeing should be treated as a complex quality, in the sense of being a function of several variables or vectors. But I will try to model this complexity, not by taking over her trio of effort, outcome, and crucialness but by turning to roles.

4. Roles as Mediators Between Individuals and Collectives We are at the intersection of two distinct and difficult philosophical literatures here, one on wellbeing and one on collective agency. Neither has a great deal to say about the other. Work on collective agency has generally focused on the ontology and ethics of collective acts, especially the assignation of ethical responsibility. The impact of collective endeavour on wellbeing has hardly had a look in. Theorists of wellbeing, conversely, in focusing on what is good for the individual, have somehow overlooked the way individuals benefit by contributing to collective endeavours.¹² As advertised in my introduction, I propose we bridge this gap ¹² James Griffin’s influential 1988 discussion is typical, in that even a chapter called ‘The Case of Many Persons’ is merely about wellbeing comparisons across individuals, not the wellbeing that emerges from joint action.

   

289

using the fact that roles mediate between individuals and collectives. Participants in a collective endeavour contribute by playing a role in the endeavour; their doing so can enhance or detract from their wellbeing. We can move towards an understanding of the vectors that shape participatory wellbeing if we first review what roles are (this section, drawing heavily on the framework set out in Barber and Cordell, Chapter 1 of this volume), and then use this understanding to map out the different ways role-playing can influence the wellbeing of participants (Section 5). As the extension of the word ‘role’ to dramatic acting makes clear, roles are typically things one can move into and out of, and they therefore have occupancy conditions, norms that specify what is to count as being, say, centre-back for Littlehampton Rovers or Chief Adviser to the Treasury. Occupancy conditions break down into entrance conditions and exit conditions. Entrance conditions specify what must hold of a person if they are to count as taking on the role, voluntarily or otherwise. Exit conditions specify how one leaves the role. For professional positions, occupying the role is usually a matter of having accepted an invitation by a designated individual to sign on the dotted line after having passed various tests of competence (the entrance condition), and having not yet resigned or been let go (the exit condition). For other roles, such as parent, friend, or pope, the occupancy conditions are either less familiar, less precise, less explicit, or just harder to pin down, but they exist nonetheless since people do move in or out of such roles, even if for some roles (monarch, birth parent) the occupancy condition is partially couched in terms of birth and death. A gap can exist between what the occupancy conditions for a role are in practice, and what they are formally specified as being (if indeed they are formally specified). Not all kings have been rightful heirs, for example. A gap can also exist between what the occupancy conditions should be and what they in fact are (whether formally or in practice). In the modern world, perhaps the role of head of state should not be something one comes to sit in simply by having the right parents, the right sex, and no older siblings. Alongside occupancy conditions—the conditions for being in the role—a role has performance conditions, norms governing what one is to do in the role. As with occupancy conditions, these are most easily illustrated with professional roles, for which they are more or less explicit. Here, we find that performance conditions fall into two kinds, the obligations one is under by virtue of being in the role, and the entitlements that make carrying out those obligations feasible, perhaps including compensation. The central role obligations for the policeofficer role in the UK tradition of policing are to enforce the law and to keep the peace, while its entitlements include a legal right to use force, such as power of arrest, various support structures, and a salary and pension. As with occupancy conditions, we need to be alert to a difference between what the performance conditions are said to be and what they are in practice; and a further gap between what the performance conditions are (on paper or in practice) and what they

290



should be. Finally, alongside occupancy and performance conditions, roles have a function, or an ostensible function, some purpose or good outcome that is supposedly achieved by the performance of the role. I will assume here that roles are constituent parts of institutions (though this will involve stretching the latter word in some cases). Indeed, a good way to think of a role is as a human-sized, human-shaped component of an institution. An institution functions, in effect, through its members—individual human beings— occupying and performing its various constituent roles. By fulfilling a role’s occupancy conditions, one becomes a part of its host institution—or, since ‘institution’ is sometimes too grand a word, part of a stable or ad hoc group, such as a friendship group or family or a one-off expedition. No one is a member of the Roberts family unless they occupy some familial role—sister, stepson, uncle, as it may be. Moreover, it is through the performance of these constituent roles by their occupants that the institution functions. If no member of the Roberts family acts as a good (or minimally effective) brother, aunt, etc., whatever that involves, then the family will simply not function, and at a certain point one might even say it will cease to exist save as an abstraction—a mere family tree—as opposed to a functioning social unit. The same is more obviously true of a formal institution such as the Treasury, and less obviously true of friendship groups (hence my qualification about having to stretch the word ‘institution’ to include such groups).¹³ A central question for any theory of roles concerns the nature of the bindingness of roles, and this has led many (quite rightly) to worry about the nature of the norms attached to roles. Are they contractual? Expressions of virtue? Conventions (Taylor, this volume)? The ontology of roles outlined above is neutral on such matters, and it allows us to explain something else about roles, something unrelated to the analysis of their bindingness: that they are enormously important to wellbeing. A role’s being well designed (role performance is not impossible, suitable occupants exist, etc.), its bringing esteem, its fitting with the aspirations of the occupant, all have an effect on individual wellbeing. In the next section, I set out to impose some system on these thoughts using the framework just outlined. Among other things this will help us to see why the simple accounts of participatory wellbeing canvassed in Section 3 were inadequate in isolation, even if they each also seemed to get something right.

5. Using Roles to Chart the Vectors of Participatory Wellbeing Roles, we have just seen, are complex entities. But they are not that complex, and their complexity is useful to us in the present context since it can be harnessed and

¹³ For complications regarding friendship, see Jeske, this volume.

   

291

used to explain the complexity of participatory wellbeing. The thought here will be that wellbeing for a role occupant/performer is composed out of several distinct vectors. The attempts to understand participatory wellbeing charted in Section 3 each acknowledged at most a single vector, meaning they both captured and missed something significant about participatory wellbeing. In this final section, I first identify the three vectors of participatory wellbeing, conceptualized using the apparatus of roles-as-mediators sketched in Section 4. I then show why the resulting account is immune to the objections that sank the attempts in Section 3, while flagging some remaining challenges.

The Three Vectors of Participatory Wellbeing Since a role serves to mediate between an individual and a collective (an institution or a simple group), we can distinguish three different vectors of participatory wellbeing, each tied to a different relationship in this trio of elements (i.e. individual, role, collective). First vector: the goodness of fit between the individual and the collective Second vector: the goodness of fit between the individual and their role Third vector: the goodness of fit between the role and the collective

Talk of ‘vectors’ is not meant to be smuggling in a pseudo-scientific veneer. Participatory wellbeing is potentially quantifiable only to the extent that talk of wellbeing is quantifiable more generally; that is, it is best limited to making comparisons (‘better off ’, ‘much worse off ’, etc.). The point of vector-talk is to bring out how an individual’s overall participatory wellbeing will vary according to these three types of fit, which I now address in turn. Each vector corresponds to a relationship between two variables, and by altering one variable while keeping the other the same we can bring out how that relationship affects participatory wellbeing. This can help us to see each vector as one of the ‘factors’ Bradford was after. The first vector is about degree of fit between the individual (with a particular character, a set of values, certain preferences) and the collective (with its goals, activities, and achievements). A simple contrast can help to make the point. (Names beginning with ‘L’ and ‘H’ in the scenarios that follow are meant to be suggestive of relatively low or high participatory wellbeing respectively.) Contrast 1a — Lu cares about the environment and is a copyeditor for Plastox, a heavily polluting company. — Huw cares about the environment and is a copyeditor for Greenpeace.

292



Here, a central value of the individual is kept constant (as is their role) but the collective is varied so that it first is not, then is, a good fit for the individual. But we can appreciate the significance of this same vector by instead varying a key trait of the individual to put their concerns in harmony or discord with a fixed collective (while again keeping the role constant): Contrast 1b — Huw cares about the environment and is a copyeditor for Greenpeace. — Leah thinks global warming is a conspiracy and is a copyeditor for Greenpeace.

Contrasts 1a and 1b bring out the dependence of participatory wellbeing on how well the concerns of the individual mesh with the nature of the collective (what it stands for, what it does, what it is for—on which, see Cordell, this volume). Clearly there will be subfactors at play in this first vector, which we could try to tease out. Perhaps Lu was working to sabotage Plastox, and so found her role highly rewarding. Moreover, the role does not have to be a stable social role, let alone a professional one. The same contrasts could be drawn using one-off collective acts, such as a group of academics teaming up to rob a bank or going for a hill walk together. The second vector is also about goodness of fit, this time between the individual (as before) and the role. As before we can treat the relata as variables to bring out the significance of the vector to participatory wellbeing. The pair in Contrast 2a differs in respect of the character of the individual while in Contrast 2b it is the role that changes, with the collective kept constant throughout. Contrast 2a — Larry, a born leader, is an office clerk at a large widget manufacturing firm. — Harry, a born follower, is an office clerk at a large widget manufacturing firm. Contrast 2b — Harry, a born follower, is an office clerk at a large widget manufacturing firm. — Lara, a born follower, is CEO at a large widget manufacturing firm.

As with the first vector, we could dig deeper into the different subfactors that influence and complicate quality of fit between the individual and their role by looking more closely at a variety of examples. Someone who is forgetful, for example, could do well in a role that requires imagination instead, yet in a role

   

293

that requires memory they might equally benefit from overcoming a known weakness. The third vector, finally, has to do with how the role sits within the larger body. We can highlight the impact of this vector on participant wellbeing by keeping the participant’s personal dispositions and values constant throughout (we might suppose they somewhat enjoy the role and somewhat approve of the institution) while varying their role in Contrast 3a or the nature of the collective in Contrast 3b. Contrast 3a — Liam is a senior blogger at NASA. — Helen is a space scientist at NASA. Contrast 3b — Helen is a space scientist at NASA. — Luís is a space scientist at the University of Lisbon.

