Hume on Art, Emotion, and Superstition: A Critical Study of the Four Dissertations (Routledge Studies in Eighteenth-Century Philosophy) 9781138351462, 9780429435263, 1138351466

Citation preview

Hume on Art, Emotion, and Superstition

This book proposes a bold new thesis on the development of Hume’s thought, challenging long-held opinions. It demonstrates that Hume’s account of the passions is more central to his positions on other topics than has up to now been assumed and makes a strong case for why it should be recognized to have changed over time, contesting the received view that Hume’s philosophical opinions were constant throughout his lifetime. —Lorne Falkenstein, University of Western Ontario, Canada This book offers the first comprehensive critical study of David Hume’s Four Dissertations of 1757, containing the Natural History of ­Religion, the Dissertation on the Passions, and the two essays Of Tragedy and Of the Standard of Taste. The author defends two important claims. The first is that these four works were not published together merely for ­ convenience, but that they form a tightly integrated set, unified by the s­ubject matter of the passions. The second is that the theory passions they jointly present is significantly different—indeed, of the ­ ­significantly ­improved—from that of the earlier Treatise. Most strikingly, it is anti-egoist and anti-hedonist about motivation, where the Treatise had ­espoused a Lockean hedonism and egoism. It is also more ­cognitivist in its analysis of the passions themselves, and demonstrates a greater awareness of the limits of sympathy and of the varieties of human taste. This book is an important contribution to the scholarly literature on Hume’s work on the passions, art, and superstitious belief. Amyas Merivale works at the Faculty of Philosophy, University of Oxford, UK. He is the co-editor of Hume’s Enquiry concerning the ­ ­Principles of Morals, with Writings on Aesthetics and the Passions, and the developer of davidhume.org. His published work has appeared in the British Journal of Aesthetics and Hume Studies.

Routledge Studies in Eighteenth-Century Philosophy

Kant’s Inferentialism The Case Against Hume David Landy Rousseau’s Ethics of Truth A Sublime Science of Simple Souls Jason Neidleman ant and the Scottish Enlightenment K Edited by Elizabeth Robinson and Chris W. Surprenant ant and the Reorientation of Aesthetics K Finding the World Joseph J. Tinguely ume’s Science of Human Nature H Scientific Realism, Reason, and Substantial Explanation David Landy ume’s Moral Philosophy and Contemporary Psychology H Edited by Philip A. Reed and Rico Vitz ant and the Problem of Self-Knowledge K Luca Forgione ant on Intuition K Western and Asian Perspectives on Transcendental Idealism Edited by Stephen R. Palmquist ume on Art, Emotion, and Superstition H A Critical Study of the Four Dissertations Amyas Merivale For more information about this series, please visit: www.routledge.com/ Routledge-Studies-in-Eighteenth-Century-Philosophy/book-series/SE0391

Hume on Art, Emotion, and Superstition A Critical Study of the Four Dissertations Amyas Merivale

First published 2019 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2019 Taylor & Francis The right of Amyas Merivale to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested. ISBN: 978-1-138-35146-2 (hbk) ISBN: 978-0-429-43526-3 (ebk) Typeset in Sabon by Apex CoVantage, LLC

In all compositions of genius, therefore, it is requisite, that the writer have some plan or object; and though he may be hurried from this plan by the vehemence of thought, as in an ode, or drop it carelessly, as in an epistle or essay, there must appear some aim or intention, in his first setting out, if not in the composition of the whole work. A production without a design would resemble more the ravings of a madman, than the sober efforts of genius and learning. An Enquiry concerning Human Understanding, 3.5

Contents

Acknowledgements Conventions

ix xi

Introduction

1

PART I

7

  1 My Design in the Present Work

9

  2 Some Late Philosophers in England

30

  3 Founded on Pain and Pleasure

49

  4 A Considerable Adjustment

70

PART II91

  5 The Religious Passion

93

  6 The First Religious Principles

111

  7 The Object of the Passions

128

  8 The Combat of Passion and Reason

145

  9 The Causes of the Violent Passions

162

10 The Predominant Passion

176

11 The Sentiments of Beauty

191

viii Contents 12 The Laws of Criticism

210

Conclusion

226



Appendix 1. The Meaning of All the Terms230 Appendix 2. Comparison of Ideas233 Index237

Acknowledgements

This book started life as my doctoral dissertation at the University of Leeds. I am enormously indebted to my principal supervisor there, Helen Steward, for her consistently close and insightful readings of countless drafts, and our many enjoyable and profitable supervision sessions. When I started out, I had little idea where I would end up; without her to bounce ideas off, to give me new ideas, and to keep me on track when my mind wandered, I  do not know how I  would ever have got here. I am also very grateful to my secondary supervisor, Matthew Kieran, for his assistance with the later chapters on aesthetics. His knowledge was indispensable. Before moving to Leeds, I took the BPhil at Oxford, largely under the supervision of Peter Millican. He had moved there from Leeds just as I started the BPhil, and we first met at a welcome drinks reception. We got chatting about Hume, realized he had been assigned to supervise me, and I mentioned nervously—not yet being familiar with his work—that I wasn’t so keen on the Treatise, and wanted to concentrate my studies on the first Enquiry. We have been firm friends ever since, and when I started my doctorate at Leeds he offered to act as informal supervisor, as a favour to his old department (which, notwithstanding Helen’s wisdom and general interest, lacked a dedicated Hume scholar). It is impossible to say how much this book owes to his input; we agree to such an extent, and have discussed these things so much, that I can no longer recall where his ideas end and mine begin. In the writing of my thesis, the development of it into this book, and more generally the starting of my academic career, he has been an invaluable mentor, collaborator, and friend. I am very grateful to Aaron Meskin and Simon Blackburn, the internal and external examiners for my doctorate respectively, for their constructive feedback on the thesis, and for their encouragement in seeking to publish it. Aaron has since been a great support in the forging of my career, and it was Simon who put me in touch with Routledge. After submitting my thesis, I was awarded the Blackham Fellowship by the Conway Hall Ethical Society, Humanists UK (formerly the British Humanist Association), and the Rationalist Association, providing

