Justifying Emotions: Pride and Jealousy Kristjan Kristjansson

J U S T I F Y I N G E M OT I O N S

Justifying Emotions is a topical and controversial discussion of the ethical and moral problems surrounding emotions. Kristján Kristjánsson challenges the usual view of emotion as a negative influence on the formation of proper moral judgement, using pride and jealousy as examples of two emotions that are essential to harmonious human existence. He argues that experience of the traditionally ‘negative’ emotions of pride and jealousy is not evidence of moral failing, but rather that these supposed vices contribute to a well-rounded, virtuous life.

The book begins with a critical introduction to cognitive theories of the emotions, before going on to consider the place of the emotions in moral theories such as utilitarianism and virtue ethics. A discussion of the nature of moral and emotional excellence is followed by detailed defences of both pride and jealousy. A final chapter is dedicated to issues surrounding the teaching of virtue and the education of the emotions.

Kristjánsson’s first aim in this book is to explore the moral justification of emotion, and the link between this justification and the notions of moral and emotional excellence. His second aim is to give a more sympathetic hearing to the emotions of pride and jealousy, arguing that a certain kind of pride is actually necessary for personhood and that jealousy is necessary to maintain pride and self-respect. Kristjánsson concludes that not experiencing the emotions of pride and jealousy, when called for, would be evidence of a moral failing.

Justifying Emotions: Pride and Jealousy makes a thought-provoking and practical contribution to the current debate on the emotions and is sure to spark greater concern about the ‘negative’ emotions in general. It will be of interest to the general reader, in addition to students and professionals working in the areas of philosophy, psychology and education.

Kristján Kristjánsson is Professor of Philosophy at the University of Akureyri, Iceland. He is the author of Social Freedom: The Responsibility View (1996).

ROUTLEDGE STUDIES IN ETHICS AND

MORAL THEORY

1 THE CONTRADICTIONS OF MODERN MORAL PHILOSOPHY

Ethics after Wittgenstein

Paul Johnston

2 KANT, DUTY AND MORAL WORTH

Philip Stratton-Lake

3 JUSTIFYING EMOTIONS

Pride and Jealousy

Kristján Kristjánsson

J U S T I F Y I N G

E M OT I O N S

Pride and Jealousy

Kristján Kristjánsson

London and New York

First published 2002

by Routledge

11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canada

by Routledge

29 West 35th Street, New York, NY 10001

Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2004.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Kristján Kristjánsson.

Justifying emotions : pride and jealousy / Kristján Kristjánsson.

p. cm. – (Routledge studies in ethics and moral theory) Includes bibliographical references and index.

1. Emotions (Philosophy). 2. Pride. 3. Jealousy. 4. Ethics. I. Title. II. Series.

BJ1535.P9 K75 2001

179'.8–dc21

2001034890

ISBN 0-203-16588-8 Master e-book ISBN

ISBN 0-203-26053-8 (Adobe eReader Format)

ISBN 0–415–26667–X (Print Edition)

FOR MY SON, HLÉR, WHO HAS TRIED HIS

BEST TO TEACH ME PATIENCE AS A

MORAL AND EMOTIONAL VIRTUE, AND

MY WIFE, CHIA-JUNG

CONTENTS

Preface

Mapping out the field

1.1 Introduction 1

1.2 Cognitive theories and their precursors 8

1.3 Taking stock: some critical comments on cognitivism 20

1.4 Preliminary remarks on responsibility, moral justification, and the ‘negative’ emotions 36

Justifying emotions: the need for moral theory 49

2.1 Human nature as the foundation of moral theory 49

2.2 The shortcomings of virtue ethics: moral and emotional conflict 63

2.3 Utilitarian naturalism and the emotions: an untapped source 76

Something to be proud of: the nature and conditions of moral and emotional excellence

3.1 Personhood, integrity, and self-respect 91

3.2 Aristotle’s megalopsychia 99

3.3 Pridefulness: pride and shame 104

In defence of pridefulness

111

4.1 The value of pridefulness 111

4.2 The dependence upon luck 119

4.3 The extra value of the extraordinary 126

4.4 Moral equality, modesty, and humility 130

vii

C O N T E N T S

In defence of jealousy

136

5.1 Jealousy as a type of envy 136

5.2 Contrasting views 145

5.3 The peculiarities of sexual jealousy 155

5.4 Jealousy as a virtue 161

Teaching emotional virtue

170

6.1 Educating emotions 170

6.2 Why all the lingering doubt? 179

6.3 Didactics 187

6.4 Teaching the virtues of pride and jealousy 196

Concluding remarks

205

Notes

211

Bibliography

242

Index

253

viii

P R E FAC E

Whenever people ask me ‘When did you first become interested in emotions?’, I give the pat answer ‘Pretty soon after I was born’. This terse reply does not rest so much on platitudes about man being by birth a pondering animal, sparing no pains to dig out philosophical and scientific truths, as on a much simpler observation: Everybody is interested in the emotions for they constitute a core ingredient, if not the essence, of human life. However, the question should perhaps be understood in a narrower sense to mean ‘When did you first become academically interested in emotions?’ In that sense, a truthful response requires a piece of philosophical autobiography.

The development of my academic interest in the emotions coincided with a growing disillusionment with certain trends in contemporary political philosophy, the field to which I had devoted much of my academic attention since the completion of my doctorate. Consider a group of well-educated people of different nationalities or ethnicity sitting together in a street café, convincing each other – with mutually understandable arguments – of the essential impossibility of mutual understanding, and you have a striking, if a little over-simplified, image of much of what has been going on in political philosophy of late. Let it suffice here to say that such philosophy does not offer proper sustenance for one, such as the present author, who is an Aristotelian naturalist at heart, a universalist, and an inveterate believer in the ‘Enlightenment Project’. Nor does its lack of serious engagement with foundational conceptual issues give satisfaction to one who considers the analytical way of doing philosophy the remnant of a certain passionate seriousness which has gradually been disappearing from many other ‘traditions of inquiry’ or – to use more fashionable jargon – ‘discursive fields’.

Allow me to be even more personal here. I imagine that we are all familiar with the perennial question ‘What book would you take with you if you had to stay for a year on a desert island?’ Arguably, I answered that question for myself a few years ago, through action rather than words, when I chose a book to accompany me on a long journey. The destination was admittedly not a desert island, yet it was a place where experience had taught me that ix

P R E FA C E

my verbal communication with the locals would be scant. I was going to Taiwan to attend my father-in-law’s funeral; the book, Andvökur [ Wakeful Nights], was a selection from the works of the celebrated Icelandic poet Stephan G. Stephansson (1853–1927) who spent most of his life in rural Canada but wrote his verse in Icelandic. During the preparatory days of chanting, while tonsured monks and nuns read from right to left about Buddha’s omniscience, I read Stephansson’s poetry from left to right.

Perhaps it was the climate’s fault; his verse warmed the cockles of my heart as never before. What is more, an important truth was borne in on me, one closely connected to the ambience of the place. Despite the smog of urban Taiwan where, as the farmer Stephansson would have put it, no ‘unclouded eastern sun / blazes up glen and grade’, and ‘the space of life’ is ‘narrowed in every respect’, I felt as if the poet’s spirit imbued every person in sight, be it the chanting monk, the nouveau riche businessman, or the street vendor. His insights and emotions were with us and in us, in them as well as in me. I have never sensed the presence of an inter-human denominator as strongly as during those days in Taiwan. I realised that if it was myopic to think that the justification of social arrangements could only appeal to a limited group of ‘us’, for instance ‘us in Western liberal democracies’ as now seems to be the received wisdom in political philosophy, then such relativism would be even more ill-considered in the case of human emotions which surely are the same all over the world. I decided that if I should ever put pen to paper to discuss emotions, my task would be essentially the same as that of Stephansson – himself an avowed son of the Enlightenment – to try to convey the universality of the nature and justification of our emotions; to try to capture, or if necessary re capture, their ‘root flavour’. My fundamental acknowledgement, at the beginning of this book, must go to Stephan G. Stephansson.

As a matter of fact, I had first thought of writing about jealousy and pride during my postgraduate days in St Andrews. I am grateful to my erstwhile supervisor Gordon Graham for sounding all the correct warning signals at that time and persuading me to postpone the project. I picked up the thread again during my sabbatical year at the University of East Anglia in 1996–7. I am indebted to Alec Fisher who at that time chaired the philosophy section, and most particularly to Martin Hollis who showed an acute interest in my exploration of concepts of self-assessment and encouraged me to write about them, but sadly died before I could present him with any results. His alert mind and incisive comments would, I am sure, have helped me steer clear of many subsequent errors.

Work on the present book began during my sabbatical stay at the University of Konstanz in the spring semester of 1999. Special thanks must go to Professor Gottfried Seebaß for providing me with office space and access to the wonderful library there, and to his co-workers, Margit Sutrop and Holmer Steinfath, for stimulating discussions. Comments and criticisms x

P R E FA C E

received when drafts of particular sections were presented at departmental seminars in Konstanz and at the Technical University in Darmstadt, as well as at the 2000 Conference of the British Society for Ethical Theory, proved most helpful. I would also like to thank the Philosophy Department at Cornell University, in particular Professors Terence Irwin and Gail Fine, for inviting me as a Fulbright Research Fellow in the spring of 2001 and for providing me with the facilities which enabled me to put the finishing touches to this book.

I gratefully acknowledge Professor Aaron Ben-Ze’ev’s early and continuing encouragement, and his generously extended comments on an early draft of the whole book. My friend Barbara B. Nelson tried her best to eliminate any infelicities of language from the text, and provided some valuable philosophical insights along the way, and my ex-student Björn Sigurðarson (who unfortunately opted for a career in computing rather than philosophy!) provided me with invaluable editorial assistance. My intellectual debt to my friend and mentor Mikael M. Karlsson should not go unmentioned here, nor the comments and counsel of my friends and colleagues Guðmundur Heiðar Frímannsson and Haraldur Bessason at the University of Akureyri.

The following academics in the Faculty of Social Sciences, University of Iceland, read and patiently commented on particular chapters: Jakob Smári and Sigurður Júlíus Grétarsson (ch. 1), Andri Steinþór Björnsson (chs 1–2), and Sigrún Aðalbjarnardóttir (ch. 6). Among the various other people who have at one time or another read and/or advised me on particular sections and issues are: Atli Harðarson, Ólafur Páll Jónsson, Logi Gunnarsson, Róbert H. Haraldsson, Vilhjálmur Árnason, Dan Farrell, and referees and editors of many of the journals listed below.

While I have profited greatly from the advice of all the persons mentioned above, none of them should be taken to endorse what I argue for in this book: too often, I have foolishly resisted the changes that they urged on me.

The Icelandic Council of Science, the Research Fund at the University of Akureyri and the Iceland–US Educational Commission (Fulbright Visiting Scholar Program) did not provide me with any academic advice, but they deserve thanks for financing parts of my research.

Many of the ideas expressed in this book started to take shape in articles that I wrote in Icelandic which were published in two collections of my philosophy papers: Þroskakostir [ Ways to Maturity] (1992), and Af tvennu illu [ The Lesser of Two Evils] (1997). Different sections of the book incorporate material already published in the following articles (see bibliography for further details): ‘Why Persons Need Jealousy’, The Personalist Forum (1996),

‘Stephan G. Stephansson: A Philosophical Poet, a Poetic Philosopher’, Canadian Ethnic Studies (1997), ‘Casual Sex Revisited’, Journal of Social Philosophy (1998), ‘Self-Respect, Megalopsychia, and Moral Education’, Journal of Moral Education (1998) (with kind permission from Taylor & Francis Ltd), ‘Liberating Moral Traditions: Saga Morality and Aristotle’s xi

P R E FA C E

Megalopsychia’, Ethical Theory and Moral Practice (1998) (with kind permission from Kluwer Academic Publishers), ‘Stórmennska’ [ Megalopsychia], Skírnir (1998), ‘A Prolegomena to “Emotional Intelligence”’, Philosophy in the Contemporary World (1999), ‘Liberalism, Postmodernism, and the Schooling of the Emotions’, Journal of Thought (2000), ‘Virtue Ethics and Emotional Conflict’, American Philosophical Quarterly (2000), ‘Teaching Emotional Virtue: A Post-Kohlbergian Approach’, Scandinavian Journal of Educational Research (2000) (with kind permission from Taylor & Francis Ltd), ‘Utilitarian Naturalism and the Moral Justification of Emotions’, International Journal of Applied Philosophy’, (2000), ‘The Didactics of Emotion Education’, Analytic Teaching (2000), and ‘Some Remaining Problems in Cognitive Theories of Emotion’, International Philosophical Quarterly (2001). I am grateful for the permission to reprint material from those sources.

Last, but not least, I would like to thank my wife, Chia-jung Tsai, for her care and advice, and my son, Hlér, to both of whom this book is dedicated.

xii

M A P P I N G O U T T H E F I E L D

1.1 Introduction

My boss’s nephew, and incidentally also a good friend of mine, gets promotion in my company just because he is the boss’s nephew, while I, a much better qualified candidate, am left behind to sweat in a low-ranking post. At the school fair, only Kate’s poster is chosen to hang in the hall although Betsy’s poster was at least equally well done. Cindy, who catches her lover in flagrante delicto with another girl, feels her world has crumbled to dust.

Are Cindy, Betsy, and I morally justified in experiencing jealousy or is that emotion invariably the sign of a malicious mind? Jack has made considerable personal sacrifices to help an ailing relative whom others, nearer and closer to the poor fellow, had left in the lurch. Is Jack morally justified in taking pride in his deeds, as well as expecting and demanding some external recognition of his efforts, or would that be the sign of a deadly sin?

This book has two main objectives. The first is to explore what, generally speaking, constitutes a moral justification of an emotion and how such a justification is connected to the notion of moral and emotional excellence.

The second is to give the two emotions that figure in the examples above, both traditionally vituperated as psychologically debilitating and morally flawed, a more sympathetic hearing. These emotions are pride – or pridefulness as I shall call it, by focusing on one of the many different uses of the word ‘pride’ – and jealousy. How often have we not heard pride proscribed as the root of all vice, and jealousy as one of its distasteful concomitants?