Being a space scientist at NASA may not in fact be every space scientist’s dream, and a senior blogger could be deeply enthusiastic and informed about space, and the examples are merely intended to elucidate this third vector, not nail it down. They bring out how a role’s significance within the institution can make a difference to the participatory wellbeing that accrues from occupying and performing the role. Whether wellbeing does indeed accrue in actual cases will inevitably depend on the details, as with the two previous vectors. Ultimately, the fit that matters is a complex three-place relation between three complex relata, the individual, the role, and the group. The three vectors were here illustrated using relatively stable professional roles, simply because the point is more clearly made that way. The same contrasts could also be made, less lucidly, using examples of roles within families or friendship groups. They can also be made about one-off collective acts, since it is possible to think of individuals as having roles here too. Imagine a group of friends climbing to the top of a mountain. While less will be formally specified, the same structures will be in play. Some will be more invested in the expedition, or care more for the friendship group itself, than others (first vector). The role of gate-closer will be less significant in this context than the role of principal navigator (third vector). Some will be better suited to one position of responsibility—arranging a good route or whatever it is—than to some other position (second vector). To go further would involve taking a stand on certain questions in the theory of wellbeing. A preferentist about wellbeing, for example, would eschew the notion

294



that there is something objectively good (for the individual) about being a space scientist at NASA as opposed to a blogger, over and above the individual’s wanting that role in that organization. Someone with more objectivist leanings might see things differently. The three-vector framework above shows just that, whatever our background theory of wellbeing, we can use the apparatus of roles to structure our thinking about participatory wellbeing. My chief aim here has been to highlight and explain the different factors at play in the determination of participatory wellbeing, not to applying that understanding by drawing out any implications for the promotion of participatory wellbeing or for distributive justice. But participatory wellbeing is demonstrably important to human lives, and roles are clearly key to it irrespective of whether the analysis above is correct. So, while the wellbeing of individuals in a collective is rarely the sole consideration in that group’s design (sometimes we may just want the organization to function well, reducing our concern for the wellbeing of individuals almost entirely to its making for better soldiers, so to speak), a next step in this discussion would be to develop and adapt this understanding so as to identify and promote roles that benefit individuals in a just way. Just as wellbeing can be affected by the supply, quality, and distribution of homes, so there can be a problem with the supply and quality of roles, our homes in a figurative but just as vital sense. This, however, is beyond the scope of the present chapter.

Comparison with the Simple Accounts In this section I will show how the present role-centric approach to participatory wellbeing fares when compared to its competitors, the false starts surveyed in Section 3. It turns out to lack their shortcomings, and in some cases can be used to diagnose their failure. I began Section 3 by looking at three flawed strategies. The first—that it springs from the value of the product—was criticized because it targeted the wrong thing. This criticism is not applicable to the present proposal because the three vectors are all about the process of production, and not about the value of the product to the individual outside of that process. That product may perchance benefit the individual, but such benefit is distinguishable from participatory wellbeing. The second flawed strategy was to deny that participatory wellbeing corresponded to anything real, holding that when the agency being exercised is collective, the beneficiary—setting aside the benefits of the product—is either the collective agent or it is no one. This was a move of desperation, since neither disjunct is appealing. The present account gives a positive account of participatory wellbeing, removing the pressure to be sceptical about participatory wellbeing in that way. The third flawed strategy was to treat participatory wellbeing as one might treat the wellbeing that arises from individual agency, after first isolating the individual’s

   

295

agency from that of the rest of the collective. This seemed to ignore the nature of collective agency, and so failed to capture something distinctive about the benefit of participation, namely, that it depends in part on what others do, and not entirely on what the individual does. The present account does not fall into the same trap. A flop’s being a flop has an impact on the wellbeing of its brilliant star performer, just as it can matter to a brilliant space scientist that they are stuck in one of the world’s lesser space agencies. I then followed Bradford in considering several analyses of participatory wellbeing, all of them different ways of understanding how the wellbeing generated per hypothesis by collective achievement redounds upon the participants or contributors. These included strict equality, or departure from strict equality on grounds of effort, outcome, and crucialness. None of these seemed right on its own, but Section 3 ended with the hope (also expressed by Bradford) that some function of these or other simples might work. The present proposal is an implementation of that approach.

6. Conclusion Existing discussions of wellbeing in the context of role occupancy tend to look at either status, esteem, and identity, or at the notion of flourishing in a role. In this chapter I have begun to explore the overlooked notion of participatory wellbeing, the wellbeing that can accrue to individuals through their participation in collective endeavour. This is an important ingredient of individual wellbeing as a whole, but it is also puzzling. After some preliminary work identifying what the puzzle is, which involved distinguishing it from any puzzle over the attribution of wellbeing to collective agents, I followed Bradford in concluding that the puzzle was one of a complexity that made it hard to conceptualize. I then went on to argue that we can model this complexity by conceiving of participation as a matter of occupying and performing a role within a collective. Participatory wellbeing has a number of vectors, in that it depends on the goodness of fit between the individual and the collective, the individual and their role, and the role and the collective. There is still plenty of work to be done, of course. Each of these vectors is itself complex, and nor have I said anything about how they combine (it could be additive, though perhaps they interact). But this structure gives us something to work with and may perhaps give us enough that we can begin to think about how to promote good and just roles.¹⁴

¹⁴ I am grateful to colleagues in The Open University Philosophy Department, to participants in the Role Ethics Network, for extensive feedback on earlier versions of this chapter, and to the Arts and Humanities Research Council for funding the latter network (Project ref: AH/N006321/1).

296



References Alexandrova, Anna. 2017. A Philosophy for the Science of Well-Being. Oxford: Oxford University Press. Aristotle (1995 [c.350 ]) Politics. Ernest Barker and R. F. Stalley (trans. and eds.). Oxford: Oxford University Press. Barber, Alex and Cordell, Sean. This volume. ‘An Overview of Social Roles and Their Ethics’. Beggs, Donald. 2003. ‘The Idea of Group Moral Virtue’. Journal of Social Philosophy 34 (3): 457–474. Bentham, Jeremy. 1789. An Introduction to the Principles of Morals and Legislation. London: T. Payne and Son. Bradford, Gwen. 2015. Achievement. Oxford: Oxford University Press. Campbell, Stephen. M. 2016. ‘The Concept of Well-Being’. In The Routledge Handbook of the Philosophy of Well-Being. Guy Fletcher (ed.), pp. 402–414. New York: Routledge. Cassam, Quassim. 2007. The Possibility of Knowledge. Oxford: Oxford University Press. Clark, Samuel. This volume. ‘Three Relations Between Roles and the Good’. Cordell, Sean. 2017. ‘Group Virtues: No Great Leap Forward with Collectivism’. Res Publica 23: 43–59. Cordell, Sean. This volume. ‘Role Ethics and Institutional Functions’. Crisp, Roger. 2021. ‘Well-Being’. Stanford Encyclopedia of Philosophy. Edward. N. Zalta (ed.), Winter 2021 Edition. . Goffman, Erving. 1990 [1959]. The Presentation of Self in Everyday Life. Harmondsworth: Penguin. Griffin, James. 1988. Well-Being: Its Meaning, Measurement and Moral Importance. Oxford: Oxford University Press. Isaacs, Tracy. 2011. Moral Responsibility in Collective Contexts. New York: Oxford University Press. Jeske, Diane. This volume. ‘Are Obligations of Friendship Role Obligations?’ Kagan, Shelly. 1994. ‘Me and My Life’. Proceedings of the Aristotelian Society 94: 309–324. MacIntyre, Alasdair. 1981. After Virtue: A Study in Moral Theory. Notre Dame, Ind.: University of Notre Dame Press. Mill, John Stewart. 1863. Utilitarianism. London: Parker, Son and Bourn. Miller, Seumas. 2010. The Moral Foundations of Social Institutions: A Philosophical Study. New York: Cambridge University Press. Murphy, Mark C. 2001. Natural Law and Practical Rationality. New York: Cambridge University Press.

   

297

Nougayrède, Natalie. 2017. ‘Angela Merkel Knows she must Defuse Donald Trump’s Threat to Europe’. The Guardian 17 March 2017. . Nussbaum, Martha C. 2001. ‘Adaptive Preferences and Women’s Options’. Economics and Philosophy 17 (1): 67–88. Pettit, Philip. 2007. ‘Responsibility Incorporated’. Ethics 117 (2): 171–201. Rawls, John. 1990. A Theory of Justice. 2nd Edition. Cambridge, Mass.: Harvard University Press. Ritchie, Katherine. 2018. ‘Social Creationism and Social Groups’. In Collectivity: Ontology, Ethics and Social Justice. Kendy Hess, Violeta Igneski, and Tracy Isaacs (eds.), pp. 13–34. Lanham, Md.: Rowan and Littlefield. Suits, Bernard. 2014. The Grasshopper: Games, Life, and Utopia. 3rd edition. Peterborough, Canada: Broadview. Taylor, Erin. This volume. ‘All Together Now: When Is a Role Obligation Morally Binding?’ Tuomela, Raimo and Miller, Kaarlo. 1988. ‘We-Intentions’. Philosophical Studies 53: 367–389. Wiland, Eric. 2022. ‘What is Group Well-being?’ Journal of Ethics and Social Philosophy 21 (1): 1–23. Wohlleben, Peter. 2017. The Hidden Life of Trees. Jane Billinghurst (trans.). London: William Collins. Zheng, Robin. 2018. ‘What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice’. Ethical Theory and Moral Practice 21: 869–885.