x Acknowledgements me with a grant for developing my doctoral research into a publishable monograph. As a result of this, I was able to complete the first rewrite of my thesis within a year, when I would otherwise have needed to take on more teaching work to keep afloat. My thanks to Jim Walsh and to everyone who made this possible. In 2010 and 2011 I  had the pleasure of delivering some of my earlier ideas to the Hume Society at two of its annual conferences, first in Antwerp and then in Edinburgh. I was immediately struck, as so many others have been, by the convivial and supportive atmosphere created by the Society’s members. Disagreements could scarcely be pursued in a more agreeable manner, and I consider the Society to be both a model of how philosophical debate should be conducted and an honour to Hume’s memory. After a long absence, I had the pleasure of its company again just as this book was going through the press, at the 2018 conference in Budapest. I am grateful to Jane McIntyre for her comments on my paper, and to the participants of the enjoyable discussion that followed. Some of the questions raised, I believe, are already anticipated here, but there was no time to address those that are not; I look forward to pursuing these debates in the future. At various times in the course of this project I have benefited greatly from discussions with James Arnold, Filippo Contesi, Don Garrett, Lorenzo Greco, James Harris, Dan O’Brien, Hsueh Qu, and Samuel Rickless. I am particularly grateful to James Arnold and Peter Millican for volunteering to read the entire manuscript through just before submission, and for pointing out a number of errors and infelicities. This is my first experience of publishing a book, and I would like to thank Andrew Wekkenmann, Alexandra Simmons, and their colleagues at Routledge’s New York office for making it such a pleasant one. From the very start, their handling of the process has been open, swift, and professional. I hope it is always this easy. I am also hugely indebted to three anonymous reviewers of the penultimate manuscript, for their valuable and constructive feedback. They were directly responsible for a substantial rewriting of Part I of this book, and for countless additional changes throughout Part II. I have been unable to address all of their concerns, and no doubt the ways in which I have addressed others remain unsatisfactory, but I hope on the whole they will be pleased by the improvements. The last six years of my life have been eventful. While studying at Leeds, I met Esther Collado, who had come from Barcelona to take her doctorate in England. We have since got married and had a daughter, Sofía. Towards the end of my doctoral studies my mother was diagnosed with cancer, which killed her two years later, on the eve of what was to have been Esther’s and my wedding. I am lucky to be part of a very strong and loving family (enriched now by three marriages and five children, including my own), which has somehow been adapting to this loss of so vital an organ. Thanks would seem trite, and in any case beside the point. I love you all.

Conventions

Quotations from Hume’s texts throughout this book are from the electronic editions, edited by me and Peter Millican, available at https://davidhume.org. References are given in a standard format, following the conventions also adopted on that site: a code for the text, followed by a paragraph number, followed by a page reference to the most widely used modern reference edition. The main text codes are as follows: ‘T’ refers to the Treatise of Human Nature, ‘L’ refers to the Letter from a Gentleman, ‘E’ refers to the Enquiry concerning Human Understanding, ‘M’ refers to the Enquiry concerning the Principles of Morals, ‘D’ refers to the Dialogues concerning Natural Religion, ‘N’ refers to the Natural History of Religion, ‘P’ refers to the Dissertation on the Passions, ‘Tr’ refers to Of Tragedy, and ‘ST’ refers to Of the Standard of Taste. Other essays are abbreviated in a similar fashion, but only after they have been introduced by name, so that it will be obvious which essay is being cited; if in doubt, a full list is available at https://davidhume.org/notes/abbreviations. In addition, ‘HL 1’ and ‘HL 2’ refer, respectively, to volumes 1 and 2 of Greig’s edition of The Letters of David Hume, and ‘NHL’ refers to Kablinsky and Mossner’s New Letters of David Hume. Page references are to the following editions (page references for the Treatise are from Selby-Bigge’s edition; page references for the Letter from a Gentleman are from the Nortons’): Beauchamp, Tom L., ed. (2007). A Dissertation on the Passions and the Natural History of Religion. Oxford: Clarendon Press. Greig, J. Y. T., ed. (1932). The Letters of David Hume. Oxford: Oxford University Press. Kablinsky, Raymond, and E. C. Mossner, eds. (1954). New Letters of David Hume. Oxford: Oxford University Press. Kemp Smith, Norman, ed. (1747). Dialogues concerning Natural Religion. Edinburgh: Nelson. Miller, Eugene F., ed. (1987). Essays, Moral, Political, and Literary. Indianapolis: Liberty Fund. Norton, David Fate and Mary J., eds. (2007). A Treatise of Human Nature and A Letter from a Gentleman. Oxford: Clarendon Press.

xii Conventions Selby-Bigge, L. A., ed. (1975). Enquiries concerning Human Understanding and concerning the Principles of Morals. 3rd edition, revised by P. H. Nidditch. Oxford: Oxford University Press. Selby-Bigge, L. A., ed. (1978). A Treatise of Human Nature. 2nd edition, revised by P. H. Nidditch. Oxford: Oxford University Press.

Other primary sources are referred to in the ordinary fashion, with references given in the bibliographies at the end of each chapter. Unless otherwise stated, quotations and page references are from the original editions, taken from the scans available at Eighteenth Century Collections Online; where they are not, details are given in the relevant bibliographic entry. As a rule, I give page references to a modern edition, where one is easily available. I also quote from these editions, unless the editor has modernized the text to a significant degree, in which case my preference is to quote from the original. I cite these works using the original publication year, since I consider it desirable to be able to see at a glance when the thing in question was first said. The works often went through several editions in their authors’ lifetimes, receiving additions and revisions in the process. In such cases I use the date of the earliest edition in which the quoted passage appears. This has the effect that the same work is sometimes referred to with a different year. Butler’s Sermons, for example, are typically referred to as ‘Butler 1726’; but when I am quoting from the preface, which was first added in the second edition, I use ‘Butler 1729’. A glance at the bibliography will quickly reveal that these refer to different editions of the same text. In line with the style guide for the journal Hume Studies, the three main divisions of Hume’s Treatise are referred to as ‘Books’ (with a capital ‘B’), while all other divisions in all other works are referred to in lowercase (e.g. ‘part’, ‘chapter’, ‘section’). For ease of distinguishing internal from external references, however, the ‘Parts’, ‘Chapters’, and ‘Appendices’ of this book are referred to with an initial capital letter, while its sections are referred to using the ‘§’ symbol.

Introduction

In 1739 and 1740, David Hume published the three Books of his Treatise of Human Nature, the first on the understanding, the second on the passions, and the third on morals. Although it is now widely considered a philosophical masterpiece, and is the primary source of Hume’s present reputation, it was not well received by his contemporaries, and Hume himself was very dissatisfied with it. Having had greater success with a subsequent set of Essays, Moral and Political (1741–42), Hume later returned to the subject matter of Books 1 and 3 of the Treatise, publishing the first edition of his Enquiry concerning Human Understanding in 1748, and the first edition of his Enquiry concerning the Principles of Morals in 1751.1 This naturally prompts the question of what happened to Book 2. Where, we might ask, is Hume’s Enquiry concerning the Passions? Hume never published anything with this title, but the standard view is that the work standing in the appropriate relations is his Dissertation on the Passions, first published in 1757 (though probably drafted some years earlier, around the same time as the moral Enquiry). The analogy was made by L. A. Selby-Bigge in the introduction to his edition of the two Enquiries (1975, p. viii), is repeated by Tom Beauchamp in the introduction to his edition of the Dissertation and the Natural History of Religion (2007, p. xiv), and to the best of my knowledge has never been contradicted. Now, if Books 1 and 3 of the Treatise have generally overshadowed their Enquiry successors, Book 2 has positively eclipsed the Dissertation on the Passions. Nor is it hard to see why: while the two Enquiries immediately strike the reader as quite substantial reworkings of their Treatise predecessors, containing much that is different and new, the Dissertation can easily look like nothing more than, in Selby-Bigge’s words, ‘verbatim extracts from Bk. II of the Treatise, with some trifling verbal alterations’ (1975, pp. xx–xxi). Beauchamp’s computer collations of these works seem to corroborate these intuitive reactions, showing that there is no significant correlation between the two Enquiries and their corresponding Books of the Treatise, but that the Dissertation contains only around 2,850 new words. Beauchamp concludes that the latter ‘was essentially