My aim is, by contrast, to show that we have been much too hard on emotions which we have not properly understood, and that both these so-called negative emotions can, in the proper dosage, be seen as virtues or as ingredients in virtues: as parts of a good human life. To put it as succinctly as possible, I challenge the received wisdom about pride by claiming that a certain kind of pride, namely pridefulness, is psychologically necessary for the formation and sustenance of personhood, and also morally necessary for a self-respectful person who wants to live a well-rounded virtuous life. In addition, I argue that jealousy is necessary to maintain pride and self-respect.

M A P P I N G O U T T H E F I E L D

Although I shall avoid using the cumbersome and semi-technical term

‘pridefulness’ in the sequel, where it is possible to do so without causing misunderstanding, it should be made clear from the start that my ultimate defence is of pride qua pridefulness (see ss 3.3–4.1).

As I noted in the preface, everybody is interested in the emotions. From an early age, our own and others’ emotions take up a substantial part of everyday conversation; they guide our actions, inform our evaluations, and kindle our interest in the mundane and the sublime: in everything from eating porridge to enjoying art. Without emotions there might be a number of Mr Spocks of Star Trek fame around, but surely no human beings.

Fortunately, in recent years philosophical interest in the emotions has reached new heights after sinking to its nadir for decades. Although the emotions, in general, have received renewed attention, much less has been written about the ‘negative’ emotions specifically – and by ‘negative’ let me here tentatively mean those emotions typically evaluated negatively from a moral perspective; there are other uses abroad as we shall later learn. A notable exception is Gabriele Taylor’s Pride, Shame, and Guilt, an insightful if somewhat disconcerting study of those three emotions and their interrela-tionships.1 A few papers have appeared about specific ‘negative’ emotions, such as Daniel M. Farrell’s important analyses of jealousy and Jerome Neu’s recent reappraisal of pride.2 Most of these studies, however, have been primarily conceptual and have not come much to terms with the substantive moral standing of the ‘negative’ emotions in question, except as a side issue.

The book which perhaps comes closest to my orientation, and from which I have learnt a great deal, is John Casey’s Pagan Virtue,3 but he is more concerned there with the traditional virtues and vices than with (their connected) emotions. In addition, there is the steadily growing mountain of literature on ‘emotional intelligence’, but it tends, typically, to evade questions of the value or disvalue of ‘negative’ emotions. The imbalance of these evasions must, I think, be rectified.

In the field of emotion research, too many a cobbler has tenaciously stuck to his last: a lamentable state of affairs in a field that cannot and should not by its very nature be the privileged domain of any one discipline. Unfortunately, philosophical and psychological explorations of the general nature of emotions, or even of the same specific emotions, frequently seem to run on parallel tracks with only the barest mutual acknowledgement. The different camps seem, so to speak, to be building similar pyramids on both sides of a huge ocean without much idea of what is happening on the other side.4 Since I think of the field of emotions as a buffer zone into which incursions can and should be made from various sides, the focus of the present book is interdisciplinary. I do not intend to hide that I am a philosopher; that will be amply evidenced by the tenor of my discussion and the choice of topics. For instance, I believe that my understanding of emotions has benefited from my earlier engagement with 2

M A P P I N G O U T T H E F I E L D

political issues, and that an influx of ideas from political philosophy as well as from ‘pure’ moral theory can aid us in looking at particular emotions from a moral point of view. After all, no moral concerns are completely apolitical and no political ones are extra-moral. This belief will be reflected in the book’s approach. However, I want to bring to bear as many insights from other areas as possible: from psychology, education, literature, and, last but not least, anecdotal evidence from daily life. I believe, like Heraclitus, that even in the kitchen ‘divinities’ are present.5 Most of the important lessons of life I have learnt at the kitchen table: first in my parents’ house; later with my wife. It is also more than a half-truth that the best descriptions of emotions tend to be provided by artists rather than academics.6 More perhaps can be learnt about the subtleties of emotions by a careful reading of the poetry of Stephan G. Stephansson,7 or the novels of Dostoevsky, than through any scholarly treatise. I shall be adding tonal shadings from various literary sources to my arguments. However, among my frequent fellow travellers will also be the academic philosopher par excellence, Aristotle. Generally speaking, the reader will, I believe, gain more from a synthetic, interdisciplinary approach than from a more narrowly-defined focus, even if that may at times mean sacrificing depth for breadth. Moreover, since I consider my exploration of specific emotions and their cultivation to be of general importance, I hope that my book will be accessible to more readers than those who are already well versed in psychology or philosophy.

My interest in the emotions has been heightened by my employment as a professor of philosophy in a recently established department of education with a strong philosophical orientation. In such a department, questions about the emotional schooling of the young are continually relevant and pressing. We need to know what the emotions are, how they are formed and cultivated and, in so far as they are psychologically under our jurisdiction, which of them are morally justified. ‘Is children’s jealousy amenable to any rational control?’, ‘Should pride be nourished or uprooted?’, ‘Under what circumstances, if any, can envy be non-malicious?’ are just a few examples of the questions crowding in. When preparing lecture notes for my classes in moral education, I realised how the old ideal of education as character formation had given way in educational theory to scepticism about moral education in general and schooling of the emotions in particular. This scepticism seemed to be propagated by a wave of psychological theories that either postponed the systematic cultivation of morally commendable emotions to ‘later stages’ in the child’s life, or simply reduced the emotions to steam rising up from internal kettles, mostly, if not wholly, impenetrable to reason. The subject of moral and emotional education had not only not been learnt in recent years, but had rather, in a fundamental sense, been unlearnt by teacher-training students, leading to teacher neutrality and parental uncertainty on these important issues. David Carr aptly refers to 3

M A P P I N G O U T T H E F I E L D

this situation as a ‘conspiracy of silence’8 (see further in s. 6.4). What the textbooks in pedagogy and educational psychology provided for my students about the nature and cultivation of emotions was so meagre that, without other resources, they were bound to starve on it. That this Spartan diet was not what famous educational theorists such as Piaget or Kohlberg had intended is more or less beside the point. It is often more relevant in practice to consider what the secondary (or even the watered-down ‘tertiary’) literature tells university students about the ideas of important thinkers than to concentrate on what may be revealed by careful scholarly exegesis of these ideas themselves. I would not have written the present book if I did not believe that at least some of the things I have to say are of practical significance for moral education.

Philosophical views tend to be shackled together with the heavy chains of social and personal history. It is probably not just coincidental that my historical background, as an Icelander, is in a proud, assertive ethical tradition of ‘saga morality’9 which accepts other-regard and self-regard as necessarily intertwined and understands morality as rules of demeanour and conduct in a society made up of free, sovereign persons. If we bring up the question of the moral justifiability of emotions as a question about what it really means to be a person, that is, one who can make claims, who can incur and acknowledge obligations, can be wronged, can be the object of and can reciprocate love, respect, hatred, contempt, etc.,10 then a range of interesting considerations starts to emerge that may threaten those accounts which automatically saddle pride and jealousy with a bad name. In this emphasis on the formation, maintenance, and recognition of personhood we can see the glimmering of a point that will be variously explored and pressed in the sequel.

For convenience of exposition, I shelve a direct defence of the emotions of pride(fulness) and jealousy themselves until chapters 4 and 5. The rest of chapter 1 will instead concentrate on a number of preliminaries. A writer wrestling in vacuo with issues such as the moral legitimacy of jealousy, or the various emotional manifestations of love, is likely to encounter severe trials. To make progress with such an inquiry, his general point of departure must be clear; he must have a fair idea of what an emotion is. I proceed in the following section with the briefest of surveys of recent emotion theory, especially of the cognitive kind to which I essentially adhere. Rather than opening new vistas, that section aims at a short survey of the present state of research. A major critical synthesis of current research is to be found in Aaron Ben-Ze’ev’s recent book, The Subtlety of Emotions.11 I refer the reader to that work for a clear, interdisciplinary overview of emotion theory and penetrating studies of various particular emotions. To avoid longueurs about general issues in emotion research, I mostly confine my discussion of that area to those points where I take exception, or believe that I have something to add, to Ben-Ze’ev’s treatment. Notice that section 1.2 is primarily 4

M A P P I N G O U T T H E F I E L D

written for readers not yet initiated into the basics of emotion theory. I recommend those better versed in the field to browse quickly through it, or simply skip it and go straight to section 1.3 which addresses some of the problems inherent in the cognitive theories of the emotions (concerns about methodology, the components of emotion, etc.). Section 1.4 then aims to clarify what it means to be responsible for one’s emotions, and to classify them according to rationality and moral appropriateness, and as either

‘positive’ or ‘negative’, as well as to offer some initial suggestions about the justification of ‘negative’ emotions.

No scholarly treatment of the emotions must stray too far afield from everyday experiences. That is a crucial mission statement for any

‘Aristotelian naturalist at heart’, as I described myself in the preface. In order to understand and evaluate the emotions, we must know what people are like: what they think, say, and do in everyday encounters. In his ethical writings, Aristotle famously synthesised an account of moral virtues and emotions by considering virtue expressed in fine emotion as well as in fine action, and treating emotions as morally evaluable aspects of character. This is why the recent resurgence of Aristotelianism and the current fad for so-called virtue ethics may seem to bode well for the reinstitution of the emotions into moral and educational discourse. Indeed, some writers think that the moral significance of the emotions can only be adequately captured in terms of such a virtue-based conception of morality.12 I agree that virtue ethics has done a lot of good for emotion research, if only by reintroducing a value-laden focus on the emotions. In the end, however, I shall be tempted to reject virtue-based ethics as a general touchstone by which to judge the moral soundness of our actions and emotions, opting instead for a sophisticated form of utilitarianism.

In spite of my (rather subtle) differences with virtue ethicists, I agree with their presupposition that we need a touchstone before we can start to take measurements. As Martha Nussbaum puts it, one cannot ‘say what role emotions should play in morality […] without defending an overall normative view’.13 This is why I have written chapter 2 about the credentials of a general moral theory: a chapter that I hope the reader will not view as a distracting detour from the main line of discussion, but rather as a crucial backdrop for the moral assessment of particular emotions.

Perhaps the only serious strategic weakness in Ben-Ze’ev’s monumental work mentioned above is that he does not precede his discussions of the moral standing of different emotions by an argued defence for a substantive moral theory. Thus, the springboard from which his evaluations are launched is unclear. While it may be true that the theoretical differences between general moral standpoints are often over-emphasised and their similarities under-appreciated, there are surely more moral theories to choose from than there are snakes in Iceland. I should not be understood as claiming here that the reader will not glean enough from Ben-Ze’ev’s 5

M A P P I N G O U T T H E F I E L D

nuanced account of individual emotions, or from my defence of pride and jealousy, to be able to form his own coherent stance on the morality of emotions without a sophisticated understanding of moral theory; I am not raising a red flag for non-experts based on the intricacies of moral philosophy. Most readers, however, should be aware of enough cases of persons committing evil deeds and thinking evil thoughts because ‘they feel so right’, to sympathise with an attempt to give a moral view of the emotions a firm theoretical grounding.

More specifically, I discuss in chapter 2 the advantages of Aristotelian essentialism and a general naturalist approach to morality and the emotions, an exploration which leads me through the shortcomings of virtue ethics to the untapped sources of utilitarianism. In particular, I consider at length the way in which liberalism fails to guide satisfactorily our emotional life, and the way in which virtue ethics also fails to do so in times of emotional conflict, owing to its insensitivity to the ubiquity of tragic moments in human life.

In chapter 3 I then ponder the nature and conditions of moral and emotional excellence. As the reader will already have seen, my line of argument follows a deductive pattern: I move from general questions about the nature of emotions, via considerations of what constitutes a moral justification of emotions, to a discussion of the virtuous life. Only after having established these general points do I venture to descend to the particularities of the specific emotions under scrutiny. That transition takes place in this chapter. I begin by discussing personhood, integrity, and self-respect, and then, subsequently, shift the attention to an historically important character ideal: Aristotle’s megalopsychia, his crown of the virtues. From the explication of that ideal then flows naturally a characterisation of pridefulness as concern with, and heightened sensitivity to, (simple) pride, shame, and external recognition.

‘Is such sensitivity to be psychologically recommended or abhorred?’ is the basic question in chapter 4, which develops and argues for the value of pridefulness, rejecting one by one the most common objections against it.

The typical criticisms given, that pridefulness is a vestige from primitive

‘shame-societies’, that it makes a person dependent on moral luck, fosters respect for the extraordinary rather than the ordinary, and fails to pay heed to the virtue of humility, do not, on close inspection, detract from the merits of this emotion.

In chapter 5 I then attempt, similarly, to give jealousy its due. That can, however, only be done by relocating the emotion within its conceptual framework and, especially, by rethinking the relationship between jealousy and envy. Indeed, contrary to the almost general consensus in other, previous, accounts, I argue that jealousy is best seen as a certain type of envy. I acknowledge the importance of sexual jealousy, but also show why more lessons can be learnt about the moral standing of this emotion by 6

M A P P I N G O U T T H E F I E L D

looking at it in non-sexual, but no less typical, contexts. When defending jealousy, I pay special attention to the way in which it needs to be contrasted with meanness and spite.

In chapter 6 I provide guidance as to how the emotions, in general, and pridefulness and jealousy, in particular, should be cultivated and schooled: What is the significance of early emotion education? Why do many reasonable people entertain lingering doubts about the feasibility of moral and emotional schooling? How can the emotions really be trained in practice? Is it possible to teach pridefulness and jealousy without inculcating arrogance and meanness? Of particular concern in this chapter is my argument that education for pride and jealousy need not be elitist: a pursuit which enlists the help of an unlikely ally, Nietzsche.

Finally, in chapter 7 I summarise the threads of the foregoing discussion.

I also explain there why my defence of pride and jealousy does not necessarily apply to other commonly stigmatised emotions, some of which may be negative beyond redemption. Indeed, concerning the emotions of pride and jealousy themselves, it should be made clear at the outset, to avoid any premature misconstrual, that my objective is not to defend these two emotions in all cases. To be sure, there are instances of pride and jealousy which have no redeeming features whatsoever. My defence is that of proper pride (qua pridefulness) and proper jealousy, the fundamental idea being that both emotions can be experienced in the right circumstances towards the right objects and at the right time. That these emotions can also be improperly experienced does not, as such, tell against the plausibility of their moral justification, for, as the Latin has it, abusus non tollit usum: abuse is no argument against proper use.