15 Virtuous Chameleons Social Roles, Integrity, and the Value of Compartmentalization Luke Brunning

1. Introduction We all have many social roles. You are probably reading this chapter as an academic or student. Right now, I’m writing it as a researcher, but earlier today I occupied my administrative role. Tomorrow, I will lead some classes as a lecturer and, when I get home, I’ll be pouring drinks as someone’s partner. There are benefits to this multiplicity, as we shall see. Nevertheless, having multiple roles can be claustrophobic, especially when their demands are shrill or simultaneous. As harried role-occupants we may wonder, how do we manage our roles in practice? One natural answer is that we should compartmentalize our lives, creating distinct domains of action with corresponding mindsets, to help us meet the demands of our roles. Nevertheless, this compartmentalizing strategy, assuming it to be effective, can prompt ethical concerns. Is compartmentalization in tension with having integrity? Won’t this strategy immerse us in the demands of individual roles at the expense of the bigger picture? Would it not be better to try and integrate the roles we have, or our treatment of them, so that we are the same person in different situations? This chapter explores these questions. In section 2, I describe social roles, their norms and comprehensive goals, and distinguish them from relationships. In section 3 I argue that having multiple roles can expand the self in beneficial ways, and in section 4 I introduce compartmentalization as a way of managing roles. Section 5 is where I set out several concerns about compartmentalization, that it impedes practical reflection, undermines our integrity, or is alienating. In section 6, I reply to these worries, showing how they may rest on premises we can challenge and arguing, more positively, that compartmentalization actively facilitates good action.

2. Social Roles We are seldom far from a social role. Broadly speaking, roles ‘are bundles of expectations directed at the incumbents of positions in a given society’ Luke Brunning, Virtuous Chameleons: Social Roles, Integrity, and the Value of Compartmentalization In: The Ethics of Social Roles. Edited by: Alex Barber and Sean Cordell, Oxford University Press. © Luke Brunning 2023. DOI: 10.1093/oso/9780192843562.003.0015

 

299

(Dahrendorf 1968, 36). These expectations are typically broken down into those governing our behaviour, and those shaping how we are to appear and feel (Dahrendorf 1968, 36). To this, I add a third category of norms about how we are to relate to the role itself, at least for certain roles. A gardener, for example, may have to tend the roses, but can smile or scowl as they please; a flight attendant, in contrast, has to safeguard passengers with enthusiasm (Hochschild 2003, 6);¹ a psychoanalyst may have to remain curious and open about what being an analyst involves. Role norms will be more or less codified depending on the context. ‘Greedy’ professional roles are often more clearly delineated, regiment more of the personality, and afford less personal discretion than other roles, like being a parent or girlfriend (Nippert-Eng 1996, 25). Most roles situate us in relation to specific people in other roles; flight attendants to pilots, sons to fathers, prosecutors to judges. In turn, many roles are performed within socially recognized, normatively structured, and often spatially defined contexts. The norms that constitute a specific role are themselves often segmented by reference to specific people or contexts (Dahrendorf 1968, 36). The role university academic typically fragments into distinct segments of academic-in-classroom, academic-in-committee, or academic-in-a-research-seminar. People can acquire and lose segments of a role over time. The constituent norms of most roles are connected to what Stefan Sciaraffa terms a role’s ‘comprehensive goal’ (Sciaraffa 2011, 111). Someone who has such a goal ‘sees (to some degree at least) a point and purpose, or a closely connected set of points and purposes, underlying the complex of considerations that constitute her goal’ (2011, 112). This loose sense of a goal is rarely present to mind as people act, but it renders certain considerations ‘salient . . . against a background sea of reasons’ (2011, 111). The comprehensive goal of a university academic is to promote critical inquiry, or something to that effect. Comprehensive goals can be troublesome to define, and people disagree (Hardimon 1994, 355). What is the ‘goal’ of being a priest, say, or an aunt? Comprehensive goals take different forms. Some are end-states to be achieved, as when a football manager aims to win the league or avoid relegation, but others involve the formation of ‘ways of being to be maintained and developed’ (Sciaraffa 2011, 111). Jonathan Lear describes being a psychoanalyst in a manner that suggests it is a way of being: . . . psychoanalyst is not simply a term like newspaper reader or airline passenger, which describe things we do, even things we do often . . . Psychoanalyst describes who we are . . . as psychoanalysts we are constantly in the process of shaping ¹ ‘In the case of the flight attendant, the emotional style of offering the service is part of the service itself . . . seeming to “love the job” becomes part of the job’ (Hochschild 2003, 6).

300



ourselves as psychoanalysts . . . We strive to shape ourselves into people who can listen well and who can intervene in ways that are genuinely helpful to our analysands. This is a process of becoming a certain kind of person. (Lear 2003, 32)

Roles that are oriented towards certain ways of being may centre on specific virtues as a constituent feature. To occupy the role of lawyer, councillor, or watchmaker, someone may have to develop specific traits typical to the profession (Oakley and Cocking 2001). One role can have several comprehensive goals. A president, for instance, may strive to deliver policies based on their mandate while also protecting their country’s democratic institutions and representing the nation overseas. Some roles are directed toward ways of being and end-states to be achieved. Roles and relationships are closely connected. Roles often facilitate relationships, from office romances to friendships, and relationships can lead people to occupy new roles. Much writing on social roles focuses on the distinction between home and work, but there is much work at home, and some people are most at home at work. In this chapter I will occasionally consider roles and relationships together, since much of what applies to the management of roles applies to the management of relationships and I see no need to distinguish starkly between the two. But there are differences between roles and relationships. Roles are typically limited to certain domains, whereas relationships can spread across domains (hence their potentially destabilizing force). Roles, unlike relationships, are typically socially or institutionally framed (there is no friendship ombudsman). Most importantly, a role, unlike a relationship, can be occupied by different people over time.

3. The Expanding Self In considering whether having multiple roles is good for us, Mikael Nordenmark produced a sociological study to contest the following clash of intuitions. On the one hand, we might think more roles provide us with more resources and sources of potential satisfaction. On the other hand, the more roles we have, the more complex and potentially conflicted life can become (2004, 16). Nordenmark focuses principally on differences between home and work. He ultimately finds in favour of having multiple social roles. People with multiple roles can typically access greater resources and sources of satisfaction, both insofar as they occupy roles and insofar as the increased complexity is experienced as satisfying. Their lives may also have the kind of stability that comes from being diversified; the loss of one role is less likely to devastate an identity (Brunning 2018, 179).

 

301

Let us consider why having multiple roles is good for someone. Nordenmark’s view resonates with the approach of Arthur Aron and colleagues, who argue that human sociality is best understood in terms of the ‘self-expansion model’ (2013). On this conception of human behaviour, we are motivated to expand ourselves through relationships with others since relationships provide us with ‘resources, perspectives, and identities’ which help us ‘enhance [our] potential efficacy’ in attaining goals (2013, 91). This model is encompassing, in that roles and relationships are said to provide ‘knowledge, social status, community, possessions, wealth, physical strength, health, and everything else that can facilitate goal attainment’ (2013, 91). Aron and his colleagues also suggest that self-expansion is typically experienced positively. People enjoying grappling with novel situations and challenges, and an expanded, socially rich, self, have more opportunities for these experiences (2013, 92). A key premise for Aron and colleagues is that we expand in this way because of the cognitive process of internalization: When we include another person in the self, our cognitive construction of the other overlaps with (or shares activation potentials with) our cognitive construction of the self . . . Thus, to the extent we include another in the self, we take on the resources, perspectives, and identities of that person, and we share that person’s outcomes. The other person then informs who we are, enhances the tools we feel we have at our disposal, shapes how we see the world, and affects the costs and benefits we perceive ourselves to incur. (2013, 102)

Internalization populates our inner life. A lawyer or chef, new to their role, will acquire new beliefs, desires, concepts, patterns of attention, repertoires of emotion, and role-models. This process may also have an explicit narrative counterpart as they revise their self-understanding and sense of their life trajectory to accommodate these roles. Nevertheless, it is important to note that internalization can proceed without explicit reflection, and even in open conflict with someone’s existing narrative self-concept. Someone may occupy a role reluctantly, or with resistance, while still benefiting from the resources and mentalities it provides. In many cases, there are clear benefits to adopting new roles and from having several roles. From becoming an office temp, to parenting, assuming new roles does expand the self. Looking beyond the obvious advantages of a salary, or other forms of resource, roles are an entry point into existing structures. To occupy a role is to access new concepts, forms of reasoning, and habits of interaction; it is to be oriented towards new goals, and to contend with new norms of action, comportment, and reflection. To occupy some roles is to encounter new people, too, with all the advantages supplied by an expanded social network. A wellstocked self can withstand change, stave off hedonic adaption in which people get

302



used to, and so derive less enjoyment from, familiar activities, and experience a range of diverse emotions which, in turn, is connected to increased well-being. This said, Aron and colleagues perhaps overstate the case for expansion. Much will depend on the specific roles and relationships a person grapples with, as well as their broader social context (a point of which Nordenmark is aware (2004, 124)). Someone contending with a disability, or systemic racism, may question the scope of this model as it seems to presuppose, implicitly at least, that a person’s daily life harmonizes with their social context. What is more, the self-expansion model can seem grasping. The value of multiple roles and relationships seems to lie in the goods they provide. Is there more to multiplicity, however, than just access to resources to be stowed in our warehouse-selves for future use? It seems there is, for to have multiple roles is also to have a focused awareness of value, and expanded agency.

Value One reason to think roles enhance our agency and grasp of value is that to acquire a role is not simply to acquire new means to pursue our goals, but is also to be oriented towards new comprehensive goals; to end-states or ways of being that structure action. In some cases, our practical orientation towards these goals—an encounter necessary if we are to last in the role for long—acquaints us with new values. An idealistic academic, for example, may come to appreciate the procedural value of expediency after assuming their first administrative role. A new parent may appreciate, for the first time, the vulnerable, dependent, nature of human life as they hold their newborn. In other cases, the assumption of a role facilitates the shift from believing something to be valuable to valuing it personally (Scheffler 2010, 21). We might all believe justice to be valuable, but a trainee lawyer, attending countless brief and underfunded trials, is positioned to value justice more intimately, their belief is accompanied with emotional receptiveness to the demands of justice, and their taking those demands as reasons for action (Scheffler 2010, 29). A role is like a viewing platform which orients us towards a portion of value’s wider landscape. Like the platform, which directs and organizes our gaze, a role is a point of coherence in social space. Like the platform, which promotes attention, not mere looking, roles engage us, not just our beliefs; they require us to act, and react, in service of certain goals within a specific institutional framework. Like the platform, to encounter a new role is to encounter an orientation to goals and values that other people have shaped and experienced before. To assume a new role is to be drawn into the landscape of value from a different location. But as well as revealing value, the occupancy of roles also constrains how we value things. Just as our attention is limited, our emotional openness to value has its limits and so

 

303

we cannot value everything we consider to be valuable (Scheffler 2010). There is also a practical limit on how many roles we can occupy.² In demanding certain kinds of action and comportment, roles ensure our engagement with value is not simply cognitive. But we must not conflate engagement with a role, or aspects of a role, with endorsement, or understanding. Like the observer on the platform, whose gaze might be hesitant, averted, or askance, and who might not understand fully what they see, so the role-occupant may experience ambivalence, resistance, or even repugnance towards their role or part of their role. They may also lack clarity about the value of the role, or the nature of its goal. To be oriented towards a point of coherence in social space is no guarantee such a space is welcoming or good. Yet these orientations can solidify our deep alignment to other forms of value through experiences of confusion, contrast, or resistance. I cannot be the only academic whose appreciation of teaching and research took on a new guise as I assumed more administrative roles. I come to appreciate that I hold those rewarding roles, but may also reconsider what is worth researching and teaching.