2 Introduction created from the text of the Treatise’ (2007, p. li). And so, while Hume’s mature investigations of the understanding and of morals have recently emerged from the shadow of the Treatise as subjects worthy of independent study,2 Hume’s mature philosophy of emotion continues to languish in the dark.3 This book is an attempt to bring it into the light. Part of my case for thinking that Hume’s mature philosophy of the passions is worthy of study in its own right is that Selby-Bigge was wrong about the Dissertation. It is true that a lot of this later work is accurately described as Selby-Bigge describes the whole. Nevertheless, it is my view that not all of the alterations Hume made to the text of Treatise Book 2 are trifling or merely verbal. For example, in Chapter  4 I  will argue that a change in Hume’s definition of the terms ‘good’ and ‘evil’ entails that the very same words in the Treatise and the Dissertation convey completely opposing views about motivation. These are the sorts of points that a cursory reading or an unthinking computer collation cannot reveal.4 More than this, however, I will also be urging that the initial premise—­concerning the textual source of Hume’s mature thought on this topic—is mistaken. The true successor to Book 2 of the Treatise is not the truncated Dissertation on the Passions, but rather the complete set of Four Dissertations in which this work first appeared, including also the Natural History of Religion, Of Tragedy, and Of the Standard of Taste. In a soundbite: the Four Dissertations are Hume’s Enquiry concerning the Passions. The Four Dissertations is the only one of Hume’s substantial philosophical publications yet to receive a book-length treatment. Hume scholars have found things to say about the Natural History, but typically in the context of Hume’s philosophy of religion, and viewing the work primarily as a companion piece to the posthumous Dialogues concerning Natural Religion. The dissertations Of Tragedy and Of the Standard of Taste, meanwhile, have been studied mainly by philosophers of art, or by Hume scholars seeking parallels or contrasts with Hume’s moral philosophy. None of these works has yet been examined in its original publication context, or in relation to Hume’s philosophy of emotion.5 And the Dissertation on the Passions itself, as already noted, has scarcely been examined at all. It is my belief—or at least my hope—that by looking at these works together we can shed new light on each of them individually, and on Hume’s views on art, emotion, and superstition as a whole, while also adding to our understanding of the development of Hume’s thought over time. For instance, Hume’s novel solution to the paradox of tragedy, and his corresponding rejection of the previous solutions given by Du Bos and Fontenelle, are both well known. It is less well known, however, that in the Treatise Hume had endorsed a hybrid of the two Frenchmen’s accounts. Furthermore, Hume’s later criticism of Du Bos is disappointingly weak, depending on a straightforward misrepresentation of Du

Introduction  3 Bos’s position. I strongly suspect, therefore, that there is more to Hume’s rejection than meets the eye, and that the context of Hume’s philosophy of emotion provides the key to a fuller understanding. See Chapter 9. The book is divided into two parts. In Part I, I  motivate my project and set Hume’s views on the passions—and the Four Dissertations ­themselves—­in their textual and historical context. After a general introductory chapter, I proceed in Chapter 2 to examine some of Hume’s most important influences. In the next two chapters I then tell the story of what I take to be the major change in Hume’s philosophy of emotion. When he wrote the Treatise, I argue, Hume was a psychological hedonist and egoist, in the tradition of Hobbes, Locke, and Mandeville (­Chapter 3). Shortly afterwards, however, he read and was persuaded by Butler’s antiegoist arguments, and consequently became one of the clearest and keenest opponents of his own earlier view (Chapter 4). My thinking on these matters goes against the grain of what has perhaps been the dominant view since Norman Kemp Smith’s influential book (1941), according to which Hume’s later theory of the passions (and of morals) was already fully in place when he wrote the Treatise, and owed more to Hutcheson than either Mandeville or Butler. It is true that Hume’s famous anti-rationalism borrowed heavily from Hutcheson, and is a more or less consistent feature of his earlier and later philosophy. But this is half the story at best. On the whole, it seems to me that the significance of Hutcheson’s influence has often been overstated, while the importance of Mandeville is correspondingly overlooked. Even after his conversion to Butler’s anti-egoist motivational psychology—itself very different from Hutcheson’s—Hume remained true to many of his Mandevillean roots. While my focus is on Hume’s philosophy of emotion, this argument should be of interest to students of Hume’s moral philosophy as well, and indeed significant parts of the evidence for my case are drawn from the moral Enquiry. (Moral philosophers may also be interested in the discussion of reason and passion in Chapter 8.) I have little doubt that Hume’s reading of Butler left its mark on his views on morals as well as motivation. For instance, this development might help to explain why sympathy—which was so essential to Hume’s theories in both Books 2 and 3 of the Treatise—is so strikingly side-lined in the moral Enquiry (see Chapter 3, note 2). Perhaps needless to say, however, I cannot possibly do justice to the many difficult and controversial issues in this area, and for want of space I will have to restrict myself to matters directly relevant to Book 2 and the Four Dissertations. If I have the good fortune to meet with success, I shall proceed to the examination of morals. In Part II, I  embark on my critical study of the Four Dissertations proper, looking at the major topics of this work in roughly the order in which they appear (skipping over the aspects that came up already in the discussion of Part I). Broadly speaking, I devote two chapters in turn

4 Introduction to each of the dissertations, and thus these chapters are intended to be read in pairs. As I go, however, I will be drawing particular attention to the links between these four texts, and consequently none of these pairs is confined to discussing just one of the dissertations. In particular, my discussion of the central Dissertation on the Passions spills over on both sides, into the preceding examination of the Natural History and the subsequent treatment of the essay Of Tragedy. The importance of my central claim notwithstanding, I take it that the Four Dissertations are of general interest in their own right, even when the views they present do not differ substantially from those put forward in the Treatise, and even when the points being made do not resonate particularly strongly with others in the vicinity. The conception of this work as Hume’s Enquiry concerning the Passions is at the heart of this study, and provides its chief motivation, but the study itself does not end there. My aim, in sum, is to build up a coherent and—as far as possible within the constraints of space—a complete interpretation of the Four Dissertations, with an emphasis on how Hume’s ideas developed after the Treatise.