This book is intended neither as a textbook nor as a literature survey.

However, my discussion contains much polemical matter, and I ruthlessly ignore the message of Nietzsche’s diatribe against those bad readers who,

‘proceeding like plundering soldiers’ pick up the few things that they can use from their interlocutors and disregard the rest.14 While I do not pretend to have recorded every point in which I agree or disagree with previous accounts of the topics under discussion, I repeatedly delve into the existing literature to seek support, or polemical targets, for my own views. There are two simple reasons for this method, neither of which is, I hope, merely a camouflage for the old truth that criticising others’ arguments is easier than formulating one’s own. First, given the relatively wide audience for which this book is aimed, it is necessary to provide a feel of the existing literature. Second, I am a firm believer in J. S. Mill’s view that one’s opinions are never sharpened to a finer edge than when they collide with those of others.15 Thus, coming to grips with what previous authors have said becomes a helpful stage in the process of formulating one’s own coherent stance and of giving the reader a taste of the challenges it faces.

M A P P I N G O U T T H E F I E L D

1.2 Cognitive theories and their precursors

Over the centuries, a fair diversity of opinion has been generated on the nature of emotions. What seems at first sight to be a relatively straightforward definitional problem turns out to encompass various distinguishable, if interrelated, questions. Three of the most significant ones are: (i) To which conditions or mental states should we refer as emotions? (ii) What characterises emotions and sets them apart from other states of experience? (iii) How do we distinguish between different emotions in ourselves and in others? Caution is even required in formulating these theoretical starting-points in order not to load the questions against any of the many emotion theories. Notice, for example, that I did not ask which ‘psychological conditions’ count as emotions, for many will argue that there is nothing essentially psychological about them at all. To avoid offending anyone at the outset, one is also well advised to refrain from references to ‘behaviour patterns’, ‘cognitions’, or ‘physiological conditions’. ‘State of experience’

may be about the only term sufficiently neutral to forestall accusations of question-begging. Let us, at any rate, understand the term provisionally in a neutral-enough way.

An exhaustive list of states passing as emotions cannot be given in reply to the first question. If we compare lists of ‘standard’, ‘indubitable’, or

‘paradigmatic’ emotions compiled by different writers, we find at their inter-face the likes of embarrassment, shame, sympathy, compassion, pity, grief, pride, indignation, anger, fear, envy, jealousy, joy, sadness, and remorse. On some lists, pleasure-in-others’ misfortune, disgust, avarice, and sexual desire appear (to give a few examples), while on others they are missing. It is often unclear whether such omissions are intended or incidental. Additionally, there is the problem of individuation which concerns a range of conditions such as romantic love, friendship, laziness, and considerateness, which some writers treat as individual emotions but others as concatenations of various emotional responses. For instance, there might be a case for arguing that love is not a single emotion but rather a common denominator for various emotions and dispositions. A person in love is aggrieved when the object of his love is hurt, jealous when the beloved one starts showing a third party undue attention, joyful when his love is requited, etc. Those points granted, the poet Swift would have been making an important theoretical observation when he wrote:

Love why do we one passion call

When ‘tis a compound of them all?

To complicate matters further, there may exist specific individual emotions which nevertheless only appear as compounds of other more ‘basic’

emotions while not being reducible to them (see s. 1.3). Indeed, I argue later that this is the case with jealousy.

M A P P I N G O U T T H E F I E L D

Finally, not far removed from our emotions lurk our moods. R. C.

Solomon aptly, if somewhat cheekily, describes moods as ‘metaphysical generalizations of the emotions’:16 If I am (non-pathologically) depressed or simply ‘in a bad mood’, I feel sad, but my sadness may not be about anything in particular. This lack of object or intentionality distinguishes moods from emotions in principle, but the dividing line may not always be as clear in practice.17 Exactly what should be included in and what excluded from the list of particular emotions, and why, must remain a matter for further debate, mostly outside the purview of the present study. However, I do have a little more to say about taxonomy and the conceptual methodology of emotion studies in section 1.3.

When talking about the various emotions, it is helpful to consider in each case whether we are referring to them as episodic or as dispositional states.

The episodic ones are occurrent, that is, states with determinate durations:

‘John is jealous of Peter now, because Peter’s paper was accepted for publication while John’s was rejected’. John may also be said to be jealous of Peter tomorrow and the next day, for this or some other reason, with the use of the term ‘jealousy’ still being episodic. We could, however, also be tempted to say ‘John is by nature a jealous person’, thus shifting the focus of the term from an occurrence to a disposition: John has a strong tendency to experience jealousy in various circumstances. Such dispositional uses seem to be conceptually parasitic on the episodic ones, but dispositional emotions are nonetheless interesting in themselves and may involve references to more than one episodic emotion (for instance, pridefulness qua disposition to both simple pride and shame; see s. 3.3). Dispositional emotions should not only be distinguished from episodic ones but also from both background emotions and moods. Background emotions are persisting emotions, but often unnoticed unless certain conditions bring them into consciousness.18 A person could be mourning the loss of a spouse here and now (sadness as an episodic emotion); this sadness might continue to mark and pervade his character for the rest of his life without being consciously noticed all the time (background sadness); the person might also have a strong tendency to experience sadness over misfortunes small or large (sadness as a dispositional emotion), and finally, he could simply be ‘objectlessly’ sad or depressed as explained earlier (mood rather than emotion).

So much for the complexities of trying to respond to question (i) above.

As to questions (ii) and (iii), the answers to those are usually found combined in full-fledged theories of emotion. Holding the field in the early and middle part of this century were, on the one hand, sensory theories of emotion and, on the other hand, behaviour theories. Both are difficult to summarise in a few paragraphs, as are the cognitive theories, which are discussed later in this section.19 The reader should be forewarned that such summaries cannot avoid trading in oversimplifications (summaries of summaries) and references to ‘straw men’. Given the variety of accounts 9

M A P P I N G O U T T H E F I E L D

subsumed under each of these theories, what one risks describing in the end is some sort of a ‘weighed-average view’, not attributable to any particular thinker. Behaviourism is likely to be the hardest hit in this regard since there are almost as many behaviourisms as there are behaviourists. But as the uninitiated reader needs some guiding lights, the following will have to do.

According to the sensory theories, what characterises a state of experience as an emotion and individuates it from others is its ‘feel’: the presence in consciousness of a felt quality, wholly (and only) accessible to introspection.

You are jealous when you experience a certain unique feeling that only you know, a feeling that distinguishes itself from other feelings with which you also have first-hand acquaintanceship, such as those of shame or anger.

Sensory theories differ among themselves as to the origin or cause of the relevant feelings, whether they are primarily psychological or physiological.

Is shame an ‘inner feel’ of an essentially mental nature or is it simply the perception of physiological processes: of blood running through your veins making your cheeks red, etc? The philosopher David Hume is often quoted as a representative of the former variety, and the psychologist William James (along with the physiologist Carl Georg Lange) as the author of the second.

A commonly cited guide to Hume’s view of emotions is Book II of his Treatise of Human Nature.20 However, many contemporary Hume scholars consider the views expressed therein uncharacteristic blunders, since what he says elsewhere about the emotions is both more sophisticated and more in line with modern cognitive theories. William James’s view is spelled out in his Principles of Psychology,21 and other writings.22 James’s adherence to sensory theories is less controversial than Hume’s. In a famous article written in 1884, he asserts that ‘the bodily changes follow directly the PERCEPTION of the exciting fact, and that our feeling of the same changes as they occur IS the emotion’. Not only that; it is possible that ‘no shade of emotion, however slight, should be without a bodily reverberation as unique, when taken in its totality, as is the mental mood itself ’.23 As opposed to the sensory theories, which equate emotions with feelings, behaviour theories claim that emotions have nothing whatsoever to do with inner feels, but rather with behaviour patterns. You are, for instance, jealous when you behave in a certain way: when you respond to external stimuli in a fashion characteristic of that emotion.

It must be said that both these theories of emotion count as somewhat passé nowadays after numerous writers have criticised them steadily for decades. Let us rehearse cursorily some common objections. As to sensory theories, the first point is that the method of introspecting lived experiences, suggested there as a reliable guide to our emotional life, has failed to prove its worth in scientific experiments – so much so, in fact, that it now counts as the bête noire of most psychologists and philosophers. Its results turned out, on most accounts, to be hopelessly subjective and incommensurable. One person might describe his experience of remorse as a pain in his stomach, 10

M A P P I N G O U T T H E F I E L D

another as a load on his back, the third as feeling conscience-stricken, etc.

The same person could even describe the feel of the same emotion differently at different times. The ‘hard data of consciousness’ proved not to be so hard after all. Not only was there something contingently murky and indecisive about the results of such introspection, but, after Wittgenstein, philosophers began to suspect that, qua scientific method, introspection was inherently misguided as an attempt to describe allegedly ‘private’ objects of consciousness, lacking public criteria for identity and individuation. If one person forms a theatre group, who can then be his prompter? These radical Wittgensteinian doubts go hand in hand with a view of language as a tool of communication rather than as a camera depicting inner reality, and with a rejection of the possibility of forming a ‘private language’ by means of which we could systematically and coherently describe the ‘beetles in our own little boxes’: the inner experiences to which no one else can, in principle, have access.24

A second, but related, point is this: irrespective of methodological doubts, there simply does not seem to be any necessary connection between particular emotions and particular sensory states. Neither introspection nor other more ‘sophisticated’ methods have revealed any sensory experiences that uniquely pick out an emotion such as fear, anger, or joy, and this is evidently more than the fault of the methods. At one time, a dry mouth may be an emotional sign; at another it may be caused by simple thirst. Well-documented physiological research points in the same direction: in tests where bodily reactions were mimicked by the use of drugs, subjects turned out to be unable to distinguish phenomenologically between the three apparently diverse emotions mentioned above, that is, until they were tipped off as to what the reason for (as distinct from the cause of ) the emotion might be.

Then those cued to react angrily reported themselves to be experiencing anger, and so forth.25

Third, the sensory theorists’ under-appreciation of reasons for emotions may betray an inadequate grasp of logical connections. For a devout Humean, it is simply a matter of contingent fact, and quite an inexplicable one at that, that we only experience remorse with respect to actions which we have already performed, would have been able to refrain from performing, and now consider morally defective.26 Why could it not so happen at some point in time that we feel remorse with regard to a future act, or one which we consider morally praiseworthy? The sensory theorist has no means within his repertoire to exclude such a possibility. As will become apparent in the sequel, however, the right thing to say is that such events are beyond the bounds of logic. The relevant emotion is determined by its objects by necessity: a way that has nothing to do with its characteristic or uncharacteristic ‘feel’.

Combining this insight with the earlier objections leads to yet another observation: a person can be mistaken about his emotions in a sense that is 11

M A P P I N G O U T T H E F I E L D

ruled out in case of mere feelings. Compare ‘I wasn’t ready to accept it at the time but only realised it later that when my grandma died, leaving me all her money, deep down I was overjoyed although I thought I was sad’ with ‘I thought I had a toothache last night, but now I realise I didn’t’. While the first statement works, there is something mysterious about the second statement, unless of course it simply means that the pain I experienced last night turned out to originate somewhere else than in a tooth: in my gum or tongue perhaps. One cannot be mistaken about a mere feeling, like the existence of pain, as one can be about the occurrence of an emotion. The statement ‘I felt as if I were in pain’, written by an early twentieth century neo-romantic poet in Iceland, was thus rightly held up for ridicule.

Furthermore, emotions can disappear with newly-gained knowledge, whereas mere feelings cannot. Once I realise that what I did to you in the past was morally sound, and not shameful as I had previously imagined, I typically stop being ashamed. By contrast, a person waking up after an operation, complaining of a pain in his leg, does not stop feeling pain as soon as he realises that the leg has been amputated. Admittedly, a person suffering from pain may experience the pain more or less intensely after gaining knowledge of its causes. One of the reasons why many doctors are notoriously reluctant to prescribe pain medication for infants may be their belief that suffering caused by physiological pain does not lie so much in the sensory perception of that pain as in its interpretation as a sign of an underlying disorder. Since infants do not interpret pain in this way, their pain-experiences are not thought to be as serious as those of adults.27 If I am told that the pain in my chest stems from indigestion rather than the onset of a heart attack, I may worry less about the pain and gradually notice it less. However, if it is a real physiological pain, it does not disappear immediately upon my hearing the good news, and, even if it did, the reason would be causal rather than logical. All in all, one can safely say that although emotions may typically and even essentially involve feelings, emotions are not feelings – a conclusion which has made sensory theories fall into desuetude.28

As a general philosophy of mind, behaviourism has long been considered suspect by the majority of the philosophical community. Since criticism has revealed the weakness of the foundations of behaviourism in general, not much needs to be said here specifically about the behaviour theory of emotion.29 The two-tiered objection commonly levelled against this theory is a rather straightforward one: first, an emotion can exist unaccompanied by any specific behaviour pattern (however common their coexistence may be), and, second, genuine-looking emotion behaviour may be displayed without the relevant emotion. An angry person will typically shout, clench his fists and grind his teeth, but it seems naive to hold that one cannot possibly be angry without showing any of these external signs. The emotion might be submerged, or the person strong-willed enough to hold it in and avoid 12

M A P P I N G O U T T H E F I E L D

expressing it in the open. The behaviourist might answer that the emotion of anger need not be identified with angry behaviour but simply with a disposition (repressed or not) to anger-behave. However, the recourse to dispositions jeopardises the explanatory value of the theory. If it cannot be shown that the disposition to anger-behave is, if nothing else, displayed in inconspicuous but measurable signs such as muscle trembling or twitches, the talk of dispositions sounds rather vacuous. However, the invariable existence of such signs has not been supported by any scientific data. Memory reports – for what they are worth – of people who were wholly paralysed by curare and without any muscular activity whatsoever even suggest that they nevertheless experienced emotions.30 As an explanation, then, the recourse to dispositions has become redolent of the infamous ‘explanation’ expressed in one of Molière’s plays that opium makes you sleep because of its soporific power.