Agency But roles do more than provide us with new values or help us appreciate the roles we already value. They also provide contextually delineated ways of experiencing our own agency, of assuming new modes of agency, and in so doing of increasing our autonomy. In this way roles function, in a more expansive and consequential way, like games do, at least on C. Thi Nguyen’s (2020) understanding of what games involve. Games require us to act within certain constraints, often cooperatively or in opposition to other people, in pursuit of goals which, for a time, we assume as our own and strive to fulfil. Nguyen suggests that, In game playing, we take on temporary agencies. These agencies have been shaped by others, and are passed to us via the game. In other words, games are a medium for storing and communicating forms of agency. A collection of games can, then, constitute a library of agencies. Games can store, offer access to, and offer immersive experiences of different agencies, as well as different social arrangements of agencies. So you can expose yourselves to many different modes of agency, by exploring the library of agencies. One game might be focused on high-speed reflexes, another on calculative look ahead, or on diplomacy and bargaining, or on manipulating alliances and shared incentives. Wide ² This metaphor has its limits. Most importantly, we cannot assume the domain of value has an integrity akin to that of a visual landscape. Values may conflict.

304



exposure to games can enhance the autonomy of agents by making them aware of alternative modes of agency. (2020, 78)

‘Roles’ can replace ‘games’, here, without loss of meaning. In assuming a role, we take on a form of agency that other people have shaped and which is passed on to us within social institutions. Different roles enable us to experience the varied facets of human agency, and their social configurations. The kind of activity required of us, in occupying a role, may be temporary if role-occupancy is temporary. Exposure to new roles exposes us to new forms of agency. If we have multiple roles, our autonomy may be enriched and enlarged. The agential mode of the game player is like a social role in miniature. Like the game player, we can move in, and out of, a role’s agential mode in a fairly fluid way (more about these shifts in a moment).³ Like the seasoned game player, the more roles we occupy, or have occupied, the more agential modes we can access, so the greater the agential options at our disposal. Nguyen argues of games that increased agential options may constitute a kind of minimal freedom. Someone acquainted with the agential mode encapsulated in three games is better off than someone acquainted with that encapsulated in two, because they will have greater agential options when faced with other contexts of action (2020, 84). Specific agential modes also help focus our attention to specific kinds of reason, in a context, and thus can help align our goals with our motivations and desires. As he puts it, ‘agential modes are tools for exerting willpower by acquiring momentary focuses, which exclude whole ranges of reasons’ (2020, 87). The same is true of social roles.⁴ Someone who is a mother and a judge has more options, when it comes to styles of practical agency, than someone who only occupies one of those roles, and less than someone who also plays saxophone in a jazz band. The ability to enter into an improvisational agential mode may aid the judge in seeking justice, or in caring for a growing child. Being able to assume a judicial mindset may help the mother with the focused task of teaching her child to take responsibility for a minor wrongdoing in a context where they suffer weakness of will. Having several roles does not just equip us with agential reportions which may help us with other tasks, external to those agential roles; this expanded agency is valuable within each role too. Having multiple roles can provide us with important practical resources, bring us close to new values, deepen our engagement with what we already value, and expand our agential capacities. These benefits accrue within limits, however. A game can be boxed and put away, but the same is rarely true for our more ³ ‘Game playing, because it is so formalized, makes it easier to see that we have this capacity to fluidly shift within our agency. Once we see it happen so sharply in games, it is easier to see the softer forms of agency shifting that we perform elsewhere in our lives’ (Nguyen 2020, 80). ⁴ A point Nguyen alludes too in suggesting, ‘I am not arguing that games are the only path to alternate modes of agency. Different careers or hobbies could do it, too’ (2020, 91).

 

305

involved roles and relationships, and so there will be limits on the number of roles we can occupy at once. But the considerations set out above also apply to the roles we occupy over time within a life.

4. Compartmentalization Our selves can expand when we take on additional roles or relationships, but this multiplicity needs management. For many people, the benefits of multiple roles can appear intangible without a practical sense of how to organize them. A core feature of role management is what sociologist Christena Nippert-Eng calls ‘boundary work’ (1996). Boundary work describes the practical efforts we expend in managing the mental dimensions of having multiple roles (1996, 7). These efforts include the task of placing boundaries between roles and their domains, and the activity of moving between roles and their domains. Consider the playing of a board game. Typically, we play the game in a certain time and place, and transition into the game mindset. On a Friday night after work one of us assembles the board, another sorts the pieces, another grabs the beers, another dims the lights. We segue out of the daily grind and into the world of the game. Now consider shifts within the day of an academic. We lecture in some places, write in others, attend meetings in others. Each activity requires its distinct mentality and focus, and we erect barriers between domains to maintain this. We will have our idiosyncratic ways of shifting between spaces and roles, taking time to make a coffee and gather our thoughts after a long class and before a detailed meeting, or striving to adopt an empathetic attitude before an office-hour after a morning spent squinting at spreadsheets. Boundary work is constrained by how much socially and institutionally afforded discretion we have. Some employers, for example, strive to deprive employees of transitional spaces; some people may lack the time to shift gear. Since personal relationships afford us the most discretion to be who we want to be, compared to the strictures of the working world, the boundary work there often involves the management of whole idiolects. This fact is most visible in the transitions between different intimate relationships, for example, the segue from being a romantic partner to being a sibling. Sometimes boundaries cannot be maintained and paralysis can ensue, as when a teenager is caught in a room between their friends and their parents. Broadly speaking, two styles of boundary work lie on a continuum, the integrating and the compartmentalizing. Integrators make it easier to move between different role domains and mindsets, but struggle to erect boundaries between them. Compartmentalizers place clear boundaries between domains, but have to work harder to transition mentally between them (Nippert-Eng 1996, 27). As Nippert-Eng puts it, using ‘segmentation’ for compartmentalization:

306



The more we segment, then, the more we create dedicated, single-purpose spaces and times in which to embed and protect our dedicated, single-purpose selves. And the more we integrate, the less we need protective territorial borders for our more all-purpose selves. (Nippert-Eng 1996, 42)

Compartmentalizers separate out aspects of their lives associated with different roles. Each domain may have its own norms, etiquette, modes of communication, forms of self-presentation, tools and accessories, and so on. Home is distinguished from work, teaching from writing, friends from colleagues. Moving between domains, and different roles, is the difficult task for the compartmentalizer. They need to ‘gear up’ or ‘wind down’, to ‘get their game face on’, and so on. Compartmentalizing boundary work can seem advantageous. Erving Goffman, for example, discussed the dangers of ‘role strain’ when encountering people who relate to you in terms of divergent roles. He suggested, ‘the answer to this problem is for the performer to segregate his audiences so that the individuals who witness him in one of his roles will not be the individuals who witness him in another of his roles’ (1990, 137). It is easy to recognize the usefulness of his suggestion that this segregation can also ‘allow a few moments in between performances so as to extricate oneself psychologically and physically from one personal front, while taking on another’ (1990, 38). I am grateful when my timetable affords me such moments. That said, the premise underlying this thought seems to be that roles, and their audiences, are discrete and ought to be kept separate; not simply that compartmentalization is idiosyncratically useful, but that it enables people to occupy their roles correctly. Expressed simply, Goffman seems to favour compartmentalization because he presumes society is itself compartmentalized. Alasdair MacIntyre also seems to accept this premise. He thinks that modern societies are compartmentalized, and ‘each distinct sphere of social activity comes to have its own role structure governed by its own specific norms in relative independence of other such spheres’ (1999, 322). He views the ‘divided self ’ as a product of this social order; such a self has ‘developed habits of mind that enable it not to attend to what it would have to recognize as its own incoherences, if it were to understand itself apart from its involvements in each of its particular roles in each distinct sphere’ (MacIntyre 1999, 326 (my emphasis)). But, as we shall see, MacIntyre denies that these habits of mind are ultimately a justifiable response to the social structures which gave rise to them. Even if we are sceptical that ‘society’, writ large, is compartmentalized, it could be a contingent matter that our specific roles require vastly divergent behaviours, or relate us to different groups of people (think of the judge with children who moonlights as a jazz saxophonist). In such a life, compartmentalization can make action easier. More broadly, compartmentalization can help us avoid conflicting mindsets even if we think there is nothing problematic about appearing in

 

307

different ways before the same audience, or that this is demanded of us, due to our role (Nippert-Eng 1996, 209). A parent working from home, one eye on their manuscript, one eye on their child, may find they are torn between the demands of two different kinds of agency. It’s not that they need to segregate audiences, but rather they want to be able to enter fully into one way of thinking for a time. These considerations motivate the practical appeal of compartmentalization, but some people remain deeply wary of this approach to life. So, before I can develop a better defence of compartmentalization, we need a clearer sense of why it worries people.