Notes 1 This is a slight (but harmless) simplification. The necessary qualifications are (i) that the Enquiry concerning Human Understanding—or ‘first’ Enquiry, as it is usually called for short—was initially published under the title Philosophical Essays concerning Human Understanding, and renamed only after publication of the ‘moral’ or ‘second’ Enquiry (second in time, that is, but not in name); and (ii) that section 8 of the first Enquiry, Of Liberty and Necessity, has its origins in Book 2 of the Treatise (T 2.3.1–2, pp. 399–412), rather than Book 1. 2 See Peter Millican on the first Enquiry (2002a, 2002b, 2007, 2016), and Jacqueline Taylor on the moral Enquiry (2002, 2009, 2015, forthcoming). 3 For example, of the 198 articles published in Hume Studies so far this century (volumes 26–41), 111 refer to Book 2 of the Treatise, while only 12 refer to the Dissertation on the Passions. Of these 12, most mention the Dissertation only once or twice in passing, often in a footnote; the only two giving it any significant attention are my own article (Merivale 2009) and a more recent one by Allison McIntyre (2014). 4 Beauchamp’s collation appears to be the direct result of a computational comparison. While his algorithm correctly identifies the source of most of the Dissertation’s paragraphs, in one case it is mistaken, selecting T 2.1.6.3 (p. 291) as the ancestor of P 2.41 (p. 15) on the basis of a few common words. A human reader could tell you, however, that this passage in the Dissertation has its origin three paragraphs later, at T 2.1.6.6 (p. 292); though there are fewer words in common, the point here is the same. See Appendix 2, in which I present the results of my own independent collation, which was done with the aid of a computer, but not simply by a computer. 5 A notable exception to this rule is Jane McIntyre (1999), who finds insightful things to say about the Natural History of Religion by examining it in relation to Hume’s treatment of the passions. I will echo some of her thoughts in Chapter 6.

Introduction  5

Bibliography Beauchamp, Tom L., ed. (2007). A Dissertation on the Passions and the Natural History of Religion, by David Hume. Oxford: Clarendon Press. Kemp Smith, Norman (1941). The Philosophy of David Hume. Basingstoke: Palgrave Macmillan. McIntyre, Alison (2014). ‘Fruitless Remorses: Hume’s Critique of the Penitential Project of The Whole Duty of Man’. Hume Studies 40(2), pp. 143–67. McIntyre, Jane (1999). ‘Passion and Artifice in Hume’s Account of Superstition’. In Religion and Hume’s Legacy. Eds. by D. Z. Phillips and Timothy Tessin. London: Palgrave Macmillan, pp. 171–84. Merivale, Amyas (2009). ‘Hume’s Mature Account of the Indirect Passions’. Hume Studies 35(1&2), pp. 185–210. Millican, Peter, ed. (2002a). Reading Hume on Human Understanding. Oxford: Oxford University Press. ——— (2002b). ‘The Context, Aims, and Structure of Hume’s First Enquiry’. In Reading Hume on Human Understanding. Ed. by Peter Millican. Oxford: Oxford University Press, pp. 27–65. ——— (2007). ‘Hume’s Old and New: Four Fashionable Falsehoods, and One Unfashionable Truth’. Proceedings of the Aristotelian Society, Supplementary Volumes 81, pp. 163–99. ——— (2016). ‘Hume’s Chief Argument’. In The Oxford Handbook of Hume. Ed. by Paul Russell. Oxford: Oxford University Press, pp. 82–108. Selby-Bigge, L. A., ed. (1975). Enquiries concerning Human Understanding and concerning the Principles of Morals. 3rd edition, revised by P. H. Nidditch. Oxford: Oxford University Press. Taylor, Jacqueline (2002). ‘Hume on the Standard of Virtue’. The Journal of Ethics 6(1), pp. 43–62. ——— (2009). ‘Hume’s Later Moral Philosophy’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 311–40. ——— (2015). Reflecting Subjects: Passion, Sympathy and Society in Hume’s Philosophy. Oxford: Oxford University Press. ———, ed. (forthcoming). Reading Hume on the Principles of Morals. Oxford: Oxford University Press.

Part I

1 My Design in the Present Work 

In this chapter I pursue three related introductory themes. First, I offer some textual orientation, in the form of a précis of Treatise Book 2, together with an indication of where its contents are picked up in Hume’s later work (§1.1).1 Next, I proceed with some additional textual orientation for the Four Dissertations themselves, both external and internal, outlining first their publication history (§1.2), and then offering a provisional argument for viewing them as a deliberately unified set (§1.3). Finally, I  introduce the main contention of this first part of the book, namely that between writing Book 2 of the Treatise and the Four Dissertations Hume changed sides on the highly contentious question of whether all motivation was reducible to self-love. At the same time, I  present what I take to be a helpful—albeit necessarily simplistic—framework for viewing the relevant early modern debate (§1.4), and state where I take Hume to fit into this picture (§1.5). The outlines of §1.4 will be filled in with some more detail in Chapter 2, and those of §1.5 in Chapter 3 and Chapter 4.

1.1.  Division of the Subject Hume begins Book 2 of the Treatise with four distinctions. First, he recalls his Book 1 distinctions of perceptions into impressions and ideas, and of impressions into those of sensation and those of reflection, now rebranded as original and secondary impressions respectively (T 2.1.1.1, p.  275). This distinction has a clear Lockean ring to it, being an echo of the Englishman’s distinction between ideas drawn in the same terms (Locke 1690, p. 105). For Locke, however, reflective ideas are those that arise in us from ‘the perception of the operations of our own mind’, and are of such things as ‘perception, thinking, doubting, believing, reasoning, knowing, willing’ (ibid.). For Hume, in contrast, impressions of reflection are those that arise secondarily, via ideas of original impressions (T 1.1.2.1, pp. 7–8), and comprise our ‘passions, desires, and emotions’ (ibid.), or equivalently ‘the passions, and other emotions resembling them’ (T 2.1.1.3, p. 275).