The second tier of the objection concerns feigned behaviour. If an emotion such as anger is literally the same as a certain behaviour pattern, what are we to say about the person who pretends to anger-behave: clenching fists, grinding teeth, etc., but does not really experience the emotion? (That feigning an emotion can lead to or bring about its sincere experience is an educationally important matter, see s. 6.3, but irrelevant here.) It seems impossible to distinguish between genuineness and pretence in this respect without reference to mental states, which would undermine the behaviourist point; unless, that is, one wants to claim that feigned behaviour will always lack some detectable signs of genuineness, thus betraying itself in the end. Then, however, the discussion might take an unexpected turn in the direction of aesthetics: How good can acting be? In the end, it seems, one can always cite a Laurence Olivier as a final proof of the imperfections of behaviourist emotion theory.

The foregoing criticisms notwithstanding, there are grains of truth in both sensory and behaviourist theories of emotion: emotions are typically accompanied by feelings, and they commonly beget certain kinds of behaviour. However, in a viable general theory, both the behaviour patterns and the feelings must be kept in their proper places as handmaids and not allowed to give themselves the air of mistresses. Emotions are not the same as behaviour or dispositions to behave, and it is almost a truism in recent philosophical literature that they are not feelings either. Whereas feelings have no reference beyond themselves, emotions have a ‘direction’: an object, a focus. Saying ‘I’m angry’ is elliptical for ‘I’m angry about x’, or ‘angry that x’, where x signifies the reason for my anger, for example: ‘I’m angry that you showed up drunk for my birthday party’. Notice that the reason need not be the cause of my anger. Maybe I became angry because I had spent a sleepless night preparing for the party. Furthermore, your showing up drunk might have caused me to be angry about something else, for instance the fact that my sister got married to a jerk like you.31

M A P P I N G O U T T H E F I E L D

Feelings, such as physical discomfort after a sleepless night or euphoria after a delicious meal, can sometimes create their own objects, randomly seeking out a place for themselves like the will-o-the-wisp on a ship’s mast, when it seems as if it was the mast which created the glow. This does not change the fact that without a ‘mast’ the ‘rays’ can never become an emotion, nor does it detract from the reality of the experienced emotion. I was angry with you in that birthday party, not only sleepless. Saying to you

‘Sorry, I wasn’t really angry, just sleepy’ would be an inappropriate apology. However, I could apologise later on by saying that I realise now that my anger was irrational (because you were not so drunk after all) or morally unfitting (because I had told everyone beforehand that this would be a rave), and that I had only experienced and exhibited anger because of sleeplessness.

Observations such as these characterise a major advance in the academic analysis of emotions made by so-called cognitive theories. According to these theories, emotions are intentional states and have propositional objects in the sense that what the emotion is ‘about’, ‘of ’, ‘for’, ‘at’, or ‘to’

can in principle be specified propositionally. The relevant emotion is then given by its propositional content, and such content is characterised by what the agent takes his relations to be vis-à-vis the object(s) of the emotion.

Only if certain appreciations of these relations are in place can we, logically, be said to have a particular emotion. Thus, to feel fear, for example, you must consider yourself in the presence, or about to be in the presence, of that which can harm you; to feel remorse you must deem a past action that was under your voluntary control morally defective, etc. Hence, the primary step in identifying and differentiating specific emotions becomes that of disclosing their logical conditions: the conditions which must be satisfied for the relevant emotion to be logically possible.

The rapidly growing literature on cognitive theories is not fully homogeneous. In most accounts, emotions are said to presuppose both factual beliefs and evaluations. Some go so far as to say that emotions not only involve evaluations, but are evaluations. (I shall say more about this tangled topic in the following section.) Thus, a special strand of ‘evaluative theories’ might perhaps be distinguished when mapping out the cognitive-theory landscape. Incidentally, that strand is no modern invention but harks back all the way to the Stoics who conceived of emotions as recently formed false judgements about the goodness or badness of states of affairs. In Stoic theory, all such judgements are necessarily false because each event in the world is to be considered morally neutral, taking place within an inex-orably deterministic world-system. Feeling guilty is judging that you have wronged someone, but Stoic therapy makes you realise that no one ever wrongs anybody, because no actions are truly voluntary. Eventually, the wise person will rid himself of all emotions and reach the perfect state of Stoic equanimity.

M A P P I N G O U T T H E F I E L D

There are striking similarities between the Stoic conception of emotion and that of Robert C. Solomon, a leading modern proponent of cognitive theories. To be sure, Solomon is not a hard determinist and does not think that all emotions are inherently wrong-headed. He claims, however, that emotions are normative judgements:

‘I am angry at John for taking […] my car’ entails that I believe that John has somehow wronged me […] The (moral) judgment entailed by my anger is not a judgment about my anger […] My anger is that judgment.32

The problem about this formulation is that by its concentration on the cognitive element of emotion, it seems to overlook another equally important element: the conative one.33 Being angry with John for taking my car is more than believing (and making a normative judgement) that in doing so John has wronged me. I might believe that and still not be angry, for instance, if I did not care a whit whether my car was stolen or not. What is required here is some kind of concern about the offence. As Ben-Ze’ev puts it: ‘Emotions arise only when we care’.34 Other cognitive theorists usually recognise this by saying that an emotion is a combination of a belief and a concern (desire or aversion): here, the belief that John has stolen my car accompanied by the desire that my car should not be stolen. In addition to belief (or belief plus evaluation, when these are kept separate) and concern, some cognitive theorists want to add affect as yet another necessary component of emotion: no one can thus be (episodically) jealous unless some kind of ‘botherment’ has attached itself to other aspects of the emotion.35 Another alternative would be to see the conative element as explaining the ‘affective side’ of emotions without the need for a new independent component. If one’s desire is frustrated, must one not necessarily feel bothered? The ‘botherment’ may, however, be expressed in a variety of ways: by feelings that vary between individuals, feelings that vary within the same individual at different times, or even ‘affects’ that are unfelt because they are repressed in our subconscious. If this suggestion, to which I shall return in the following section, holds water, sensory theories would be right in that ‘affects’ are necessarily involved in emotions, although these ‘affects’

need not be conscious, nor – as mentioned earlier – follow the same pattern whenever a particular emotion is evoked.

Another general problem commonly noted with cognitive theories lies in the link which they require between emotions and beliefs. We are told that fear always goes hand in hand with acknowledged beliefs about danger, resentment with acknowledged beliefs about unfair treatment, and so on. Although this may be true in most of our emotional experiences, it is not always so. Sometimes emotions are in conflict with avowed beliefs.

While it is true that emotions can, and often do, disappear with newly 15

M A P P I N G O U T T H E F I E L D

gained knowledge, a change in belief does not always guarantee a change in emotion. Although I realise that what I did to you in the past was morally sound and not shameful, as I had previously imagined, I do not necessarily stop being ashamed. Admittedly, in such cases the relevant emotion may be irrational, but why should a mental state be any less real for that reason?

Cheshire Calhoun has tried to work out a sophisticated version of cognitive theory, which is sensitive to the fact that though some cognitive element must be built into emotion, it need not be a belief or a judgement.36

Calhoun uses the following examples:

Tess has a spider phobia. Spiders make her skin crawl and she jerks out of their way, pleading for someone else to kill them. Yet Tess believes spiders are harmless and knows enough spider biology to back up this belief.

Raised in a conservative household, Tess acquired, among other beliefs, the belief that homosexuality is unnatural and immoral.

But in college, both friends and professors challenged this. After extensive discussion and reflection, Tess came to believe that homosexuality is neither unnatural nor immoral. But several years later, she suddenly discovered that a good friend is a lesbian and she experienced feelings of shock and revulsion.37

In Calhoun’s view, dissonance can occur between our belief system, comprising a set of reflectively held, articulable judgements, and our more general cognitive system which also includes pre-reflectively held claims and an unarticulated framework for interpreting the world – a system that may, for instance, be partly the product of childhood conditioning. To have an emotion is not necessarily to believe x, but rather to see the world as x, whether or not this interpretive ‘seeing as … ’ ever emerges in our reflective belief system. Thus Tess’s emotions channel her cognitive life down well-worn paths, although these remain hidden and would not be accepted by her as her beliefs.

Robert C. Roberts shares the same orientation as Calhoun regarding possible emotion–belief conflict and comes up with a similar, if even more subtle, analysis.38 As Roberts puts it, what emotions and beliefs have in common is propositional content. If A is angry with B, it is either because A believes B has culpably offended or sees B as having done so. Roberts invokes here the notion of a ‘construal’. To return to Calhoun’s example, Tess may construe spiders as harmful although she believes they are not. A construal in this sense is the grasping of something (here spiders) in some terms (‘harmful creature’). Roberts refers to Wittgenstein’s famous example of the same image being seen either as a duck or as a rabbit by dwelling on or attending to different aspects of it. To change from one to the other requires a gestalt switch, which can at times be difficult to master. Tess’s 16

M A P P I N G O U T T H E F I E L D

construals may have her in their grip, although she would deny the corresponding judgements. In other words, the construals have for her the appearance of truth, although she would not affirm their being true. The long and short of Roberts’s analysis is that because construals form the necessary cognitive condition for an emotion’s occurring, while judgements do not, emotions are better thought of as construals than beliefs.

The Calhoun–Roberts solution is not without problems of its own. What exactly is the epistemological standing of these construals? Emotions are intentional states with propositional contents. If the construals satisfy those criteria, it is difficult to see what distinguishes them in the end from ordinary beliefs, some of which may be vague or even hidden. Again, allow me to postpone further elaboration of a lingering problem in cognitive theories until the more critical exploration in section 1.3.

Instead, let me finish this brief survey with yet another open question, if only to remind the reader that there is still a lot more theoretical work to be done in this field. What distinguishes those beliefs (or if you prefer, construals) that form the basis of our emotions from the rest of our beliefs?

I may have strong convictions about techniques and tactics in angling and car repairs – even strong enough to lead to physiological changes and distur-bances – but I do not classify those beliefs into distinct categories related to the passions of the soul. It does not help much to say, as Ben-Ze’ev does, that our emotion-beliefs are more ‘personal’, expressing ‘our profound values and attitudes’,39 for other types of beliefs (such as those about angling and car repairs) can also be deeply personal and profound. Annette Baier suggests – professedly drawing on Descartes and Freud – that the emotions have what she calls ‘deep objects’, derived from highly significant early experiences, objects which can only be properly understood by telling a biographical, associational story. A person’s view, however important to him, about how best to arrange worms on a fishhook or how to fix brakes in a Volvo can be understood without considering the trajectory of his life.

However, his revulsion when putting a cold earthworm into his mouth may only be understandable as the last link in a long chain going back to a fateful childhood experience such as the drinking of a foul-tasting liquid which looked like his mother’s milk, or to the first realisation that one day he himself will be devoured by worms in his grave. The deep object may emanate from art and literature as well as from real events: from myths and horror stories that have left an indelible mark on the person’s soul.40 Ronald de Sousa suggests, somewhat similarly, that our emotional repertoire has its origin in ‘paradigm scenarios’ mediating between the past and the present.

Every emotion is considered to be rooted in a dramatic situation or episode type associated with a characteristic feel. These paradigm scenarios, drawn from childhood experiences but later reinforced by the arts and culture to which we are exposed, provide both the characteristic object of the specific emotion-type and a set of typical responses.41

M A P P I N G O U T T H E F I E L D

One question, however, immediately arises, namely: What if my disgust and your disgust turn out to have radically different ‘deep objects’, or to be rooted in radically different ‘paradigm scenarios’; in what sense can we then talk about the same kind of disgust, or disgust at all as a separable emotion?

It does not suffice here to insist that my disgust and yours will, nevertheless, share the same basic evaluative pattern although they may be connected, in actual experience, with various other beliefs and desires; for the invocation of the ‘paradigm scenario’ was meant precisely to help us specify what this pattern is. De Sousa seems to give in to relativism by accepting that the

‘canons of normality according to which we must assess the rationality of emotions are ultimately individual’.42 Baier tries to hold the line against relativism by claiming that it so happens that we inhabit a common physical and cultural world, and that this universality provides, at least to some extent, a ‘common emotional world’.43 While that answer is convincing up to a point, at least for those of us who share Aristotle’s belief in inter-human ‘grounding experiences’ (see s. 2.1), we must remember that children experience anger, jealousy, pride, envy, and so forth from a relatively early age. It is rather implausible to suggest that they have all, by that time, gone through similar enough experiences – traumas, disappointments, and joys –

to explain the universality of those emotions, though surely the deep objects or the paradigm scenarios must be somehow isomorphic in order to give rise to the same emotions. Hence, I still consider the question of what distinguishes emotion-beliefs from other types of beliefs – be it a single marker or more diverse and subtle differences – an unanswered one.

Despite the lack of uniformity among cognitive theories and the remaining problems which harass them, they have, arguably, signalled a major advancement in comparison with their predecessors. What should be noted here is that, notwithstanding the differences between them, all cognitive theories paint an intrinsically rational picture of human emotions.

Having cut away the alleged dead wood of sensory and behaviourist theories, what emerges is a view harking back to the ancient Aristotelian conception of emotions as more or less intelligent ways of grasping situations, dominated by a desire. For Aristotle, as for recent cognitive theorists, emotions are seen as penetrable to reason: as rational states that change with our opinions, or as more irrational ones which can still, in principle, be brought to consciousness to be reflected upon and, if necessary, ‘defused’.

Instead of being beyond the bounds of reason, the emotions are placed right in the middle of reason’s kingdom itself. Human nature is no longer seen as that of a divided creature, the inevitable battleground of reason and passion, but as that of an essentially, or at least an ideally, unified being.44

In addition to its similarities to contemporary emotion theory, Aristotle’s general account of the emotions is highly impressive in its own right. Like contemporary theorists, Aristotle avoids treating emotions as irrational, uncontrolled, or essentially inappropriate responses to situations. Our 18

M A P P I N G O U T T H E F I E L D

emotions may, at times, be unwarranted, but just as often they may be warranted. He especially develops this point in the Nicomachean Ethics, in which he argues that virtue (for instance, bravery and generosity) is largely a matter of feeling the right thing (of re action as well as of action); for instance, the brave individual is neither fearless nor overwhelmed by fear in a dangerous situation.45 The emotions thus become central to his treatment of moral virtue and vice, in which virtue is famously defined as a mean of action and passions. In other words, just as the right deeds are to be performed in the right circumstances, Aristotle claims that our emotions should be appropriate to the situation: felt toward the right individual, at the right time, and in the right amount, being neither too violent nor too calm.