5. Integrity and Other Concerns In writing about compartmentalizing boundary work, Nippert-Eng suggests that, The more we segment, the more we must become something of a chameleon, able to transform ourselves from one distinct way of being to another, given a certain sociocognitive backdrop. (1996, 102)

For critics, the possibility that we turn into chameleons of this kind is precisely the problem. Three related worries stand out, that the compartmentalization of social roles encourages us to be immersed in our roles in a way that impedes practical reflection, undermines our integrity or commitment to the good, and may be alienating. MacIntyre (1999) helps us illustrate the first of these worries. He is concerned that compartmentalization impedes the reflection necessary for moral responsibility because it encourages people to be immersed in their roles, and so only occupy one practical perspective when more perspectives are needed. If we compartmentalize our roles, we may avoid conflicts between their demands. Although this avoidance of conflict can make it easier to act, MacIntyre thinks it is morally impoverishing, in so far as that self recognizes and aspires to conform to what it takes to be moral requirements, within each particular sphere of activity, it will be a morality from which the elements of potential and actual conflict are missing, a diminished morality that matches its diminished powers of agency. (1999, 325)

Viewed in this way, some kinds of role strain are morally valuable because they can prompt us to consider the demands of a role from an external vantage point. Robert Adams exemplifies the second worry in also focusing on the moral risks of compartmentalization. In keeping our roles separate, to facilitate the adoption

308



of a specific mindset in each context and so be able to serve the goals of the role, we apparently undermine our commitment to the good: In a way it would be easier to keep the values of each [social] sphere in a separate compartment in my life. But if I do that, I am an evaluative chameleon, changing my ethical colours as I move from one situation to another. If I do not have a single system of values by which I live in all the spheres in which I move, disturbing questions can be asked about me. Am I, as a person, really (fully, deeply) for the goods affirmed in my stance in any of those spheres? (Adams 2006, 207–208)

Expressed in another way, compartmentalization undermines integrity. People often understand integrity as a kind of constancy between situations. Someone with integrity has a ‘a stable and coherent sense of who one is and why one is’ (Scherkoske 2013, 11). Robert Solomon, for example, has a structural understanding of integrity understood as ‘the integration of one’s roles and responsibilities and the virtues defined by them’ (1997, 216) which seems to preclude active compartmentalization of roles. Finally, we might worry that compartmentalizing is alienating, that it may ‘bring a certain hollowness into the more prized parts of one’s life, as one tries to respond with only part of oneself ’ (Adams 2006, 207). Such alienation can be distinguished from the moral worries above. Instead, we might just worry that someone is missing out in their immersion into a role, and in the process of transitioning between roles; that there is no need to take off one ‘hat’ to don another. Pushed to extremes, we might worry that compartmentalization of roles reifies those roles at the expense of the person underneath; that our selves are more than Goffman’s coat peg, ‘on which something of collective manufacture will be hung for a time’ (1990, 245). Integration of some kind is often proposed as a response to these concerns. We should forgo compartmentalizing boundary work and make sure we do not get lost in the demands of individual roles. Adams embraces this as a general, and demanding, project, suggesting that ‘the advantages of integration, and perpetual reintegration, for Virtue, are obvious, and we surely have good reason to pursue moral integration as a project’ (2006, 210). MacIntyre concurs, and suggests we need to limit our compartmentalizing tendency. For him, ‘to have integrity is to refuse to be, to have educated oneself so that one is no longer able to be, one kind of person in one social context, while quite another in other contexts. It is to have set inflexible limits to one’s adaptability to the roles that one may be called upon to play’ (1999, 317). To have integrity in this way is, for MacIntyre, to learn how to reflect on our social roles ‘independently of the requirements of those roles’ (1999, 318).

 

309

We can aid such reflection by facilitating the existence of ‘reflective milieus’ which, would provide agents with what they otherwise lack, an understanding of themselves as having a substantive identity independent of their roles and as having responsibilities that do not derive from those roles, so overcoming divisions within the self imposed by compartmentalization and so setting the scene for types of conflict that compartmentalization effectively suppresses. (1999, 324)

Reflective milieus prevent over-immersion in roles and enable us to experience conflicts between the different demands and general ways of being our individual roles require. For MacIntyre, the ability to experience conflicts of this kind is a crucial part of adequate moral reflection, and necessary if we are to take responsibility for our actions within roles.

6. Virtuous Chameleons I shall defend compartmentalizers from the criticisms above. My argument is not simply that compartmentalization is unproblematic in some cases but, more strongly, that compartmentalizing roles is an important feature of moral life which helps us to act well.

Immersion First, let us tackle the worries about compartmentalization and immersion in roles. It is important to note that the question of whether we are immersed in roles—that is, find our selves dissolved into them, and their demands, without attending to much else—is orthogonal to that of whether we compartmentalize roles. The connection between the two is contingent. A compartmentalizer can keep roles separate, to preserve distinct role-mentalities, without being immersed or over-immersed in them. An integrator can be immersed into their wider life. They might have one vocation, or fail to distinguish between any of their social roles (the legendary mathematician Paul Erdős comes to mind). Whether this is good, or not, is an open question, but it does not hinge on their prowess as a compartmentalizer. We must not presuppose that to occupy a role, to take its demands as reasons for action, is thereby to be immersed in it, either in the psychological sense of being unable to extricate ourselves from the role-mentality, or in the constitutive

310



sense of how that becomes ours. We should resist thinking that for a role to be ours, we must endorse its constitutive norms, or identify with its comprehensive goal. Christine Korsgaard takes the former view; that to have a role is to endorse its norms and point. For her a role is a ‘practical identity’, which is a ‘description under which you value yourself and find your life worth living and your actions to be worth undertaking’ (2009, 20). She suggests that ‘one might think of a particular practical identity, if a little artificially, as a set of principles, the dos and don’ts of being a teacher or a citizen, say’ (2009, 21). Roles become practical identities, for Korsgaard, when someone ‘endors[es] the reasons and obligations to which that way of identifying [ourselves] gives rise’ (2009, 28). She often describes this process of endorsement in a way that aligns it with the notion of immersion, especially when she suggests that, you can wholeheartedly endorse even the most arbitrary form of identification, treating its reasons and obligations as inviolable laws. Making the contingent necessary is one of the tasks of human life and the ability to do it is arguably a mark of a good human being. To do your job as if it were the most important thing in the world, love your spouse as if your marriage was made in heaven, treat your friends as if they were the most important people in the world—is to treat your contingent identities as the sources of absolute inviolable laws. (2009, 23 (my emphasis))

But endorsement is a diffuse notion which comes in degrees, both in terms of how strongly we endorse a role norm, and in the sense of how many norms we endorse. More importantly, we can occupy roles whose norms we do not endorse, or which we actively resist, and this resistance can indicate that we care about the role. A teacher in a modern secondary school who resents the bureaucracy expected of them, because they think it impedes the broader task of educating children, can value their role whilst arguably failing to endorse some of its constitutive norms.⁵ There are many ways to occupy a role without being worryingly immersed. Stefan Sciaraffa offers us a different picture of how one comes to have reasons for acting according to the demands of a role. He suggests we identify with the comprehensive goal of a role, rather than its constitutive norms, where this is for someone to conceive ‘of herself as a person whom the pursuit and realization of the goal is important’ (2011, 112). On this view, our alignment with a role’s goal is more fundamental than our endorsement of its norms, and so someone can

⁵ There is room for debate about the constitutive norms of most roles. For example, some might argue that the need to be attentive to the legal and administrative dimension of teaching is not actually a constitutive norm. These are important practical matters, but here it suffices to note that there are cases where our endorsement of a role is partial. Thank you to Sean Cordell for raising this point.

 

311

identify with multiple roles without getting immersed into them, or overly stuck following their specific principles. Perhaps a more nuanced approach would suggest that we can occupy roles, and compartmentalize them for expediency, without any kind of uncritical identification with their norms or comprehensive goal. Instead, we might think that each of a role’s norms and comprehensive goal offer a reflective perspective on the other. An academic philosopher, for example, can reflect on the specific demands of their institutional role in terms of the goals of education or understanding, and reflect on the goals of understanding and education in terms of the specific tasks they are expected to perform and traits they are expected to cultivate. Each focal point illuminates the other. Finally, and most interestingly, to properly occupy some roles is to heed norms about how we are to relate to them; norms which preclude the kinds of immersion that can worry critics of compartmentalization. Consider my earlier example of psychoanalysis. Much has been written about the normative ideal of psychoanalysis, and many analysts are deeply concerned about the dangers of adhering to a rigid conception of what their role involves. Psychoanalysts are aware that they occupy a role with conventions and codes of ethics, but that this role has a peculiar character. Adam Phillips, for example, suggests, ‘An analyst should be someone you have an appetite to talk to and who has a desire to listen to you. Not a professional desire, which is a contradiction in terms’ (2014), whereas Jonathan Lear writes that ‘to be an analyst one must ever be in the process of becoming an analyst. For anyone for whom the process of becoming is over, he or she has ceased to be an analyst’ (2003, 32). One way in which this ‘process of becoming’ can end, or ossify into a ‘professional desire’, is if the analyst becomes unreflective about their role, and starts to maintain ‘fictions of themselves’ as a psychoanalyst, as the analyst Robert Caper puts it. To counter these risks, Caper enigmatically suggests that ‘it is the analyst’s task to play no role at all’ (1999, 29). I suspect this ideal is more common than the rarefied example of psychoanalysis may suggest. Part of what it is to be a parent is not to have an inflexible conception of your role; part of what it is to be a philosopher, or a composer, or scientist, or romantic partner, is that your relationship to your role is not immersive in the way that concerns critics of compartmentalization. These roles can require distinct domains and mentalities, which can require active maintenance, but part of what it is to occupy them well is to avoid stale, routine, unreflective, immersion. These roles seem open to temporally extended forms of reflection because of the rich and complex nature of their comprehensive goals, or ends. They seem to be, or involve, what Talbot Brewer calls ‘dialectical activities’, which have a self-unveiling character, in the sense that each successive engagement yields a further stretch of understanding of the goods internal to the activity, hence of

312



what would count as a proper engagement in it. If the activity’s constitutive goods are complex and elusive enough, this dialectical process can be reiterated indefinitely, with each successive engagement yielding a clearer grasp of the activity’s proper form and preparing the way for a still more adequate and hence more revealing engagement in it. (Brewer 2009, 37)

Not all roles have this character. Perhaps there are other roles which do not preclude, or even praise, immersion, but the fact that some do not means we cannot worry about immersion in general terms, and we can take comfort from the fact that some of our most valued roles encourage us to guard against this possibility if we want to occupy them well.