10  Part I Hume seems to have had a curious blind-spot with regard to Lockean reflection as a source of ideas.2 Without ever denying it explicitly, he implicitly rules it out with his copy principle—that all our simple ideas are copied from a precedent impression (T 1.1.1.7, p. 4; E 2.5, p. 19)— and also fails to appeal to it at three crucial points in his work where one might have thought it was the obvious thing to appeal to: necessary connection, belief, and volition. In each of these three cases he instead posits an impression as the source of our ideas.3 It is very unclear (to me at least) whether these impressions should count as original or secondary. There is no mention of any of them in either T 1.1.2 or T 2.1.1, and I can only assume that Hume wasn’t thinking of them when he drew that distinction. It seems most natural to me to treat each of them as sui generis, belonging to neither category.4 Be that as it may, Hume’s next move in Book 2 is to divide the reflective or secondary impressions (a category that apparently does not include the three anomalous impressions just mentioned) into the calm and the violent: Of the first kind is the sense of beauty and deformity in action, composition, and external objects. Of the second are the passions of love and hatred, grief and joy, pride and humility. (T 2.1.1.3, p. 275) The point of this distinction appears to be to separate the passions proper—what we would today call ‘emotions’—from the moral and critical sentiments. (The moral sentiments, one assumes, are what Hume meant here by the sense of beauty and deformity in action.) These distinctions, between them, afford a rough and ready way of dividing up the substance of the three Books of the Treatise: Book 1 is concerned predominantly with ideas, Book 2 with the passions or violent emotions, and Book 3 with the calm emotions of moral approbation and disapprobation. The sense of beauty and deformity in composition would presumably have been the main topic of the originally projected book on criticism, promised in the advertisement to Books 1 and 2 (Ad1739, p. xii).5 Finally, Hume divides the passions or violent emotions into the direct and the indirect, offering a brief definition of the two categories, together with the main examples on either side: By direct passions I understand such as arise immediately from good or evil, from pain or pleasure. By indirect such as proceed from the same principles, but by the conjunction of other qualities. This distinction I cannot at present justify or explain any farther. I can only observe in general, that under the indirect passions I  comprehend pride, humility, ambition, vanity, love, hatred, envy, pity, malice, generosity, with their dependants. And under the direct passions, desire, aversion, grief, joy, hope, fear, despair and security. (T 2.1.1.4, pp. 276–7)

My Design in the Present Work  11 Hume seems to have had several ‘other qualities’ in mind in connection with the indirect passions, and it will be convenient to have a label for the set; I will call them the associative qualities, since most of them involve the principles of association in one way or another, and those that do not typically set up a brute association of their own. The three most important are the double relation of impressions and ideas (which accounts for pride, humility, love, and hatred), comparison (which explains ambition, envy, and malice), and sympathy (which gives rise to vanity, pity, and generosity). I will examine the double-relation theory in Chapter 7, and sympathy and comparison in Chapter 3. The distinction between the direct and the indirect passions, aside from its theoretical interest, also informs the structure of Book 2. Part 1, Of pride and humility, explains the first four indirect passions in Hume’s list, namely pride, humility, ambition, and vanity. Following the general introduction in section 1 (which draws the four distinctions just noted), pride and humility are discussed at length in sections 2–10. Hume’s primary aim in these sections is to establish his double-relation theory, as applied to these two passions. In the final paragraph of section  10, he briefly accounts for ambition using the principle of comparison. In section 11, he explains his associative account of sympathy, before applying it to the passion of vanity. Section  12 then concludes this part with a discussion of the pride and humility of animals. In part 2, Of love and hatred, Hume explains the remaining indirect passions in his original list, namely love, hatred, envy, pity, malice, generosity, and ‘their dependants’. These dependants—judging at any rate by the contents of part 2—are the love of relations, esteem for the rich, benevolence, anger, respect, contempt, and the amorous passion.6 Since part 1 defended the theory of the double relation at length, at least as applied to pride and humility, part 2 is comparatively brief in extending its application to love and hatred in sections 1 and 2. Section 3 then begins ‘the sequel of this part’ by ‘removing some difficulties, concerning particular causes of these passions’ (T 2.2.3.1, p. 347). In the remainder of the sequel, namely sections 4–11, Hume proceeds with his examination of the other passions mentioned a moment ago. As with part 1, section 12 then concludes part 2 with a discussion of the love and hatred of animals. Parts 1 and 2 of Treatise Book 2 became sections 2–4 of the Dissertation on the Passions. Broadly speaking, part 1 forms the basis of section 2, while part 2 forms the basis of sections 3 and 4. In section 2 of the Dissertation, however, Hume cleans up the presentation of the doublerelation theory by bringing love and hatred within its scope right at the outset, alongside pride and humility. Section 3 of the Dissertation is then able to focus on the main part of the ‘sequel’ of Book 2, part 2, namely the examination of the indirect passions related to love and hatred (the first part of this sequel, i.e. the removal of ‘some difficulties’, is dropped from the later work). Section 4 of the Dissertation derives from T 2.2.2 (Experiments to confirm this system), incorporating also the second half

12  Part I of T 2.2.8 (Of malice and envy), in which Hume illustrates the importance of association in the workings of comparison. Thus section  4 of the Dissertation is a collection of additional ‘experiments’ or arguments intended to support Hume’s associationist psychology. The subject matter of the two concluding sections of parts 1 and 2 of Book 2, namely the passions of animals, is not revisited after the Treatise. Part 3 of Treatise Book 2, finally, concerns the direct passions, together with the general aspects of Hume’s theory of motivation (in contrast, that is, with some of the particulars that inevitably came up in the first two parts). This final part was split up in Hume’s later work, feeding first into section 8 of the first Enquiry, and then later into sections 1, 5, and 6 of the Dissertation. For reasons that will become clear when we get there, the contents and subsequent fate of part 3 will be described more fully in §1.3.

1.2.  Dispersed Thro’ Different Parts of This Volume Before looking at Book 2, part 3 in more detail, it is convenient to jump ahead to the Four Dissertations. This work, it should immediately be noted, came into being only after a couple of false starts. Hume recounted the main particulars of this story in a letter to his publisher William Strahan, dated the 25th January 1772:7 I am told by a Friend, that Dr Millar said to him, there was a Bookseller in London, who had advertisd a new Book, containing, among other things, two of my suppress’d Essays. These I suppose are two Essays of mine, one on Suicide another on the Immortality of the Soul, which were printed by Andrew Millar about seventeen Years ago, and which from my abundant Prudence I suppress’d and woud not now wish to have revivd. I know not if you were acquainted with this Transaction. It was this: I intended to print four Dissertations, the natural History of Religion, on the Passions, on Tragedy, and on the metaphisical Principles of Geometry. I sent them up to Mr Millar, but before the last was printed, I happend to meet with Lord Stanhope, who was in this Country, and he convincd me, that either there was some Defect in the Argument or in its perspicuity; I forget which; and I wrote to Mr Millar, that I would not print that Essay; but upon his remonstrating that the other Essays woud not make a Volume, I sent him up these two, which I had never intended to have publishd. They were printed; but it was no sooner done than I repented; and Mr Millar and I agreed to suppress them at common Charges, and I wrote a new Essay on the Standard of Taste, to supply their place. (HL 2, pp. 252–3) Having thus brought these works together, at least in part for pragmatic reasons, Hume then quickly separated them in the next edition of his