He notes that what characterises many emotions is a strong moral belief about how oneself and others should act, and that in a good person there is ideal harmony between emotions, values, and actions. In his Rhetoric, too, much can be read about the structure and moral worth of particular emotions, and how lack of emotion can be as debilitating as its surfeit.46 By seeing emotions as potential virtues or necessary ingredients in virtues, Aristotle’s theory of emotion plays a substantial role in his general moral theory.

The link Aristotle forges between actions and emotions (with both exemplifying potential virtues or vices) may strike the modern reader as somewhat odd at first glance as we are used to considering the latter less

‘active’ and ‘intellectual’ than the former – witness the entrenched distinction in ordinary language between ‘actions’ and ‘passions’.47 However, a closer look will reveal how blurred the division between the two is bound to be. A person, such as Nelson Mandela, languishing in his prison cell, cannot be said to have lost his virtue of benevolence just because he has little if any chance of displaying it through action, that is, not as long as the underlying passion (the relevant compassion) and the related disposition to act well is intact. Moreover, benevolent activities performed in everyday life tend to be inspired by compassion and related emotions. There is a strong case for saying, à la Aristotle, that talk about passions and actions simply directs our attention to different aspects of the same virtues or vices. One might even want to say that an occurrent emotion such as compassion is simply one kind of action (of the mind) which typically motivates another kind of action (of the body). What we should resist in this line of thought, however, is a temptation to which Solomon, for one, succumbs: that of explaining the difference between ‘emotion-acts’ and ‘normal rational deliberate actions’ by categorising the former as ‘rash’, ‘hasty’, and ‘dogmatic’.48 While it may be true that emotions are often brief and unstable, it does not follow that they have been hastily and dogmatically arrived at. For instance, it is unclear why an act of anger (qua judgement or qua bodily action motivated by anger) must necessarily count as more hasty and dogmatic than an action stemming from what Solomon would count as ‘rational deliberation’. After all, 19

M A P P I N G O U T T H E F I E L D

people commonly claim to have arrived, as a result of ‘rational deliberation’, at conclusions such as that it is wise to take up strenuous exercise at the age of 65, or to stop taking medicine prescribed by their doctor.

Cognitive theories have more to say in detail about the psychology of emotion than Aristotle; he, on the other hand, is stronger on the moral aspects of emotions. Combining these different insights helps to shed light on various problems which other emotion theories leave in the dark. I shall be making substantive use of these insights in what follows.

1.3 Taking stock: some critical comments on cognitivism A well-known psychologist wrote in 1983 that there were signs that the

‘cognitive wave’ was on the wane.49 However, news of its imminent death seems to have been rather premature at that time since cognitive theories still dominate emotion research within both philosophy and psychology. To the average reader, the vast majority of ‘different’ models and paradigms posited by psychologists conducting empirical research into the emotions may seem little more than variations on a common theme – the theme that cognition lies at the centre of emotion – with the other serious candidates of yore, such as pure sensory and behaviour theories, having been cast aside, much like an old hat.

The review in the previous section brought out many of the strengths of the cognitive theories. At the same time, it also indicated how they are still marred by a number of unresolved difficulties and internal conflicts. One is tempted to observe, ironically, that the virtues of cognitive theories would be overwhelming if they only had fewer vices to keep them company. The aim of the present section is to ameliorate, or at least to cast new light on, some of these ‘vices’, by bringing to bear on them insights from other areas of philosophical inquiry, particularly insights relating to the methodology of conceptual studies and the nature of moral language. Let me first say something about taxonomy, and then turn to the notion of basicness and the necessary components of emotion.

We have already seen how the lists compiled by different philosophers and psychologists of standard emotions only coincide to a limited extent.

Everyone seems to have his own shopping trolley packed with different goods. Psychologists have located hundreds of emotion terms in the normal adult’s vocabulary; including the so-called ‘aesthetic emotions’, and emotions describing alleged other-worldly experiences in religions East and West, adds even further to the disarray.50 The complexities of the emotional landscape may tempt us to conclude, as Amélie Rorty does, that ‘emotions do not form a natural class’.51 Perhaps the ‘passions of the soul’ can only be classified according to family resemblances rather than essential defining criteria; the vast number of ill-defined emotion-terms, labels, and phrases referring to our various ‘states of experience’ points in that direction.

M A P P I N G O U T T H E F I E L D

However, if such a conclusion is taken to mean that the classification of emotions is dependent simply on the vagaries of language in different societies and at different times, we are left with a disconcerting implication. For, arguably, the compassion felt for victims of misfortune in modern society is essentially the same kind of compassion as the soldiers in ancient Greece felt for the desolate Philoctetes, left to his own devices after being bitten by the serpent of Lemnos on that deserted island.52 The same kind of compassion obviously does not mean that it must be compassion felt for the same reason, but simply that we moderns are able to identify our feelings with those of the soldiers, given the circumstances in which they found themselves, and to see Philoctetes’ bad fate through their eyes.53 The universality of emotions thus sets certain limits to a relativistic interpretation and makes abandoning the classification of the emotions according to some natural schema less enticing than it might otherwise seem.

Admittedly, this universality of emotions is not undisputed, and reference to the alleged inter-human understanding of emotions across cultural and temporal borders is seen by some as question-begging.54 However, when anthropological evidence is presented for a non-universalist view – of unknown emotions appearing and others more familiar being absent in some ‘distant’ societies – its interpretation is typically marred by one or both of two common errors. It is supposed, first, that if there is no word ‘ x’ in the given remote language, y, which can be directly translated as the name of the emotion to which we refer as ‘ x’, then that emotion cannot exist among the speakers of y; second, it is taken for granted that a person cannot experience an emotion unless he attributes that emotion to himself. There is, however, no direct correlation between words and concepts. In contemporary Icelandic there is no word corresponding to the English term ‘compassion’

as distinct from ‘sympathy’; they would both typically be referred to by the same common name, ‘ samúð’.55 Yet this does not mean that Icelanders cannot experience two distinguishable emotions, or that there are not other manoeuvres within the language to differentiate between the two, for example by calling compassion ‘ djúp samúð’ (‘deep sympathy’). Moreover, as I shall discuss later in this section, people are commonly self-deceived about their emotions; self-attribution is not a necessary condition of the experience of an emotion.

To return to Rorty’s conclusion above, much depends on what she means by a natural class. If she is referring to something like the periodic table of elements, she is probably right: emotions do not come with different atomic numbers. However, if we consider the everyday classification of natural phenomena into, say, rivers and brooks, mountains and hills, what she says becomes less plausible. There, our concepts are more ‘open-textured’ than those relating to natural kinds in the strict sense (such as the chemical element gold), not to mention the non-natural ‘closed’ concepts of mathematics and logic. In the case of geometrical concepts we can, for example, 21

M A P P I N G O U T T H E F I E L D

give a strict rule prescribing the necessary and sufficient properties of a rect-angle, that is, we can fashion a complete definition which anticipates and settles once and for all every possible question of usage. In the case of most concepts of the natural, moral, and social sciences, however, this is not the case: the boundaries of the sets of things to which our concepts refer there do not typically constitute tight-shut compartments and are thus difficult to fix precisely.56 However, this does not mean that our classification of those natural classes which do not constitute natural kinds in the strict sense, nor for that matter of moral and social concepts, cannot be more or less reasonable, more or less well argued-for, within the limits of the field. As Aristotle would be the first to remind us, we must in each field of inquiry plough with such oxen as we have.57 The borderline flexibility of most of our everyday natural concepts does not imply essential contestability.58 If what Rorty means is that emotions cannot be considered a natural class only because we cannot ascertain with the same precision whether a given state of experience constitutes emotion as whether a given sample from a mine constitutes gold, her conclusion is a non sequitur.59

However, rather than dismissing Rorty’s remark out of hand, we should set it against the background of methodological considerations (or rather the lack thereof) in cognitive theories. A philosopher entering the field of emotion research from other areas will be somewhat taken aback by the paucity of ‘meta-talk’, something akin to meta-ethics or the methodological explorations conducted within philosophy of science, to be found there.

Psychologists and (even) philosophers probing the various emotions usually get down to work without worrying too much – or so it seems – about the adequacy of the tools that they are employing. In spite of the demise of the ‘ordinary-language conceptual analyses’ of the 1950s and ’60s, moral philosophers still spend considerable time ruminating over the nature of moral language and how moral truths are to be arrived at. They also have on hand a wealth of material on those issues from the time when meta-ethics ruled the day. So what should cognitive theorists learn from moral philosophers in this respect? What they could learn, I submit, from philosophers working within a broadly Aristotelian perspective, is that all satisfactory inquiries into the nature of open-textured concepts must be critical inquiries.

In our world, concepts are formed for a purpose and have a point. The aim of conceptual studies is to argue critically what these points are and/or what they should be, in the light of our existing knowledge of human beings and their environment. In other words, it must be shown why the point of the given concept is or should be of interest to people, given their common human nature, and how the term designating the concept must be defined so as to correspond to this point.60

Suppose we want to define the term ‘table’. What we do is to search for a function or a social need that the term is meant to fulfil. If someone insists on reserving the term ‘table’ for four-legged objects fulfilling this need, we 22

M A P P I N G O U T T H E F I E L D

point out the arbitrariness of the distinction: why not also three-legged ones? For what reason, that is, should the language of ‘table’ be thought to mark out this particular distinction? If, on the other hand, the definition is so broad as to include tables and beds, we would say: ‘If you use “table” in this way, you will have no term with which to refer specifically to objects you put your plates on at dinner, as distinct from those you sleep on at night’, etc. Much the same can be seen to apply to ‘jealousy’ as to ‘table’ in the example above. What we need to do is to produce critical naturalistic arguments about how the term ‘jealousy’ should be defined for its use to be internally coherent and distinguishable from the meanings of other related terms.

As a starting point we should of course, as Aristotle proposed, collocate most of the possible endoxa, the conceptions or appearances of the ‘many or the wise’ about the given concept, work through the puzzles of contradic-tions in usage and beliefs, and then formulate a new account which we bring back to the appearances, trying to retain the truth of the greatest number and the most authoritative of these.61 At the outset there is simply no other rational way to proceed. However, in the end we may be doing much more than ‘respectfully tidying up ordinary usage’, as Roberts describes his enterprise of defining ‘emotion’.62 It often happens that ‘common usage fails to honour distinctions which themselves emerge only from an analysis of common usage’,63 or we realise that distinctions entrenched in ordinary language are in fact redundant. This is in no way mysterious: definitions are created to serve a purpose; they do not fall into our lap by chance. Let me make it quite clear, however, that I am not proposing a quest for an arbitrary, stipulative definition of jealousy. What we need to show, for jealousy or any other emotion, is its point or purpose in human relations, and then to argue critically for a particular definition of the emotion that best conveys this point. In the end, what hopefully emerges is a definition that is objective in the sense of being objectively useful to those interested in certain relations between human beings, the reference being fixed by what Nussbaum aptly calls our common universal ‘sphere of experience’.64

Given the typically sparse or non-existent remarks about methods, Ben-Ze’ev should be praised for paying some attention to methodology at the beginning of his book.65 What he says, however, is not entirely clear. First, he makes a distinction between ‘binary’ and ‘prototypical’ categories, a distinction that seems to correspond substantially to that between ‘closed’

and ‘open-textured’ concepts as introduced above. Emotions, according to Ben-Ze’ev, constitute prototypical categories, inclusion being determined by the degree of similarity to the most typical cases. There is no way to list necessary and sufficient conditions for inclusion, and the borderlines are fuzzy. With this I would more or less agree. But how is the ‘best’ or ‘most typical’ example, which determines the inclusion of other examples, then to be found? Ben-Ze’ev claims the prototypical example is one that exhibits 23

M A P P I N G O U T T H E F I E L D

‘the significant features of the given emotional category and has but a few distinctive features that are not shared by category members’. Reference to

‘the significant features’ may here seem to be question-begging, but Ben-Ze’ev goes on to explain that we determine these features first by asking people to describe typical cases; second by discovering the features through conceptual analysis. The first point is true but rather trivial, given the typical disharmony of people’s examples. The second point is more promising but still begs the all-important question: What kind of method should we use in our conceptual analyses? There are signs that Ben-Ze’ev is suspicious of a naturalistic solution – witness his disparaging reference to a

‘vague discourse about some essence of emotions’.66 But then I would suggest that he is faced with a dilemma: either the ‘prototypical’ example yielded by conceptual analysis is prototypical because it best captures the natural point of the concept, in which case the concept itself and not only the particular example should be explicated, or else it is the best example for some other reason, in which case mere reference to ‘conceptual analysis’

tells us precious little about how the example can be found. This is not meant to undermine the merit of what Ben-Ze’ev or, for that matter, other emotion theorists subsequently say about particular emotions, which is often insightful and to the point. After all, the proof of the pudding generally lies in the eating rather than somewhere in the cooking. However, I believe that it would add substantially to the merits of their dishes, or at least help others to follow suit, if we received more precise information about the cooking process.