Roles and Virtues In section 3, I suggested roles offer us distinct modes of agency, which is why having multiple roles can be valuable. But some philosophers seem wary of compartmentalization as a strategy of managing roles, because they think this will undermine our ability to be virtuous or will weaken our commitment to values. These concerns seem a bit inchoate. Perhaps they worry that if someone compartmentalizes their job as a judge from their role as a parent, for example, then they may value justice in court but neglect it at home. This view, supposing anyone holds it, neglects the fact that many virtues and values are common to different roles, so the compartmentalization of roles is unlikely to decrease our commitment to those values. Instead, the bigger question is simply that of whether the apparent values internal to specific roles are actual ones. This point aside, we might think that the compartmentalization of roles is actually necessary if we are to act well, not simply compatible with this end. Put differently, on one conception of what virtues are, the drive to integrate roles and their corresponding mindsets may be inimical to at least some forms of virtue. To see this, let’s consider Christine Swanton’s discussion of virtue ethics and social roles (2007). For Swanton, the virtues internal to roles are specified in terms of a prototype, on two levels. The first is a thin specification, which functions at a high level of generality by specifying the field of application in question (2007, 211–212); the second is a thick specification where conduct is understood in terms of ‘mother’s knee’ rules, and various dispositional and emotional heuristics. Social roles help provide further detail to the thick specification of these prototypical virtues. For Swanton, it is ‘only when more specific requirements are determined by role-differentiation do we know what it would be to act well as a human being’ (2007, 214). This kind of determination happens in reference to an understanding of a role’s comprehensive goal, which itself is understood, reciprocally, in connection to the broader social institution which houses the role

 

313

(2007, 216). So although justice, fairness, and so on may be virtues common to many roles and domains of life, we can only clearly specify what they require of us in reference to the demands of our specific roles, and the goals of those roles. In some sense, then, social roles are primary and integral to the specification of virtue; they supply the content which makes virtue-talk intelligible in practice. If Swanton is right, something I cannot adjudicate here, then integrating boundary work should concern us because it blurs the boundaries between roles in favour of a more uniform practical mindset. When boundaries between roles are blurred we risk loosening the connection between a specific role, its institutional context, and its comprehensive goal. For example, a psychoanalyst who fails to adequately distinguish working life from her home life may fail to see that her ‘analytic’ attitudes towards her partner are not properly suited to the flourishing of their relationship; or an academic who fails to maintain appropriate distance from his students may lose sight of what is involved in being a responsible educator, and in being a good friend. The integrator may find it easier to shift between domains, but domains may shift into them. While we may not mind when personality colours our bartender, we are less welcoming when officiousness taints a father. What goes wrong in these cases is that people are unable to sustain a clear ‘regulative ideal’ to shape their conduct in a role. A regulative ideal is an ‘internalized normative disposition to direct one’s actions and alter one’s motivation’ in reference to a ‘conception of correctness or excellence’ (Oakley and Cocking 2001, 25). The regulative ideal of oneself as a psychoanalyst, or teacher, or mother, involves reference to role-specified virtue traits. Compassion-qua-psychoanalyst, for example, is different from compassion-qua-lover. Compartmentalization helps us maintain the integrity of the different agential modes integral to our roles by enabling us to understand what our roles are, in reference to their broader social and institutional contexts. Management of our environment, of times, places, material objects, dress, those around us, and so on—boundary work, in short— helps delineate our characters. To take a small example, consider Nippert-Eng’s observation that, ‘if we find that looking at certain objects makes us think in a certain way, then it’s easiest to keep those objects only in the places where it’s okay to think that way’ (1996, 42). To preserve the distinctness of their mindset, compartmentalizers might keep such objects apart (no family photographs on the judicial bench, say). Integrators, in contrast, might keep such objects physically together, or close to each other, thus potentially clouding their ability to focus on the requirements of a specific role. Here I am reminded of Jonathan Lear’s observation about his office, For a decade, I used the same office both to see analysands and to see students in an academic capacity. On entering my office, they would see my desk on the left surrounded by walls of books. The couch was over on the right, down at the far end of the room. As an analysand tried to make her way to the couch, the desk

314



and the books exerted a magnetic force, pulling her towards a glimpse of the rest of my life. . . . Meanwhile, when students came to see me during office hours, the couch was the one place they couldn’t look; it was as though a whole corner of the room did not exist. The presence of the couch was made overwhelming by the intensity of the effort to ignore it. . . . During this period in which I used a single office for students and analysands, it seemed that my students wanted to be analysands and my analysands wanted to be my students. (1998, xi–xii)

Here, the integrated space generates dissonance. Such dissonance is likely to grow the more we adopt an integrating mentality and try to erode boundaries between home and work, or teaching and administration. At the very least, we face a higher cognitive burden in grasping the important features internal to each of our roles, their specific and specified virtues, and the regulative ideals which guide our conduct. Put simply, integration introduces dissonance, and dissonance obscures our grasp of virtue. Instead of acting in accordance to compassion-qua-psychoanalyst, we risk acting in some other way. To be guided by compassion-qua-lover would be wrong in the consulting room, a moral error due to conflation of roles; to be guided by compassion, period, is risky because on Swanton’s view it is unspecified. To act like this is to want to live in a prototype, rather than a finished house. A similar point applies to a related worry about compartmentalization and virtue. Some, like Adams, may worry about norms of emotional comportment within social roles. His concern about compartmentalization and our commitment to the good could be articulated in terms of the emotions we are encouraged to manifest in some roles. Do we not risk getting locked in to one pattern of emotional response, due to the demands of one role, at the expense of a broader, more human response? Moreover, if coming to personally value something involves being emotionally receptive to that thing, as Scheffler suggests, might being locked in like this undermine our ability to convert the belief that something is valuable into a more intimate form of valuing? This kind of worry is not hard to motivate. Some roles, such as those in the service industry, seem to demand a kind of emotional expression that is often out of step with how people feel more broadly, whereas other roles, such as being a paramedic, may require a great deal of emotional management. The worry here is that the forms of emotional immersion or management demanded by certain roles might come to shape our emotional life more generally, and undermine our ability to respond with appropriate feeling. This concern resonates with the worry that compartmentalization leaves people alienated from themselves. On closer inspection, however, the force of this worry dissipates, for it can arise just as strongly in response to integrative role management, too. The person who looks to blur boundaries between roles may not be an emotional chameleon, but they risk a corresponding sense of alienation and emotional obscurity. They may be responding with their ‘whole self ’ in all situations, but their emotional

 

315

engagement to value, and the emotional dimension of their virtuous traits, will remain loose and unspecified by context; a prototypical emotional response, if you will. They run the risk of emotional homogenization. Compartmentalization reduces these risks. People who delineate their roles, and manage shifts between domains, are better able to grasp the virtues specific to their roles and be guided by fine-grained regulative ideals. Their emotional responses to the values exposed by their roles will be more resonant due to such understanding, and therefore they are more likely to be able to personally value the comprehensive goals of those roles rather than just believing them to be valuable. If Swanton is right, some degree of role-separateness is important, at least insofar as it enables us to grasp the virtues specific to a role and be guided by a detailed regulative ideal. This guidance is not simply useful because it makes action easier, but it is importantly connected to the task of having a virtuous character.

Reflective Milieus MacIntyre suggested that one way to combat the potential risks of compartmentalization is to foster reflective milieus, contexts in which people can reflect about the roles they have, experience conflicts between them, and avoid immersion. Underlying this suggestion is a worry about people being absorbed into forms of immoral activity, or ending up acting in harmful or neglectful ways. We can perhaps all think of someone who would benefit from the ability to reflect more on how they occupy a certain role. MacIntyre is right that we need to reflect on our roles. But his notion of the reflective milieu could give rise to the idea of a space where naked selves, shorn of their social roles, can step back, and reflect on how things are going. We might doubt whether this is plausible. Rahel Jaeggi argues it constitutes a false dichotomy. She thinks, credibly, that alienation, when it occurs, occurs to a never-nude self that is always and only constituted through its roles (2014, 74–76). We are socialized into our selves through social roles and act through them. There seems to be no time when we are not occupying some role or other. Instead, ‘it is precisely the idea of the authenticity and wholeness of persons prior to their being deformed through roles that is problematic’ (2014, 75). Understood as an opportunity to escape from the demands of roles, into a space of pure moral reflection, the idea of a reflective milieu looks unconvincing. Even if there was such a milieu, however, someone would have to manage their practical transition in and out of it, and ensure it is compartmentalized away from other role domains. To fail to do this is to risk slipping into a reflective mode in a moment when unthinking action or ‘thinking on the job’ would be more appropriate. Sometimes it is paralysing, or even wrong, to stop and think about

316



what we are doing. Boundary work of a sort would still be necessary, therefore, to manage these transitions in and out of reflection. We should also reject the idea of a reflective milieu. Instead, all roles offer a reflective perspective on each other. As a viewpoint in social space, a role is a standpoint from which to appraise the norms and constitutive goals of other areas of life. This is especially so if, as I argued above, social roles can acquaint us with forms of value that we might otherwise overlook or only appreciate abstractly (or forms of value that deepen as we continue to engage with, and thus come to understand, the goals of our role). As a philosopher, for example, I come to value, through experience, the analytical mindset which I can use to evaluate my role as an administrator. As an administrator I come to value, through experience, expediency, and institutional design, concerns that can inflect my work as a philosopher. As a romantic partner I appreciate, through experience, the visceral dependency within some relationships which, in turn, may temper the value I place on analytical thinking or efficiency. This process iterates without end, and has force precisely because these are distinct roles. Integrative efforts obscure these differences, and so obscure the value of the evaluative perspectives and agential modes provided by each role. Finally, it is worth conceding that MacIntyre is right about the moral significance of experiencing conflict between the demands of roles, and the importance of these conflicts, or uncertainty, for developing a rich sense of moral agency. But to value conflict, like this, is to have an additional reason to favour compartmentalization of our roles. Conflicts are clearer when there are boundaries between roles and their demands are distinct. The integrator, striving to be one person in all contexts, will experience conflicts only indirectly or in nascent form. They may not experience them at all, trading, instead, on undetermined notions of virtue, or ideals of conduct, which are not supplied detail from a concrete role context. If you think conflicts and ambivalence can provoke reflection, derail moral complacency, and motivate the search for alternative options then compartmentalization should appeal.