My Design in the Present Work  13 Essays and Treatises on Several Subjects (1758). In this edition, Of Tragedy and Of the Standard of Taste were appended to part 1 of the Essays, Moral, Political, and Literary, while the Dissertation on the Passions and the Natural History of Religion were placed following the first and second Enquiries respectively, a rearrangement that persisted in all subsequent collections: ‘The four Dissertations lately published are dispersed thro’ different parts of this Volume’, as Hume wrote in the advertisement to that edition (Ad1758). From this short publication history, one might be tempted to infer that the Four Dissertations never really belonged together in Hume’s mind, and hence that my project is simply misguided from the start. This is apparently the inference drawn, albeit tentatively, by Beauchamp: Hume had . . . originally envisaged a different collection of essays, and it seems unlikely that he had a philosophical vision that integrally linked geometry, religion, and the passions. Any integrity he might have envisioned, at some point in its history of shifting contents, could have been a consequence of fortuitous circumstances. (2007, p. xxix, n. 61) The conclusion, however, does not follow. It is only Of the Standard of Taste that was a late addition, and nothing suggests that Hume didn’t initially conceive the other three as part of a coherent set. Nor is there anything to suggest that, faced with the need for a new fourth, Hume didn’t then compose this late addition precisely with the other three in mind.8 As we will begin to see in the next section, and then more fully in Part II of this book, there are many substantial links between Hume’s views on art, religion, and the passions. Perhaps surprisingly, these links even extend to the abandoned material on geometry, as I will explain in a moment. The original four dissertations were all written around the same time (probably during Hume’s enormously productive two-year period at Ninewells from 1749 to 1751, when he also completed the moral Enquiry, the Political Discourses, and a draft of his Dialogues concerning Natural Religion),9 and it is hard to believe that Hume was unaware of how his own ideas were related. It is also possible that the subject of the passions was intended to be the common theme even according to the original plan, which featured a concluding dissertation concerning ‘some Considerations previous to Geometry & Natural Philosophy’ (HL 1, p. 223), before this was replaced by Of the Standard of Taste. This work sadly does not survive, but presumably it contained reworked material from Treatise Book 1, part 2 (Of the ideas of space and time, pp. 26–68), a topic that hardly features in the first Enquiry, save for three paragraphs in the context of a discussion of scepticism about abstract reasoning (E 12.18–20, pp. 156–8). While Beauchamp’s tacit assumption that there is nothing to connect Hume’s

14  Part I thought on geometry to his philosophy of emotion might seem plausible at first glance, it is in fact demonstrably false: there are two substantial sections in Book 2 of the Treatise devoted to discussing the effects on the passions of distance and contiguity in space and time (T  2.3.7–8, pp. 427–38). In the Dissertation on the Passions, these two sections—more than 3,700 words between them—are reduced to just a single sentence, which merely states a trite observation: ‘What is distant, either in place or time, has not equal influence with what is near and contiguous’ (P 6.18, p. 29). It is possible, therefore, that the abandoned dissertation on geometry combined the metaphysical discussion of Book 1 with the psychological discussion of Book 2 that the Dissertation on the Passions itself almost entirely omits. If this had been the case, the original four would have hung together very nicely; while if Hume had not discussed the relationship between geometry and the passions in this last dissertation, it is difficult to imagine what else could have saved it from sticking out like a sore thumb. This point will be strengthened in the next section, when we see how clearly the first three dissertations are connected, and how naturally a discussion of geometry and the passions would have concluded the set. Indeed, if my conjecture is right, then the withdrawal of the original fourth dissertation could even help to explain why the book ultimately fell apart; without it, perhaps, Hume felt that there just wasn’t enough for a full volume on the passions, and that Of Tragedy and Of the Standard of Taste would work better as stand-alone essays. The view that the Four Dissertations form an integrated set has been put forward before, by John Immerwahr and Jacob Sider Jost (2013). They emphasize the importance of the passions as a unifying theme, and lament the fact that the collection has received so little scholarly attention as a coherent whole. This book is intended to fill precisely that gap. To be clear, however, my characterization of the Four Dissertations as Hume’s Enquiry concerning the Passions is intended first and foremost as a claim about the content of this work, not about its author’s intentions (concerning which we can only speculate). My position is that the dissertations are in fact unified by the subject matter of the passions, whether by accident or by design, and—more importantly—that they are profitably studied together on this basis. This last claim, of course, stands or falls with the book as a whole. The proof of the pudding, as they say, is in the eating. In order to whet the appetite, however, I will expand in the next section on Immerwahr’s argument for deliberate unity in the construction of this short-lived set.

1.3.  Of the Will and Direct Passions When we look at the third and final part of Treatise Book 2 (Of the will and direct passions), and how Hume edited and reordered this material

My Design in the Present Work  15 for the Dissertation on the Passions, we find very striking evidence that Hume himself initially conceived of the Four Dissertations as a unified set. The beginnings of this argument were made by Immerwahr (1994), but as we will see he seems to have understated the case. He also failed to bring the abandoned dissertation on geometry into the picture, though there is—in line with my speculation in the previous section—a very natural extension to be made in this direction. There are ten sections in part 3 of Book 2 of the Treatise. Sections 1 and 2 are on liberty and necessity. This being a particularly important philosophical question, then as now, Hume understandably revisited it as soon as possible, in section 8 of his first Enquiry, immediately following the section  7 discussion of necessary connection, to which it is closely related.10 Section  3, Of the influencing motives of the will, contains Hume’s famous argument concerning the combat of passion and reason, revisited in section 5 of the Dissertation (albeit with some substantial and striking cuts; see §8.5). It is the remaining seven sections that matter for the present argument. Having argued in T 2.3.3 that there can never be any combat between reason and passion, strictly speaking, Hume goes on to claim that the struggle his predecessors have confusingly described in these terms is really one between calm and violent desires (T  2.3.38–9, pp.  417–8). This later distinction is not to be confused with the earlier one between calm and violent emotions. The point of the earlier distinction, as I said in §1.1, was apparently just to remove the moral and critical sentiments from immediate consideration, the better to focus on the passions proper. This later distinction, though it is based on the same factor (i.e. the felt intensity of the impressions in question), applies on the violent side of the earlier line. To put the point as clearly as possible, Hume is here distinguishing calm violent emotions from violent violent emotions. Furthermore, while the earlier distinction applies (as we would now say) to types of emotions, the later distinction applies to their tokens. The very same desire can occur either calmly or violently in the later sense, in confirmation of which notice that Hume here gives resentment as an example on both sides (T 2.3.3.8–9, pp. 417–8). With the combat of reason and passion recast as the combat of calm and violent passions, Hume proceeds in sections 4–8 to the examination of ‘some of those circumstances and situations of objects, which render a passion either calm or violent’ (T 2.3.4.1, p. 419). Section 9 then turns to the direct passions in general, offering Hume’s official accounts of desire, aversion, joy, sorrow, hope, and fear (despair and security had already come up briefly in section 4). And in section 10, finally, Hume concludes the Book with his account of curiosity, or the love of truth. The first thing to note (and this is the point Immerwahr makes) is that Hume reorders the material from these seven sections, moving his general discussion of the direct passions—including, notably, the passions