Next a few observations about ‘basicness’. A distinction between ‘basic’

or ‘primitive’ emotions, on the one hand, and ‘non-basic’ or ‘complex’

emotions (that is, emotions which are reducible to primitive emotions or to such emotions and some non-emotional factors) on the other, is central to many cognitive theories. As with emotions in general, the problem is that the lists of the basic emotions tend to differ both among philosophers and among psychologists, although one will probably find sadness, anger, and fear on most of them. Some authors, however, simply reject this distinction altogether. Yet others, such as Ben-Ze’ev, try to steer a middle course by retaining the distinction but not letting it play too important a role in their analyses of specific emotions.67 I must admit that I find the invocation of this distinction, as typically formulated, of dubious usefulness. First of all, diverse and often seemingly conflicting criteria have been invoked for considering some emotions as basic, criteria having to do with priority of development in human beings (qua species and qua individuals), functional value, universality, prevalence, and uniqueness (physiological, expressive, etc.).68 Exploring these criteria, the psychologists Power and Dalgleish admit that the arguments typically given for each of them are weak; yet they try to extrapolate from research findings exploiting all of these different criteria and end up with a ‘core list’ of anger, fear, disgust, and sadness.69

M A P P I N G O U T T H E F I E L D

However, extrapolating from weak premises yields a weak conclusion. It seems reasonable to suppose that pursuing, for instance, the developmental criterion may enlighten us about which emotions are prior in the order of emergence; those would then be ‘basic’ in a certain temporal sense, but not necessarily in the above (‘non-complex’) sense of all other emotions being reducible to them. Nevertheless, there are cases when it seems reasonable to understand an emotion as a compound of other emotions. Perhaps we may here again be aided by some philosophical insights from a different field.

Let us consider Elizabeth Anscombe’s clever discussion of the ‘bruteness’

of facts.70 Saying that my grocer carts potatoes to my house and leaves them there is more basic or ‘brute’ than a description of the grocer as having supplied me with a quarter of potatoes. The second description is, again, more brute than saying that I owe the grocer such-and-such a sum of money, but that description could, in turn, be brute relative to yet another description, say, of my being solvent. None of the descriptions is reducible to the preceding one in the order of bruteness; yet each naturally follows from there, given a context of normal procedure not implied by the descriptions themselves but not independently describable either except as the absence of all circumstances which could impair the given description as a description of the relevant action (e.g. carting potatoes to the buyer’s house). In this sense, facts are not brute per se; they are only brute relative to descriptions and contexts.

Now, there exists some interesting psychological literature about

‘blended’ or ‘compound’ emotions, that is, emotions as compounds of other emotions aroused by various and often conflicting aspects of the same object or situation.71 ‘Compound’ must here be understood to refer to a unique combination, not merely to co-existence. If jealousy is cognitively organised as a compound emotion, as I shall argue later (a compound of anger, envy, and righteous indignation), then it does not simply mean that the jealous person is angry, and envious, and righteously indignant, but that these three emotions are blended together in a unique, characteristic way in the cognitive structure of jealousy. Anscombe’s analysis helps us to understand how this can be, without presupposing that anger, envy, and indignation are more basic (non-complex), as such, than jealousy; it may only so happen that they are brute relative to a certain description of a mental state and a certain situation. There could exist other emotions or attitudes relative to which jealousy itself (along with some other emotions or non-emotional elements) were brute – for instance, disgust of Jews. Even more importantly, there could well be situations in which other emotions were brute relative to anger. A proposed hierarchy of basicness in each particular instance thus does not imply any controversial theory about essentially basic/primitive versus complex emotions. What we have, instead, is a much more flexible account of basicness according to which the same emotions can appear under different descriptions at different levels of 25

M A P P I N G O U T T H E F I E L D

basicness, just as the same behaviour (the same set of motions) can appear under different descriptions as an expression of different emotions.

Let us now turn to the components of emotion. Although all cognitive theories naturally enough consider cognition the linchpin of an emotion (hence their label), they part company as to the exact nature and status of this basic component. Most common perhaps is a multi-component variety which divides the cognition element into two ( factual belief and evaluation) and adds some other, less central but still necessary, components: concern, feeling and (sometimes) behaviour patterns.72 However, we saw in the previous section that there exist versions, such as that of Solomon, which make do with a single cognitive component, evaluation, and consider the non-cognitive elements unnecessary. This discordance may be seen as a source of embarrassment for the cognitive theories, or at least give us some cause for concern about their validity. While I shall not be adding any new positive arguments here for the viability of a cognitive approach, I shall try to defuse this ‘source of embarrassment’ by carving out a middle-ground approach between the multi- and single-component versions, an approach which follows Solomon in positing evaluation (that is, evaluative belief ) as the single cognitive component, but departs from him in adding concern (desire or aversion, see s. 1.2) as the other necessary component of emotion.

I argue that the multi- and single-component views are both anchored in certain misconceptions concerning the much discussed fact-value distinction. More specifically, I claim that the multi-component view retains too much of this distinction which should have been omitted, while the single-component view omits too much of it which should have been retained. I think that a cogent line of defence for my claim can be worked out by drawing on two separate sets of arguments, by Elizabeth Anscombe and Philippa Foot, respectively, relating to the nature of moral language.

The distinction between facts and values was long taken for granted in philosophical circles. While it has been waning in popularity there during the last few decades, it still holds its ground pretty well in the social sciences, especially in economics and within some of the prevailing research paradigms in psychology. According to this distinction, only ‘facts’ are objective whereas ‘evaluations’ (moral, aesthetic, or otherwise) are subjective responses to facts. The details of the numerous more-or-less radical subjectivisms based on this distinction are less important here than their common assumption that evaluation always presupposes some purely-descriptive, value-free foundation: i.e. that which is evaluated.

Many proponents of the multi-component view seem to be strongly influenced by this distinction. Ben-Ze’ev’s account is a case in point.73 He divides emotions into four basic components: cognition (beliefs, veridical or distorted, about the facts of the given situation), evaluation (assessment of the cognition), motivation (based on concern), and feeling. Thus, when A envies B for getting better grades, A has some information about B’s grades, 26

M A P P I N G O U T T H E F I E L D

and evaluates his own inferior position negatively, and is concerned about abolishing his inferiority, and entertains some negative feelings about the case. Ben-Ze’ev emphasises that the evaluative component is the most important one, in that it distinguishes between emotions. For example, hope distinguishes itself from expectation in evaluating the relevant future event as somehow favourable. However, he is also at pains to explain that the cognition, which ‘contains descriptive information about the object, logically comes prior to the evaluation of this object, namely, to a normative appraisal of its value. Hence, there can be cognition without evaluation’.74

In other words, to use a recently fashionable jargon, evaluation ‘supervenes’

upon brute facts.

It is salutary at this point to return to Anscombe’s piece, cited earlier in this section. Her point there is not only that the bruteness or basicness of facts is relative to descriptions and contexts, but also that our climb up and down the ladder of bruteness straddles any taken-for-granted distinction between facts and values. Recall Anscombe’s examples, where carting potatoes to my house and leaving them there was considered brute relative to supplying me with potatoes, which again was brute to my owing the grocer money, which then was brute relative to my being solvent, and so forth.

There is no particular step in this ladder of bruteness where we can say that the factual has ended and the evaluative taken over. Notice that this is much more than a slippery-slope argument about a slide from the ‘purely factual’

to the ‘purely evaluative’, for my owing the grocer money (and it thus being morally incumbent on me to pay) seems to be no less of a fact than that he left potatoes at my door.

To add further strength to this insight, it is helpful to bring in here reference to J. Kovesi’s perspicuous analysis of the nature and formation of moral and non-moral concepts. His conclusion is that their difference does not at all coincide with a difference between evaluation and description. The real difference can rather be seen as lying in the formal elements, the diver-gent reasons for collocating certain features, aspects and qualities, and for grouping them together. Thus, whereas we always describe from some point of view, we can never be said to do so from a perspective which would be called the descriptive point of view. In the case of moral concepts, such as murder, we describe from the moral point of view, but that point of view is in no way less descriptive than the perspective employed in classifying an object as a table or a kettle. It is simply another perspective; we are drawing attention to features of another sort.75 Kovesi’s considerations help to bring out the nature of the misunderstandings that flourish on the basis of the description-evaluation distinction. Evaluation is not the icing on the cake of hard facts; in the case of evaluative notions we cannot, so to speak, peel away the layers of evaluation until we touch bottom, until we reach the neutral descriptive content. We do not first have hard facts and then load them with normativity. That is to say, although we have facts such as 27

M A P P I N G O U T T H E F I E L D

another candidate getting a post for which I applied, or my brother receiving more attention from my mother than I, which would, to use Anscombe’s terminology, be ‘brute relative to’ jealousy, there is nothing in the nature of the brute facts themselves that tells us to collocate them under the heading

‘jealousy’. There is an independent evaluative rationale behind the concept of jealousy which determines the reactions brought together under that heading. If we remove the rationale, we are not left with the descriptive criteria for jealousy minus evaluation; we are simply left with nothing at all.

As Kovesi puts it most forcibly: Evaluative notions ‘do not evaluate the world of description but describe the world of evaluation’.76 All notions are formed for some reason, and they ‘describe’ from some point of view.

Dividing ‘cognition’ up into belief and evaluation or, as Ben-Ze’ev does, reserving the term ‘cognition’ for the descriptive element and loading ‘evaluation’ on top of it, obscures the way in which the evaluative component in emotion describes the world around us and picks up, for instance in jealousy, those instances of other people’s acts vis-à-vis us and a third party which fit into its grid. It is not as if we can, à la Ben-Ze’ev, neatly distinguish between my perception of the facts of such a three-party relationship and my subsequent evaluation of them as morally positive or negative; rather, as I suggested in the preceding paragraph, what is perceived of as fact is informed by the very moral point of view. That point enters into the collocating description of the jealousy-inducing relationship (as one of undeserved disfavouring; see s. 5.1) rather than supervening upon it. For in default of such a point, these facts would never have been brought together under a single description in the first place. In other words, Ben-Ze’ev’s division is both redundant and misleading.

Should we then simply embrace a single-component view of emotion where evaluation rules the roost? I have already mentioned Solomon as an advocate of such an ‘evaluation theory’. However, let us concentrate here on Nussbaum’s version, as she has argued for its superiority both more passionately and more recently than Solomon. Nussbaum defines emotions as evaluative judgements ‘about important things’, that is, things to which great significance for the person’s own flourishing can be attached but which are, at the same time, more or less outside his control. Nussbaum emphasises that such judgements are to be seen not merely as necessary constituent elements in emotion, but also as sufficient ones.77 But what about cases where the correct evaluative beliefs are in place, but the emotion is not felt –

I know that a child has been sexually abused, what a serious effect such abuse can have on the long-term well-being of the child, and how wrong it is to commit such atrocities; what if, however, I am simply not concerned about the interests of this particular child (or children in general) and do not experience any compassion towards it? Nussbaum’s answer is that in such cases we need to distinguish between ‘really accepting a proposition and simply mouthing the words’. Parroting evaluative sentences does not mean 28

M A P P I N G O U T T H E F I E L D

that I really believe them; they need to become part of my cognitive makeup and motivation for action before we can start to talk about real beliefs.78

Here Nussbaum has gone as far in the direction away from the fact-value distinction as possible; it is not only that evaluation involves, rather than supervenes upon, description; it also incorporates motivation. This is, I believe, throwing out the baby with the bathwater. There remains a valid distinction, not between description and moral evaluation, but rather between morally evaluative description and motivation. This distinction is commonly obscured in moral theory by talk about ‘moral’ or ‘normative’

judgements which can cover both; thus the social-science textbook staple about ‘normative conclusions only following from normative premises’. If this means that evaluative descriptions can only follow from other such descriptions, Anscombe’s example refutes it: The non-evaluative description of how a grocer carries potatoes to my house conversationally implies through a number of steps the evaluative description that it would, ceteris paribus, be morally right of me to pay the grocer money. However, it does not follow that I must wish that the grocer be paid or that I actually choose to do so, all things considered. Theorists, whether engaged in emotion theory or pure moral theory, would do well to pay heed to Foot’s point that it is one’s will, ‘something the agent wants or which it is in his interest to have’, which turns a moral evaluation into a motivation.79 An agent can hold true evaluative beliefs about the plight of sexually abused children – beliefs which have really sunk in and are not only superficially parroted – without being stirred at all to compassion and compassionate action, because the motivation element is lacking; a drug baron in Columbia can be correctly informed about the effects of his drugs on their addicts, and about his own moral failings, without having the slightest inclination to mend his ways because he is, as it happens, not concerned about the suffering of others or about the demands of morality. If we deny this, we end up in the cloud-cuckoo land of believing that all moral callousness and wrongdoing must be attributable to ignorance or mental illness, an assumption which (contrary to Socrates and the optimistic nineteenth- and early twentieth-century reformation theories of punishment) simply flies in the face of all experience. The worst evildoers are, indeed, those who have perfect understanding of the moral evaluations in question and decide to act contrary to them. Moral evaluative judgements thus provide at best, as Foot states, a set of hypothetical imperatives:80 if you are concerned about the substantive conditions involved in the evaluation, then you experience the relevant emotion(s) and (perhaps) decide to stir a finger. Hence, I conclude that it is necessary to include the concern-component in the specification of emotion. Without it, that is, without desires and aversions, there would be no experienced emotions and no motivations to act upon them. Or as Richard Wollheim puts it: ‘That emotion rides into our lives on the back of desire is a crucial fact about emotion, as well as a crucial fact about us.’81

M A P P I N G O U T T H E F I E L D

What about the third traditional component: feeling? It would take me too far away from our present concerns to respond here to those who believe in (episodic) desires without feelings: to behaviourists or those artificial-intelligence theorists who claim that a computer program can, in principle, have the same desire to win a game as the ordinary chess-player. However, at least in the area of emotions, the idea of a frustrated desire without, so to speak, any frustration, seems thoroughly implausible. How ‘calm’ can one’s desire be for it to remain a desire?82 If no ‘botherment’ occurs after my car is stolen, how can I still claim, other things being equal, that I had a desire that it should not be stolen, and that I am really angry? I do not think I can. If I am right in this, it means that we do not have to add the feeling element to emotion as an independent (third) necessary component, but rather consider it an implication of the conative component: if my desire is satisfied, I feel content; if it is frustrated, I feel bothered; but the feelings of contentment or botherment can express themselves differently at different times, or in different emotions, or even (physiologically and phenomenologically) identically in otherwise diverse emotions, as we saw in section 1.2. I still want to retain the possibility that the feeling is unconscious. This is not tantamount to claiming that there are emotions without the feeling component. When we speak of having been unconsciously jealous of B, we do not mean that we were jealous without any feelings; what we mean is rather that we were self-deceived about either (a) our evaluative beliefs concerning B, (b) our uncomfortable feelings, or (c) the connection between the two. In a typical case, we undoubtedly felt bad (maybe we suffered from headache or stomach upset) but we did not realise until later where these feelings stemmed from. So the feelings were unconscious qua feelings of jealousy, not qua feelings per se. In this sense, the reference to unconscious feelings need not carry the dubious psychological baggage at which many contemporary writers look askance.