Boundary Work is Reflective Work To defend compartmentalization is not to defend uncritical immersion in a role. Indeed, much of the compartmentalizer’s boundary work is aimed at moving between roles; focused, that is, on active processes of extrication and redirection. Critics of compartmentalization, armed with a caricature, neglect the fact that placing, maintaining, and moving between roles is itself a reflective activity, at which we can be more or less skilled. What does good boundary work look like? Expressed simply, we need to be able to place boundaries between our roles, immerse ourselves in the demands of a role

 

317

when appropriate, and move between roles. Of great value in doing this is the trait Valerie Tiberius calls ‘attentional flexibility’, which involves being ‘open to considerations that are not the focus of our current practical perspective (perhaps not at every moment but eventually), and to be able to make judgments on the basis of these considerations about our current perspective’ (2008, 83). Attentional flexibility is a general trait, not something exercised in a specific context. It is the ability to occupy a role without getting mired in its demands. Perhaps another name for attentional flexibility is that we cultivate space for irony. We can always slip a reflective cigarette paper between the mindset of one role, and that of another. To see motherhood through the eyes of a bureaucrat; philosophy through the eyes of lover, a least for a moment. This flexibility helps us organize our roles and make use of the discretion we are sometimes afforded, in permissive social contexts, to ‘be ourselves’. It is not to be confused with what MacIntyre understands disparagingly, by ‘flexibility’, which he describes as the virtue of ‘knowing chameleon-like how to take on the color of this or that social background. [Someone] exhibits this virtue in managing [their] transitions from one role to another, so that [they appear], so far as possible, to be dissolved into [their] roles’ (1999, 325–326). Someone who compartmentalizes their multiple roles, and manages them with the kind of boundary work I praise, need not take on the colour of their social background, or dissolve into a role. Some contexts may allow for such immersion, others not. The norms, or comprehensive goals, of some roles may be endorsed, others not. The task of managing roles is an ongoing one, subject to change of perspective and deepening understanding, especially since roles, and their institutional contexts, also change over time. To risk an analogy to which I am not entitled, the virtuous compartmentalizer resembles a musician improvising with others. This musician occupies their role, which is defined by rules and boundaries. In turn, those rules and boundaries make sense in the broader context of the group’s musical goals. The musician plays their part, within these constraints, but with some freedom to deviate when contextually appropriate, and the context is changing all the time. Most importantly, they are absorbed into their playing, but not dissolved into it; they listen, and respond, to their bandmates, to the music, to the room.⁶ The virtuous compartmentalizer is attentive to their roles as sources of constraint to be embraced or reacted to; evaluates the placement and transcendence of role boundaries against the social goals and goods of their roles; acknowledges those goods and goals can be reinterpreted; recognizes that the salience of a role’s ⁶ Here I am indebted to Gary Hagberg’s fantastic article on ‘Jazz Improvisation and Ethical Interaction’ (2008). Of jazz improvisation he writes, ‘it is an art form that itself constitutes a welcome antidote for the trivialization of complexity and impatience with particularity. What, jazz improvisation asks, is happening in this moment, with these people, in this setting, under these conditions? Given that this musical gesture has just been made in this circumstance, what is possible now?’ (2008, 282).

318



goal or good can change when viewed from the perspective of another role; and is sensitive to the particular discretion they have to manage their roles, and the limits of their character. This view can accommodate the fact that boundary work can shift and change over time. For example, we might become more integrating as we become more proficient; or as our role changes; or the institutional context changes; or as we shed additional roles and can thus accommodate more integration. This view also gestures towards a recipe for avoiding alienation which stresses the dynamic options for relating to the roles we have over attempts to retreat to milieus where roles are shed.

7. Conclusion We have many roles, and in so doing our expanded selves benefit from resources, experiences of value, and different agential mindsets. Compartmentalization helps us manage the demands of our roles and, if done well, does not risk failures of integrity, moral blindness, or alienation. Maintaining clear boundaries between roles and their attendant mindsets enables us to be guided by fully specified virtues, and flexible boundary work can help us avoid inappropriate forms of immersion in roles at the expense of others. Chameleons we may be at times, but virtuous ones.

References Adams, R. M. 2006. A Theory of Virtue: Excellence in Being for the Good. Oxford: Oxford University Press. Aron, A., Lewandowski Jr, G. W., Mashek, D., and Aron, E. N. 2013. ‘The SelfExpansion Model of Motivation and Cognition in Close Relationships’. In Jeffry A. Simpson and Lorne Campbell (eds.), The Oxford Handbook of Close Relationships. Oxford: Oxford University Press, pp. 91–115. Brewer, T. 2009. The Retrieval of Ethics. Oxford: Oxford University Press. Brunning, L. 2018. ‘Cultivating an Integrated Self ’. In M. Dennis and S. Werkhoven (eds.), Ethics and Self Cultivation, Historical Perspectives. London: Routledge, pp. 174–196. Caper, R. 1999. ‘Psychoanalysis and Suggestion, Reflections on James Strachey’s “The Nature of the Therapeutic Action of Psychoanalysis” ’. In A Mind of One’s Own, London: Routledge, pp. 9–18. Dahrendorf, Ralf. 1968. ‘Homo Sociologicus’. In Essays in the Theory of Society. London: Routledge & Kegan Paul, pp. 19–88. Goffman, E. 1990 [1959]. The Presentation of Self in Everyday Life. Harmondsworth: Penguin.

 

319

Haberg, Gary. 2008. ‘Jazz Improvisation and Ethical Interaction’. In G. Hagberg (ed.), Art and Ethical Criticism. Oxford: Blackwell Publishing, pp. 259–285. Hardimon, M. O. 1994. ‘Role Obligations’. The Journal of Philosophy 91 (7): 333–363. Hochschild, Arlie. 2003. The Managed Heart, Commercialization of Human Feeling. 2nd edn. Berkeley: University of California Press. Jaeggi, Rahel. 2014. Alienation. New York: Columbia University Press. Korsgaard, Christine. 2009. Self-Constitution, Agency, Identity, and Integrity. Oxford: Oxford University Press. Lear, Jonathan. 1998. Love and Its Place in Nature. New Haven: Yale University Press. Lear, Jonathan. 2003. Therapeutic Action. New York: Other Press. MacIntyre, Alasdair. 1999. ‘Social Structures and Their Threats to Moral Agency’. Philosophy 74 (3): 311–329. Nguyen, C. Thi. 2020. Games: Agency as Art. Oxford: Oxford University Press. Nippert-Eng, Christina E. 1996. Home and Work: Negotiating Boundaries Through Everyday Life. Chicago: University of Chicago Press. Nordenmark, M. 2004. ‘Multiple Social Roles and Well-Being: A Longitudinal Test of the Role Stress Theory and the Role Expansion Theory’. Acta Sociologica 47 (2): 115–126. Oakley, Justin and Cocking, Dean. 2001. Virtue Ethics and Professional Roles. New York: Cambridge University Press. Phillips, A. 2014. Interview. The Paris Review no. 208. Scheffler, Samuel. 2010. Equality and Tradition: Questions of Value in Moral and Political Theory. Oxford: Oxford University Press. Scherkoske, G. 2013. Integrity and the virtues of reason: Leading a convincing life. Cambridge University Press. Sciaraffa, Stefan. 2011. ‘Identification, Meaning, and the Normativity of Social Roles’. European Journal of Philosophy 19: 107–128. Solomon, Robert. 1997. ‘Corporate Roles, Personal Virtues: An Aristotelian Approach to Business Ethics’. In Daniel Statman (ed.), pp. 205–226. Virtue Ethics: A Critical Reader. Edinburgh: Edinburgh University Press. Swanton, Christine. 2007. ‘Virtue Ethics, Role Ethics, and Business Ethics’. In Rebecca L. Walker and Philip J. Ivanhoe (eds.), pp. 207–244. Working Virtue. Oxford: Oxford University Press. Tiberius, Valerie. 2008. The Reflective Life: Living Wisely With our Limits. Oxford: Oxford University Press.