16  Part I of hope and fear—from the penultimate section of Treatise Book 2 to the very first section of the Dissertation, leaving the Dissertation to end instead with his examination of the causes of the violent passions. This latter discussion, moreover, now focuses almost exclusively on Hume’s so-called conversion principle. The substance of this principle does not matter for now (I examine it in Chapters  9 and 10). What matters is simply to point out that it is the very principle Hume goes on to apply, in Of Tragedy, to the puzzling phenomenon of the pleasure that we take in tragic drama. In a similar way, hope and fear are the main emotions that connect the Dissertation on the Passions to the Natural History of Religion, since they are at the heart of Hume’s account of the origin of religious belief (see Chapter 6). Given these two clear links, there is an obvious explanation for Hume’s rearrangement, if we assume that he was thinking of the Four Dissertations as a unified set. For in this set, the Dissertation on the Passions is sandwiched in between the Natural History of Religion and Of Tragedy. There is thus a natural progression of ideas from the first of these dissertations to the second (via hope and fear), and from the second to the third (via the conversion principle). This reordering of material from Book 2, part 3, is not the only thing that points to deliberate unity. There are also some very telling facts regarding what Hume left out from his earlier work (which Immerwahr does not mention). As well as shifting the penultimate section of Book 2, which included the account of hope and fear, to the start of the Dissertation on the Passions, Hume also removed altogether the final section on curiosity, or the love of truth. I can find no evidence that Hume came in later life to doubt his account of this passion, and accordingly assume that he omitted it merely so as to keep things relatively short. But there is a particular reason why this section should have marked itself out as a suitable candidate for dropping, should length have been a concern, and this reason is also to be found in the Natural History. In the second section of this dissertation, Hume writes: It must necessarily, indeed, be allowed, that, in order to carry men’s attention beyond the present course of things, or lead them into any inference concerning invisible intelligent power, they must be actuated by some passion, which prompts their thought and reflection; some motive, which urges their first enquiry. But what passion shall we here have recourse to, for explaining an effect of such mighty consequence? Not speculative curiosity surely, or the pure love of truth. That motive is too refined for such gross apprehensions; and would lead men into enquiries concerning the frame of nature, a subject too large and comprehensive for their narrow capacities. No passions, therefore, can be supposed to work upon such barbarians, but the ordinary affections of human life; the anxious concern for happiness, the dread of future misery, the terror of death, the thirst of revenge,

My Design in the Present Work  17 the appetite for food and other necessaries. Agitated by hopes and fears of this nature, especially the latter, men scrutinize, with a trembling curiosity, the course of future causes, and examine the various and contrary events of human life. And in this disordered scene, with eyes still more disordered and astonished, they see the first obscure traces of divinity. (N 2.5, pp. 38–9; my emphases) Here, Hume not only places hope and fear at the centre of his account of the origin of religious belief, but also explicitly discounts the love of truth. It is understandable, then, that when he came to rework the relevant material in the Dissertation on the Passions he chose to begin with the former passions, and cut out the section on the latter altogether. A similar point can be made with regard to the connection between the second and third dissertations. Section 6 of the Dissertation on the Passions, containing Hume’s mature discussion of the causes of the violent passions, derives as I have said from sections 4–8 of Treatise Book 2, part 3. From this original, Hume reproduces almost all of section  4, which contains his extended treatment of the conversion principle. The single paragraph reproduced from section 5, meanwhile, is the one paragraph from that section that concerns this same principle (T 2.3.5.2, pp. 422– 3). Section 6 is condensed into five short paragraphs (P 6.13–17, p. 28), while sections 7 and 8, as noted in §1.2, are reduced to just a single sentence (P 6.18, p. 29). As with his account of curiosity, I can find no evidence that Hume was dissatisfied with the abandoned material from sections  5–8. But now, if the Natural History explains why curiosity was cut, and Of Tragedy explains why the conversion principle wasn’t, what explains the absence of the other material from sections  5–8, most notably the discussion of the influence of space and time on the passions? I  suggest that this material is absent from the Dissertation on the Passions because Hume had instead worked it into the abandoned dissertation on geometry. For notice how naturally it would have followed in the original context: section 5 of the Dissertation contains Hume’s argument concerning the combat of passion and reason; section 6 then begins his account of the causes of the violent passions, by explaining the conversion principle; Of Tragedy elaborates on this principle, applying it to the phenomenon of the pleasure that we take in tragic drama; the dissertation on geometry could then have rounded things off nicely with the rest of Hume’s account of the causes of the violent passions, namely the effects of distance and contiguity in space and time, either before or after addressing the metaphysical side of this topic.11 However plausible this last speculation may be, the known facts are these: Hume made significant cuts from Treatise Book 2, part 3, preserving in the Dissertation on the Passions only the material that was relevant

18  Part I to the Natural History of Religion and Of Tragedy, while at the same time moving this material so that it appeared immediately following and preceding these other two dissertations respectively. This strongly suggests that Hume originally conceived of the Four Dissertations as a unified set, with the passions as its central theme.

1.4.  Sects That Naturally Form Themselves in the World The thematic links among the Four Dissertations constitute only half of my case for viewing this work as Hume’s Enquiry concerning the Passions. The other half is that there are important and substantial differences between Hume’s earlier and later philosophy of emotion, making the Four Dissertations worthy of study independently of Treatise Book 2. In Part II of this book, I will point out various local differences in the course of reconstructing Hume’s mature position. In the remaining chapters of Part I, however, I will offer a sustained argument for one particularly important global difference, namely Hume’s change from egoism to anti-egoism in the theory of motivation. By way of introducing this argument, I turn now to some general historical orientation. The philosophers of the seventeenth and eighteenth centuries can be grouped and labelled in various ways, for various different purposes. Any set of labels we might choose will inevitably involve a degree of simplification, but as long as this general warning is borne in mind, the practice does not seem to me inherently dangerous. To the contrary, I  find that labels like these can serve as helpful landmarks when navigating the complex intellectual terrain. Moreover, when it comes to Hume, at least with regard to the passions and morals, there is an obvious framework to adopt, since it is one to be found explicitly in his own writings. In The Epicurean, the first of his four essays on happiness, Hume announces his intention to ‘deliver the sentiments of sects, that naturally form themselves in the world, and entertain different ideas of human life and of happiness’, giving to each ‘the name of the philosophical sect, to which it bears the greatest affinity’ (Ep n1, p. 138). The ancient schools chosen are Epicureanism, Stoicism, Platonism, and Scepticism. While nothing like these schools and their card-carrying representatives still existed in the eighteenth century, of course, Hume and his contemporaries were familiar with (and often referred to) the classical background to their own views and debates, and there is a legitimate sense in which we can think of these four traditions as still very much alive at that time.12 In Chapter 2, when introducing the British debate in the years leading up to Hume’s Treatise, I will identify Thomas Hobbes, John Locke, and Bernard Mandeville as early modern Epicureans, and Lord Shaftesbury, Francis Hutcheson, and Joseph Butler as eighteenth-century Stoics. I will also briefly mention some exemplars of the Platonic tradition here, notably Samuel Clarke; his views, together with those of the French Platonist