Accepting the redundancy of a value-free ‘cognition’ and of a logically independent feeling component does not mean that we are yet home and dry as far as the essential components of emotion are concerned. The course of our discussion now brings us to a point where another and perhaps greater difficulty awaits us, for we still need to ask whether the evaluative component necessarily involves a belief or not. Recall the short exploration of

‘inert’ emotions in section 1.2: coming to know that a person did not really, as I previously believed, wrong me does not mean that my anger is automatically terminated.83 It might still hold me in grip, however irrationally. How can that be if the emotion is really grounded in my evaluative belief ? Rather than giving up the cognitive component altogether, many thinkers have wanted to cast the notion of belief in the role of villain here. What lies at the heart of emotion – so the argument goes – is not a full-bodied belief but rather the experience of ‘seeing-as’ (recall the Calhoun–Roberts solution): dwelling on or attending to in consciousness, grasping one thing in terms of 30

M A P P I N G O U T T H E F I E L D

another. Hence, instead of beliefs, we should talk about ‘construals’, ‘modes of attention’, or ‘recognitions’: propositional attitudes with an intentional focus which still fall short of strict beliefs.84 Nancy Sherman even attempts to show that Aristotle had thought of this problem and suggested a solution along the construal-lines,85 an attempt which seems to me based on a rather strained reading as Aristotle explicitly said that ‘imagining’ is not sufficient to ground emotion. By merely imagining what is fearful, according to Aristotle, like a person looking at a painting of some dreadful scene, we may become physically aroused, but we do not experience the emotion itself.

That only happens after we come to think something to be truly fearful or threatening.86

If we relax the demand that our emotional evaluations involve beliefs, we evade the conceptual problem of inert emotions suggested above; we are then, when such emotions occur, simply evaluatively construing a state of affairs in which we do not believe. However, I am afraid that this solution may import as many difficulties as it removes. The problem is that more conceptual freight seems to be heaped on the notion of ‘construal’ (or its equally contrived sister-notions) than it can bear. Calhoun and Roberts do not want to abandon the condition that the construal has propositional content. It must, in principle, be articulable (linguistically formulable), although it is not articulated. Yet it is not allowed to have the status of a belief, not even of a ‘vague’ or ‘insufficiently articulated’ belief. At the same time, the construal is more than an ordinary perception in that it must have (prospective) truth-value. Perhaps we should envisage it as some kind of belief-piece from a jigsaw puzzle which has not yet been arranged. But the snag is that this specification of construal forfeits its similarity to belief in direct proportion as it achieves its relation to perception, and vice versa. It thus becomes an all too easy prey for Ockham’s razor: is there really a place for an intermediate mental entity between clear perception and vague belief ?

If the reader has the same impression here as I have, that the whole construal-thesis is but a sledgehammer to crack a nut, it is because a much simpler solution seems to suggest itself: does the person still experiencing anger after realising that he was not wronged, or a fear of spiders in spite of knowing that spiders are harmless, not simply engage in doublethink, that powerful manoeuvre which Orwell so famously defined in 1984 as holding two contradictory beliefs in one’s mind simultaneously? More academically oriented people would probably want to call it wilful self-deception: holding

‘deep down’ or ‘high up’ with lessened awareness unacknowledged beliefs (about being slighted by the person at whom our anger is directed; about the harmfulness of spiders), beliefs which can still hold us captive for shorter or longer periods, while attending, in direct consciousness, to the contradictory rational and veridical beliefs.87 The common discrepancy between emotion and belief is then not between emotion and belief per se but rather between emotion and avowed belief.

M A P P I N G O U T T H E F I E L D

Roberts is unhappy about the self-deception solution, finding it implausible that I can ‘either simultaneously or in rapid succession’ make contradictory judgements (about spiders, being slighted by another person or not, etc.).88 His doubts echo those of many philosophers who see something inherently misguided or even strictly illogical about the very notion of self-deception: How can I at the same time be the source and object of deception, the deceiver and the one deceived? And how can I decide to ignore a certain part of my belief-set without having precisely located what I want to ignore, thereby constantly attending to that which I do not want to attend? Self-deception has fallen on hard times in philosophical circles, and that in spite of ingenious suggestions as to how its logical paradox can be solved by assuming that the contradictory beliefs are held at ‘different levels’

or attended to in ‘different respects’.89 This presents, however, a typical case of academic misgivings being at odds with ordinary intuitions, and in such cases it is not always the latter which must be relinquished.90

As Nussbaum notes, it is the most common of truths that we often hold contradictory beliefs, especially in cases involving long habituation – citing her own example of still irrationally believing deep down that the US

Supreme Court is in California, owing to a deeply-ingrained childhood misunderstanding, although her sober self knows perfectly well that this belief is false. If this can happen with matters on which nothing depends, such as the location of the court, for the person holding the belief, then it is even more likely to be true with strong evaluative beliefs, often laid down in childhood or in times of affective vulnerability.91 However, owing to the relative brevity and instability of most (if, obviously, not all) episodic emotions as compared with other propositional attitudes, their grip on us, in cases of self-deception, tends to be relatively short-lived. For instance, while the belief that all blacks are stupid, held by a person who deep down knows better, can hold sway unchallenged for extended periods, the belief that spiders are harmful will typically only induce the emotion of fear in us sporadically, hitting us in unguarded moments.92 If the emotion lingers on, it will do so at a subordinate level of consciousness while the contrary veridical belief operates and rules at a higher level.

The idea of beliefs lodging deeply enough in us to alter our cognitive lives, while not avowed or directly attended to, is an idea entertained on the grounds of daily experience. Philosophers, who continue to ask in bewilder-ment how that can happen, may here learn a lot from psychologists who are normally more interested in how it does happen. While the former tend to be stuck in unrewarding misgivings about the Freudian notion of the unconscious, the latter typically state that ‘it is possible to be conscious of one interpretation of an event whilst also holding an unconscious and contradictory interpretation’93 – and then go on from there. Power and Dalgleish, for instance, propose a plausible psychological model according to which there are two main routes to the generation of emotion: a rational (paradigmatic) 32

M A P P I N G O U T T H E F I E L D

‘schematic-appraisal’ route, and an associatively driven ‘automatic’ route, where a rational, non-self-deceptive level of meaning is ‘short-circuited’

owing to some personal inhibitions.94 Admittedly, this model is a little more advanced than the simple generalisations of dinner-table talk about irrational emotions and self-deception, but the upshot is more or less the same in accounting for contradictory beliefs. It is tempting to conclude that the invocation of construals is but a distraction in the discussion of the ingenious tactics employed by our mind to preserve and manipulate entrenched evaluations and inure them against conflicting evidence. We do not need to give up the view of evaluation as grounded in the much less conceptually cumbersome notion of belief to make sense of those tactics.

It might be argued at this point, however, that I have only traded one loathsome thing for another. Although the construal-solution is even less capable than the belief one of accounting for emotional inertia, it does not mean that the latter is satisfactory. Perhaps – in order to make full sense of our emotional landscape – we must abandon even more drastically than the construal-theorists dream of the condition that the cognitive component of emotion is linguistically formulable. Perhaps it need not have any propositional content at all. This is the point of John Deigh’s attack on cognitive theories of emotion, the most serious and full-blooded one mounted in recent years.95 Deigh does not reject the insights of cognitivism altogether; for instance, he readily admits that emotions are intentional, and thus distinguishable from mere feelings. What he opposes, however, is the ‘intellectual sabotage’ of cognitivism which has gradually taken place during the last decades. He suggests a return to an older form of cognitivism, represented in works produced during the early part of the century by English psychologists and philosophers. There, ‘cognition’ referred to any mental state in which the subject was cognisant of some object (including the states of perceiving, imagining, or remembering). Nothing in this ‘traditional’ understanding implied that the cognition must have propositional content: that the person affirm or even consider any proposition whatsoever. In other words, something can count as an intentional object without the subject entertaining any beliefs about it, however vague, or being in the grip of a linguistically formulable (if as yet unformulated) ‘construal’: ‘When a baby or a cat stares at you, you are the object of its stare. Yet it does not follow that the baby or the cat has any beliefs about you. When a dog relishes a bone, the bone is the object of its delight. Yet it does not follow that the dog has any beliefs about the bone.’96

Robin Dillon has recently used Deigh’s insights to reformulate the cognitive-theory explanation of how successful women entertaining true beliefs about their self-worth can still suffer from irrational bouts of shame and resentment. The reason for these anomalous emotions, Dillon argues, is not, in all likelihood, that there are other irrational beliefs or even construals operating somewhere within these women – at least not in cases where we 33

M A P P I N G O U T T H E F I E L D

have no independent rationale to posit the beliefs’ existence except as an explanatory tool to make sense of the emotional incongruity. Rather, she says, we need to distinguish between two modes of understanding, ‘intellectual understanding’ and ‘experiential understanding’, where the latter involves experiencing something directly and feeling the truth of what is experienced without involving any actual beliefs or judgements, that is to say, without any thought expressed or expressible in propositional representations which have a truth-value. These women can then be considered to both understand (intellectually) and not understand (experientially) their worth at the same time.97

The main objection that I want to raise to this ‘back-to-basics’ version of cognitivism is that it fails to exclude the possibility that emotions can be ascribed to infants and animals. Now this may seem to be a rather obtuse objection, perhaps even defiantly so. The main spur of Deigh’s writing is specifically to counter the cognitivist ‘inattention to the emotions of beasts and babies’,98 along with the intellectualisation of those primitive emotions of fear, anger, love, sexual passion, and sorrow which we have in common with the animals and are, for Deigh, essentially unresponsive to reason. To argue my case, let me first point out that emotion-terms such as ‘fear’ are not always used to refer to emotions. A person in an objectless state of anxiety is often said to be suffering from ‘fear’. A person caught with an eerie feeling of discomfort upon entering premises remotely similar to the site of an earlier horrifying experience may be described as being in a ‘state of fear’. A person suddenly looking down from a precipice may be ‘scared as hell’. However, there are more reasonable ways of accounting for these states of experience than referring to them as the emotion of fear: the first person is the grip of a mood, that ‘metaphysical generalization’ of an emotion (see s.

1.2); the second experiences a conditioned reflex; the third has a startle-response.99

These distinctions are not arbitrary; they help to mark out psychologically and morally important aspects of human experiences which distinguish them from real emotions. By this I am not saying that emotions cannot be constituted by, among other things, automatic responses. Such responses were, for instance, plausibly invoked by Power and Dalgleish, as mentioned above, to account for irrational emotions. However, automatic responses of that kind need to be distinguished from objectless moods, and from pre-linguistic biological responses/reflexes which do not involve any beliefs, not even self-deceptive ones. That distinction is acknowledged in everyday speech; when pressed, most people would be willing to grant that, while often referred to as ‘fear’, the three kinds of experiences described at the end of the last paragraph are conceptually distinguishable – for example, from one’s rational fear of street muggers and even from a typical sporadically appearing irrational fear of spiders. Furthermore, this distinction is helpful when coming to terms with alleged infant emotions. For instance, there was 34

M A P P I N G O U T T H E F I E L D

a time when my son reacted anxiously toward patches of light on the wall above his bed, caused by a sudden reflection from moving curtains. I think there is every reason to keep those experiences distinct from the fear he expressed later, after acquiring language, of these patches as ‘animals biting people’. Of course, there may have been an intermediate period, around the time when he was learning to speak, when his ‘fearful behaviour’ was difficult to categorise as either indicating a startle-response or a fear, but that does not change the fact that there is a valid distinction to be drawn between the two. My complaint about Deigh’s ‘traditional’ cognitivism is that it obliterates such distinctions and makes the notion of fear become bloated beyond good sense: an error from which the more mainstream varieties of cognitivism escape by leaning, like Hope on her anchor, on belief as the cognitive component of emotion.

Strangely enough, claiming that animals do not experience emotions (as distinct from moods and automatic responses) is likely to arouse even greater fury in people holding the opposite view than a similar claim about infants. Even as sober a thinker as Ben-Ze’ev states defiantly that any claim about animals’ lack of emotion is ‘refuted by ethological research as well as by common sense observations (at least by those who have animals around them)’.100 As a matter of fact, I often used to have animals around me as a child, being a frequent visitor on my aunt’s farm, where I even, officially, owned a sagacious old dog. This dog was so ‘wise’ that when a new puppy appeared on the farm who disturbed his midday siestas and arrested attention at the old dog’s expense, he took the puppy in his mouth, walked with it a considerable distance (even swimming across a river) and left it there. This became a repeated ritual, happily explained by my aunt as an example of the old dog’s jealousy: he ‘resented’ the ‘undeserved’ attention the puppy received, and ‘knowing’ that the further away from the farm he took it, the

‘less likely’ the puppy would be to find its way back home, he ‘decided’ to embark on this long journey. Not only do I now think that my old dog lacked the intellectual repertoire to perform any of the complex logical and statistical manoeuvres posited by my aunt, I do not see the need to ascribe to him any emotion either. He was not jealous but rather, I believe, biologically driven to distance himself from a negative stimulus as much as possible or, more accurately, to distance the stimulus as much as possible from himself.

Perhaps dogs do have beliefs; that is basically an empirical issue.

However, given what we know at present, my dog’s behaviour admits of a much simpler and theoretically less cumbersome explanation than that given by my aunt. People’s reluctance to use Ockham’s razor in making sense of animal behaviour can be accounted for in many ways. One reason is the well-known attraction of the pathetic fallacy; another (and quite an opposite one at that) is the temptation to assimilate some of our own ‘negative’

emotions symbolically to alleged animal emotions in order to explain the 35

M A P P I N G O U T T H E F I E L D

former away as ‘brutish’ or ‘bestial’.101 In any case, I agree with the psychologist J. R. Averill that our understanding of the nature of a specific emotion

– be it fear, anger, or the arguably more complex pride or jealousy – as distinct from the evolutionary story behind it (see ch. 7), is not enhanced much by studying superficially analogous behaviour in animals.102

1.4 Preliminary remarks on responsibility, moral justification, and the ‘negative’ emotions

After having descended temporarily to some of the specifics of the ongoing debate within cognitivism, where, as always, close cousins make the worst enemies, it is now time to ascend again to a more general perspective of our field. Since my aim is to evaluate morally and defend certain kinds of emotions, it is in order to say something first about the connection between moral evaluation and responsibility.