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. academic 10–11, 187, 189–90, 194–5, 203–4, 223–4, 292, 299, 302, 305, 311 achievement, individual and collective 139, 203, 272, 285–6 Achilles 25–47 affect 53, 56, 60–1, 67, 76–7, 117, 226–8, 232–3, 299n.1, 301–3, 314–15 love in friendship 225–9 Løgstrup on loving thy neighbour 92–3 agency, roles as a medium of 34, 49, 61, 67, 134, 137–8, 143, 278, 303–5 alienation 256–60 see also compartmentalization of selves, merits and drawbacks of; waiter (Sartre’s example) Analects 53, 62–3 Analects for Women 57–8, 60–3 aristos as a role 34 Aristotle’s function argument 191–3 Ban Zhao, see Lessons for Women bindingness of roles: see normativity of roles, source of its content and force biography 43–4, 263–76 Biographies of Women (Liu Xiang) 58 see also martial autobiographies Bradley, F. H. 1, 72n.2, 88–9, 90n.7 careers 2, 249, 265–6 see also professional roles characteristic activity of institutions 190–5 citizen as a role 11–12, 49, 98, 187, 200, 209–11 class and role differentiation 138, 140, 184 codification of roles 7, 10–11, 59–61, 128, 179–81, 299 coerciveness of roles: see oppressiveness of roles and voluntariness of roles collective contexts and agency, roles and 15, 139–40, 150, 200–3, 281 compartmentalization of selves, merits and drawbacks of 305–9 and alienation 308–12 and boundary work 305–7, 316–18 and reflection 307, 315–16 moral risks of and integrity 307–8, 312–15

conflict between multiple roles: see friction between and juggling of roles conflict between roles and morality: see morality, exhausted by or in conflict with role morality Confucian tradition of role ethics 49–64, 88n.2 Confucian role ethics (CRE) 49 contrasted with Western tradition 49 neglect of women by 53–4 neglect of women contributors to 52–3 conventional norms 107–13 as theorems 113, 127–8 as binding by default 122–4 conventions 108 minimal theory of 114–18 defective roles 8–9, 36, 111, 129–30, 170–1, 223–4 ameliorating 130, 145–53, 200–15 see also privileging and disadvantaging nature of some social roles; slaves and slavery definition of social role 4–9, 12, 135–9, 177, 219–22, 238 and paradigm cases 4–5, 25, 107–8, 221–2 Dewey, J. 66–84 dirty hands 14, 16 division of labour 67, 143, 201–2, 204, 206, 248 doctor 55, 126 see also moral monsters emotion: see affect ergon: see characteristic activity of institutions and Aristotle’s function argument etymology of ‘person’ 2–3, 270 of ‘role’ 3 expanding self 300–5 see also self-realization; self-transformation externalism and internalism about reasons 27, 37, 45–6, 156–8 evil roles: see defective roles familial roles and relationships 4–5, 59–61, 138, 184–8, 194, 204–5, 219–20, 278, 290 see also in-law relationships

322



feminist philosophy, roles considered by 43–4, 52, 63, 141, 148–9, 224 friction between and juggling of roles 11, 18, 70, 81–2, 298–318 and unity of self in Dewey 75–6 friendship 6, 8, 15–16, 18, 53, 73–4, 100–1, 179n.7, 217–34 as involving intimacy 227, 229 characteristics of 226–7 functions of 227–8 not a social role 230–3 obligations arising from 228 function of a role 8–9, 299–300 and function of institutions 178 comparison with biological and artefact functions 159–60, 183, 186, 238, 249 distinguished from function of the occupant 239, 254 function-first approach 8–9, 178 of occupational roles 238–41 function of institutions 177–96, 207–8, 240 as a guide when roles generate moral conflict 209 etiological approach to 183–5 morally suspect instances of 211–14 teleological approach to 187–90 use-function approach to 185–7 see also characteristic activity of institutions function of nuclear family 185–6 gender roles 11–12, 15–18, 49–64, 140, 224 and biology 43–4, 59, 231–3 see also social justice and roles; women’s conduct Goffman, E. 15–16, 26n.1, 71, 258 Hardimon, M. O. 4–5, 136, 177, 179–80, 204–5, 219–23, 263 Hegel 88 see also Sittlichkeit heroism and hero as a role 28–37, 41–2, 46, 79, 268, 276 hierarchy in role relations 59–60, 111, 134–5, 137–8, 143–4, 201–2, 204 see also knowing one’s place; oppressiveness of roles Homer’s Iliad 25–47 I vs. Me in American pragmatism 70–1, 73, 75 impartiality 15, 78–81 see also friendship; familial roles and relationships in-law relationships 53–7 see also marriage; familial roles and relationships

indeterminacy of a role’s prescriptions 37–8, 97–8, 177, 179–83, 195, 249 individualism 1, 63, 139, 282–3 institutional context of roles 10–11, 13–14, 18, 100–1, 135–8, 177–96, 200–15, 251, 256, 278, 290 integrity: see compartmentalization of selves, merits and drawbacks of intelligibility of an action as dependent on role 72–3 interdependence of conventional norms, including roles 116, 120 internalism about reasons: see externalism and internalism about reasons internalization of a role 69–70, 97–9, 231, 301 see also waiter (Sartre’s example) interpretive character of roles 177, 180 involuntary roles: see voluntariness of roles is-ought gap 37–9, 41–2, 46n.27, 246 Kant and Kantianism 63, 70–1, 79, 88, 213, 271 knowing one’s place 1, 89, 134–7, 142–4, 150, 224 lawyer 107, 180–3, 302 Lessons for Women 54–8, 60–1 liberalism 1, 6, 263 Løgstrup, K. E. 87–101 his ethical demand 92–9 on internalization of a role 97–9 on law, morality, and convention 93–5 on parenting 97–8 MacIntyre, A. 27–40, 68–9, 266n.2, 306–9, 315–17 marriage 10–11, 54–5, 146–7, 177–8, 186–7, 223–4, 230–1 see also in-law relationships martial autobiographies 265–73 masks and personas 2–3, 256–8, 270–1 master roles 161–3, 166–7 Mead, George Herbert 66–73 Mencius 50, 53, 61, 63 moral agency 66–84, 109, 112–13, 139–40 Confucian 49–52, 59–60, 63–4 of organizations 213–15 moral monsters 87, 90–2 see also defective roles morality, exhausted by or in conflict with role morality 50, 90, 100n.12, 109, 111–12, 181, 200–1, 205–14 multiple roles, occupation of: see friction between and juggling of roles names given to roles 10, 237, 240–1 narratives 30–2, 42–3, 269–70, 301 naturalism 83, 169, 202, 236–8

 nei–wai distinction 59–61, 59n.4 normativity of roles, source of its content and force 13, 17–18, 77–81, 101, 107–30, 243–4, 290 see also reason-givingness of roles obligation to alter one’s own role 200–15 obligation to resist oppression in view of one’s role 147–8, 152–3 occupancy conditions of a role 5–6, 133 as entrance and exit conditions 6, 157, 288–9 oppressiveness of roles 1–2, 47, 51–2, 257 see also defective roles; social justice and roles organizations: see institutional context of roles participatory wellbeing 278–95 Bradford, G., on 285–8 vectors of 291–4 Paul L. A. 2, 273–5 performance conditions of a role 7–8, 133 as obligations and entitlements 7, 135, 289 blurred relation to occupancy conditions 8 persistence of a role having different occupants 6, 219, 241 persona: see masks and personas personhood 3–4, 49–50, 61–3, 69 and persona 2–3 see also virtues and ‘establishing the person’ positions (contrasted with jobs) 250–6 power and inequality 51–2, 82–3, 138, 143 see also hierarchy in role relations; knowing one’s place; social justice and roles pragmatism, American 66–84 privileging and disadvantaging nature of some roles 134–5, 138 see also defective roles; social justice and roles pro tanto norms 119, 246 professional roles 10–11, 59–61, 100–1, 137, 178–9, 236–60 see also academic; doctor; lawyer; soldier; teacher public nature of knowledge of roles 25–6, 110–11, 118, 204, 206 see also standpoint (epistemological) race, roles and 4–5, 140–1, 148–9 see also social justice and roles Rawls 1n.1, 112, 178n.2, 280–1 realization of roles 11–12 recognition requirement for 25–6 reason-givingness of roles 25, 35–6, 45, 156–72, 222–4 hierarchical or non-hierarchical accounts of 161 indirect and direct accounts of 160–1 see also normativity of roles, source of its content and force

323

reflective detachment from one’s role 28–31, 34–5, 76 see also role-recalcitrance; standpoint (epistemological) role ethics (as a theoretical perspective as opposed to as a topic area) 49, 88 role pluralism 158, 163–5, 167–70 role-recalcitrance 25–47, 249, 307 rolescape 11, 16, 82–3, 188–9 self and self-conception 2–4, 14–16, 18, 63, 69–70, 271–2, 300–5 as constituted by roles occupied 49–50, 64, 251 see also waiter (Sartre’s example) self-discovery 264, 268–73 see also self and self-conception self-realization 74–6, 90, 264, 268–73, 280–1 and self-formation in Mead 74 as requiring removal of mask 270 see also self and self-conception; expanding self self-transformation 265–6, 273–6 rationality of 273–6 see also Paul, L.A. Shakespeare’s As You Like It 5–7, 11 Sittlichkeit 42, 88 slaves and slavery 14, 45, 81–2, 229n.23 see also defective roles ‘social’ in ‘social role’ 9 social justice and roles 14, 16, 82–3, 133–53, 295 social-group membership as a role 142, 144–8, 150, 153 and good allyship as an obligation 149–50 and moral burden 147–8 sociology 1–3, 8–9, 41, 66–9, 71–2, 83–4, 201n.1, 300, 305 see also Goffman, E. soldier 27, 265, 269 and monasticism 266 Song Sisters: see Analects for Women special relationships 15 Confucian conception of 49 see also familial relationships; friendship standpoint (epistemological) 28–30, 34–5, 46–7, 130, 135, 140–1, 151, 316 status and social standing 80, 141, 295 see also hierarchy in role relations teacher 10, 26–7, 182, 240–9, 310 Thomson, J. J. 236–8 tools, roles as useful 66–7, 76–8, 94–5, 110, 143 for generating sui generis goods 266–7 for satisfying pre-existing desires 264–5

324



tools, roles as useful (cont.) for self-revelation and self-realization 268–73 for grasping values 302–3 see also division of labour; selftransformation transformation: see self-transformation utilitarianism and utility 78–9, 88–9, 115, 211, 226n.17, 280n.2 maximizing expected utility 273–4 value, roles as a guide to 302–3 virtue ethics 49, 88, 179n.5, 312 virtues 80–1 and ‘establishing the person’ 61–3 and unity of self 80–1

contrasted with roles and relationships 50–2, 56 see also virtue ethics voluntariness of roles 6, 36, 46–7, 89, 100–1, 138–9, 224–5, 232–3, 243–4 waiter (Sartre’s example) 15–16, 28n.5, 244, 256–8 wellbeing, roles and 14–16, 18, 89, 263–76, 279–83 and self 272–3 see also participatory wellbeing women’s conduct 40, 57 see also Lessons for Women Zeus 30–2