Generally speaking, we can only be held morally responsible for that which is within our voluntary control. If I break the glass door of a jewellery store in order to burgle it, I can be held responsible for my action since it is presumably the result of my deliberation and choice: I have decided to commit this offence. However, if my body is blown through the glass door in a hurricane, I am free of responsibility because that unfortunate result is not one over which I have any voluntary control. At most, I could be held responsible for deciding to seek shelter in front of a glass door in a hurricane, rather than in some other equally convenient place nearby, when I should have known that this might well cause an accident.

It should be noted here, if only as an aside, that the fact that an outcome is under our voluntary control, while necessary for moral responsibility, is not sufficient for it. The fact that it is within my voluntary control to tie the loose shoe laces of all the children in my neighbourhood, by devoting my attention to that and nothing else, does not make me morally responsible for every local accident resulting from loose laces, for there is simply no good reason (moral, statistical, or otherwise) why I should be expected to make such a project my sole preoccupation. Ascribing moral responsibility thus involves considerations of reasonable expectations. Even when I am immediately causally responsible for an outcome, it does not necessarily follow that I am morally responsible for it: I am not morally responsible for your being late for work simply in virtue of the fact that my driving in front of you at a reasonable speed prevented you from speeding and getting to work in time.103 In the following, whenever I mention ‘responsibility’, it should be taken to mean ‘moral’ rather than ‘causal’ responsibility.

However, as any interesting human relationships involving pride and jealousy – the emotions primarily under scrutiny in the present work – are bound to be more intimate and direct than those between me and unknown children with loose shoe laces, or between me and you where I coinciden-36

M A P P I N G O U T T H E F I E L D

tally disturb your preferred rhythm of driving to work, the caveats mentioned in this paragraph about disharmony between voluntary control and moral responsibility may well be of limited relevance for what follows.

An objection frequently levelled at cognitive theories of emotion is this: by pointing to a cognitive element as the alpha and omega of emotions, you have placed them within the purview of responsibility. If emotions really are rational and purposive rather than irrational and disruptive, then they are very much like actions, and you will have to hold that we choose an emotion much as we choose a course of action. However, this flies in the face of the accumulated experience of mankind. Simply look at our literature which is filled with tales of people who are captives of their emotions:

‘struck’ by jealousy, ‘paralysed’ by guilt, ‘plagued’ by remorse, ‘felled’ by shame, ‘blinded’ by passion (witness the notorious crime passionel ), and so forth.104 Could anyone but the most naive rationalist claim that all these locutions are but convenient feel-good excuses for reactions which are in fact within our voluntary control? Indeed not, the objector would say: emotions are occurrences that happen to us; they bring to the surface that part of our animal-physical nature which often jars with our rational-intellectual life; they are like natural secretions from within, invading our freedom. Maybe one can be held responsible for failing to comport oneself with dignity – putting up an appearance of nonchalance – even if one’s soul is burning with the vapours from those natural secretions. But one can surely not be held responsible for failing to turn off the secretions themselves, any more than one can be held responsible for not stopping the production of saliva in one’s mouth. Whether or not it should be spat out or swallowed is another story.

No doubt a lingering sense of dissatisfaction with cognitive theories of emotions will remain until this objection has been rebutted. A possible first recourse might be to grant that while emotions qua dispositions are outside our voluntary control, and hence beyond the bounds of responsibility, each particular occurrence of an emotion is not. The parallel there could be alcoholism. A reasonable view of alcoholism would hold that while the alcoholic’s desire for liquor is uncontrollable in the sense that he cannot stop having the desire, it is controllable in another sense: the alcoholic may use his will-power to refrain from acting on this desire. He decides not to drink, however unquenchable the underlying desire is.

Unfortunately, this parallel does not work in the case of emotions. A person with a strong tendency to jealousy does not decide in each particular case whether to be jealous or not. Given a certain context, he becomes jealous. Likewise, the phlegmatic laggard does not decide to become angry, even if he should be, if he does not have in him the disposition to react angrily when seriously wronged. In other words, emotions seem to differ from ordinary actions in that they do not allow for the same kind of discrepancy between disposition and occurrence, for instance refraining from x, as 37

M A P P I N G O U T T H E F I E L D

a result of deliberation in the particular case, when you have a strong tendency to x. When locating responsibility for emotions, our focus must thus be on dispositions rather than on particular occurrences. Justin Oakley reaches a similar conclusion, though from a different angle, when he argues that not only in the case of emotions, but also of many physical acts, the ability to stop something here and now is not a necessary condition for assigning responsibility. We need also to consider how the person came to be in the position where his doing, or having, this ‘something’ is now unavoidable. Oakley’s example is that although I cannot stop myself from driving off a cliff due to the failure of my car’s brakes, I can still be held responsible for driving off the cliff if I have carelessly failed to have the brakes checked. The resulting view is that ‘we are responsible for something only if we could have at some time avoided doing or having it’. So even if it is true that we cannot cut off our emotions at will once we have them, it may be that we could have taken various measures earlier to control their onset.105

The crucial question then becomes how and/or to what extent can we generally be held responsible for our personalities, for the formation of our own dispositions and character traits? To answer that question, we need to enter a field much broader than that of emotion theory, a field which has to do with fundamental questions of (moral) upbringing and (moral) education. Aristotle is particularly enlightening on these points, but since it would be getting ahead of my argument to introduce his insights here, let me shelve the details of his discussion until section 6.1 – simply presupposing in the meantime, for argument’s sake, that we may well be, and in fact are, morally responsible for various emotions although their appearance at the present time is not under our direct control.

We should at this point, however, pause to ask whether we may have been mistaken in viewing the possible lack of responsibility for our emotions as a threat to the book’s objectives of morally justifying emotions. My underlying assumption here is a widely held and powerful one in moral philosophy: that moral evaluations can only be properly directed at that for which we can be held responsible. R. M. Adams argues in a paper defiantly entitled ‘Involuntary Sins’ that this common assumption is misconceived.106

Vices are vices and virtues are virtues, however they have been come by.

Similarly, an emotional condition, such as invidious envy, to which opprobrium is typically attached, will, in Adams’s view be a proper object of censure in its own right, even if it is an involuntary state of mind. We should bear in mind that blameworthiness does not imply the necessary open expression of blame.107 There might be overriding (utilitarian) reasons for remaining silent: every parent knows that even after a child becomes responsible for its actions, it would be counter-productive to scold it for every misdeed. This reservation, however, in no way detracts from the seriousness of Adams’s challenge to the received wisdom about the relationship between responsibility and blameworthiness.

M A P P I N G O U T T H E F I E L D

Adams considers the beliefs of an officer bred up from childhood in the Hitler- Jugend, and accepts that such brainwashing could debar the person from forming autonomous beliefs and passing reflective judgements. Still, according to Adams, his beliefs would be evil and make him a fitting object of reproach. How he acquired them would not warrant exemption from blame. If this is so, we do not need to consider people’s personal histories before assigning blame or credit to them for their emotions. It does not matter whether we have any chance of changing our existent emotions, or whether we could have brought it about earlier that our underlying emotional dispositions were any different from what they happen to be. If the emotion is of a vicious nature, then we are blameworthy for having it; if it is of a virtuous kind, then having it counts as a moral credit. Adams anticipates an obvious objection: that attitudes, such as those of the Hitler- Jugend officer, can be appropriate objects of other moral attitudes, although not of praise or blame. We can, in other words, ‘think poorly’ of the person for having such attitudes, pity him or even look down on him, but we cannot blame the person, for his emotions and beliefs are not voluntary.

However, Adams rejects this distinction, insisting that heinous attitudes are blameworthy as such.

Now, it is true that we sometimes do blame or praise things without ascribing responsibility to them. For instance, I might blame my broken-down car for my being late for work (and that is not necessarily short for blaming the car mechanic who last fixed it, or the person who sold it to me).

Or I might praise my new computer which makes the word processing of this book so much easier than it would otherwise have been. In this sense

‘blaming’ simply means ‘giving the reason for a misfortune or why something does not work’, and ‘praising’ the opposite. However, as part of our language game featuring emotions and attitudes, ‘blame’ is rarely used in this way. When I blame my wife’s anger yesterday for my sadness and disappointment today, I am typically giving more than a functional explanation of my present state of mind; I am making a moral claim, expressing a grievance. To be sure, if my friend were killed by a robot, I might blame the robot for his death in the same sense as I blamed the car for my being late to work. But if I were concentrating on the robot’s motives for its deed, it would be most outlandish to say that I ‘blamed’ the robot for its undeserved

‘anger’ at my friend, which led to his killing. Even granting that it makes some sense to talk of a robot being angry (and I am not sure it does), any possible blame of this kind would be ascribed to the person who programmed the robot. Moral blame for emotions requires responsibility, which again presupposes voluntary control. What Adams trades on seems to be little more than the fact that blaming has other roles to play in (perhaps) equally interesting contexts. However, what he says about its role in the

‘game’ in which we are interested here remains unconvincing.

Still, we should not be too carried away by this counter-argument, for 39

M A P P I N G O U T T H E F I E L D

Oakley suggests some much more initially plausible reasons to be sceptical of the assumption that moral evaluations require responsibility. He does not, like Adams, attack the unlikely targets of fixing praise and blame, but claims instead that there are other significant moral evaluations which can be legitimately applied to us in the absence of responsibility, for instance to those emotions of ours which we are not responsible for having.108 He calls these evaluations moral assessments of estimability and disestimability. They distinguish themselves, among other things, from evaluations of blame- or praiseworthiness in being directed at features of the persons themselves, rather than at the situations in which these persons find themselves and the actions they perform there. Thus, esteem or disesteem can be fixed on us because of our beauty or ugliness, intelligence or lack of intelligence, riches or poverty, and in the case of emotions, to our compassion even if/when this is a purely natural sentiment, or to our malice even if/when its onset is completely outside our control (now or earlier). Nevertheless, these evaluations are morally significant in that they can be relevant to the general estimation of a person’s moral character, and hence they undermine the traditional ‘evaluation-requires-responsibility’ assumption.

Far be it from me to argue against Oakley that esteem and disesteem are not morally significant attitudes. However, I think he may underestimate the way in which we commonly assume, when passing judgements of the kind he mentions, that the persons so judged have played some part, however small, in the creation or non-suppression of the (dis)estimable condition. For instance, we may feel that an ugly person could have done a little bit more to look presentable, or that, while of a compassionate constitution, the amount or kind expressed in a particular case goes beyond the person’s natural tendency. Beyond this, I would also say that Oakley seems to overstate the common assumption that he attacks in order to provide a focus for his own thesis. The idea behind saying that moral evaluation requires responsibility has, I think, never been that assessments of outcomes for which a particular agent is not morally responsible cannot be morally significant. To return to Adams’s example of the Hitler- Jugend officer, an expression of disesteem of his character can be morally significant in various ways. It can set a moral example to others by providing evidence for what happens when young souls are inculcated with evil; it can give a morally significant instrumental explanation for the plight of my life as marred since I got into his hands (just as explaining why I feel psychologically and morally debilitated after being struck by a serious illness can give a morally significant reason for my present state without ascribing moral responsibility to the disease), and, additionally, it can be read as an indirect imputation of blame on those who brainwashed the officer. In other words, various kinds of moral evaluations of a person’s character or behaviour (call them esteem and disesteem if you like) are possible without presupposing moral responsibility. It should be noted, however, that (a) they are not moral evaluations of the person qua 40

M A P P I N G O U T T H E F I E L D

agent, and (b) they do presuppose a world in which there are other responsible agents, to whom blame can be ascribed, who can learn from example, etc. An evaluation of the Hitler- Jugend officer qua agent must be one of aesthetic, rather than moral, significance, and in a world where no one could be held responsible for anything, all evaluations would be reduced to the aesthetic kind. Thus, I claim ( pace Oakley) that the moral evaluation of a particular emotion, in so far as the evaluation relates to the person having the emotion as a moral agent, cannot be more than an appendage to the assumption of responsibility for our character traits, for being the kind of persons we are.

As a final comment here, let me anticipate the response of a reader who has taken in my earlier message (s. 1.1) about being guided, inter alia, by Aristotle. ‘Does not Aristotle’s moral code’, the reader might ask, ‘essentially presuppose what you deny, namely, that moral evaluations of persons, their actions and emotions, often rely heavily on circumstance: conditions completely beyond the persons’ control?’109 It would take me too far afield to give a satisfactory answer to that question here; let me simply say for the moment that this may be an oversimplification of Aristotle’s view, and even if it was his view, there are other ways of getting to the conclusions that he intended (see s. 4.2) without scrapping the ‘evaluation-requires-responsibility’ assumption as I have interpreted it above.

The importance of responsibility notwithstanding, knowing that an emotion falls within the province of the agent’s responsibility is only the first step in morally evaluating it as worthy of praise or blame. In general, we can say that an emotion is morally justifiable ‘if and only if the evoking object or situation warrants the emotion’.110 Given the pervasiveness and salience of the emotions as sources of moral evaluation (‘He is a jealous bastard’!) and motivation (‘She acted out of sympathy’), testing the justifiability of emotions according to this criterion is a constant challenge. It is no coincidence how many of the traditional Christian virtues and vices bear the names of emotions: pride, envy, and anger (as deadly sins), and hope (as a theological virtue), to name but a few. De Sousa hardly exaggerates when he says that ‘most of what is morally interesting about human life is played out in the domain of the emotions’.111 Indeed, if we consider the everyday judgements passed about other people at the dinner table or while doing the washing-up, many of these will, I think, be found to relate to how other people reacted emotionally to us during the day in morally appropriate or inappropriate ways. In academic discussions about morality, emotions may not be as predominant as in kitchen-talk, but they still chart a considerable part of the terrain of ethical inquiry.

How, then, do we perform the test of moral justifiability: when does the