My Design in the Present Work  19 Nicolas Malebranche, will be discussed further in Chapter 8. René Descartes belongs in this camp as well, though I will not have occasion to examine his philosophy here. Prominent Sceptics of the period include Pierre Bayle (whose views are also not examined here), and Hume himself (see §1.5). The classical Epicureans were hedonists, who held that pleasure was our chief or sole good. In the early modern period, this manifested itself in the popular doctrine of (what we now call) psychological egoism, the view that all motivation is reducible to self-love, the desire for personal happiness. This doctrine was also invariably presented in hedonistic terms, with happiness understood as pleasure and the absence of pain. In matters of religion, meanwhile, the Epicureans were frequently branded as atheists, because of their materialism and their denial of providence and divine design. Hobbes was (at least perceived to be) an Epicurean in all these senses. Locke defies such easy classification, since his metaphysical and moral Epicureanism was combined with a more Platonic attitude to religion. The classical Stoics, by contrast, held that virtue was our main or only good. In the eighteenth century, this became the doctrine of sentimentalism or psychological altruism, defended vigorously in opposition to Hobbes and the other Epicureans. We have genuinely benevolent sentiments, according to this view, and happiness and virtue coincide in the exercising of them. The Stoics, both ancient and modern, also emphasized the beauty and order of the universe, and in religion were keen proponents of the design argument. Given their emphasis on order, however, they were generally hostile to the ideas of miracles and particular providence, and at the more freethinking end of the spectrum (e.g. Shaftesbury) were thus attracted to deism. At some level of abstraction, the Stoics and the Platonists of the early modern period belong together, united by the common cause of refuting what they took to be the irreligious and immoral principles of Epicurus and his followers. With Hutcheson, however, the difference became more pronounced, since unlike Shaftesbury or Butler he was explicitly opposed to moral rationalism. In understanding Hume, moreover, the difference between these two traditions becomes increasingly important. Not only did Hume follow Hutcheson in opposing moral rationalism, but he was also—notwithstanding his own Epicureanism in this regard—­considerably kinder to Stoical than to Platonic attitudes to religion, as we will see. Where the Epicureans placed pleasure at the heart of their moral philosophy, and the Stoics placed virtue, the early modern Platonists had instead an Augustinian God (corresponding to Plato’s form of the good). God was for them our sole good and the author of all our happiness. This tradition also had at its heart Descartes’s real distinction between mind and body, and a belief in the absolute superiority—both metaphysical and moral—of the former over the latter. With Malebranche, as we will

20  Part I see in §8.1, Cartesian philosophy received a further injection of Augustine’s neo-Platonism in the form of the doctrine of original sin, and an increased aversion to the body and its lustful passions, which turn us away from God and our one true good. The Platonists were rationalists, with a firm conviction in the powers of a priori reason both to establish the existence of God and to regulate our base and bodily passions in the service of our creator. Their favourite proofs for the existence of God were the ontological or the cosmological arguments, whereas the Stoics, as already noted, preferred the a posteriori design argument based on the observable order and beauty of the universe. Accordingly, the Platonists were not averse to the idea of miracles and of a particular providence interrupting the natural order of events; indeed, they often saw these as proofs of divine authority that paved the way to revelation. In terms of the distinction that Hume sets up in his essay Of Superstition and Enthusiasm, Platonism tends towards superstition, while Stoicism tends towards enthusiasm. As I will argue in §5.2, Hume was openly hostile to superstition, but rather less critical of enthusiasm. Hume’s general opposition to the Platonic tradition was with him from the start, and is a consistent and unambiguous feature of his earlier and later philosophy. What is less clear, however, is whether he leans more towards Epicureanism or more towards Stoicism in his positive views. There are many fairly obviously Epicurean aspects to his thought, such as the associationism he picked up from Hobbes and Locke, his determinism, and his rejection of providence and life after death. (His argument against a future state in the first Enquiry, section 11, is explicitly put in the mouth of Epicurus.) These points notwithstanding, however, there has been a strong tendency in the last hundred years or so to view Hume’s moral and motivational philosophy as closely in line with Hutcheson’s sentimentalism and moral sense theory. Three particularly prominent interpretations in this vein (which of course differ among themselves in the details) are from Norman Kemp Smith (1941), David Fate Norton (1982), and Don Garrett (2015). Kemp Smith’s narrative of Hume as Hutcheson’s protégé has been brought into question in recent years, by commentators such as James Moore (1995), James Harris (2009, 2015), and Mikko Tolonen (2013). While acknowledging a degree of similarity or even continuity between Hutcheson and Hume (notably their shared insistence that morality is founded on sentiment more than reason), these interpreters have emphasized the important Epicurean aspects of Hume’s moral philosophy— the artificiality of justice, for example, and the central importance of sympathy—­while reminding us that these represent fundamental points of disagreement with Hutcheson. Harris and Tolonen, moreover, have insisted on the considerable importance of Mandeville as a formative influence on Hume’s early views. There remains no consensus, and given

My Design in the Present Work  21 the very deep opposition between Hutcheson and Mandeville, different views on this matter will naturally coincide with very different views on Hume’s own moral theory.

1.5.  Sceptical Solution of These Doubts Before weighing in with my own angle on this debate, it will be convenient to complete the general framework begun in the previous section, and to give Hume himself a place within it. For if Hume is to be affiliated with one of the classical schools more than any other, the obvious identification is with neither the Epicureans nor the Stoics, but the Sceptics. Hume himself explicitly associates with this group, for example when he describes the philosophy of the Treatise as ‘very sceptical’ (A 27, p. 657; see also T 1.4.7.15, pp. 273–4, T App.21, p. 636), when he endorses a ‘mitigated scepticism’ in the conclusion of the first Enquiry (E  12.25– 6, pp.  162–3), or when he famously offers a ‘sceptical solution’ to his ‘sceptical doubts’ about induction (titles to E 4 and E 5, pp. 25, 40). He implicitly sides with the Sceptics in the aforementioned essays on happiness, by giving them the last and (by far) the longest word, and having them endorse some recognizably Humean ideas.13 And his sympathies are again explicit at the start of section 5 of the first Enquiry, where he singles out Scepticism as the only species of philosophy that is not liable to bias the mind by fostering some predominant inclination: The academics [or sceptics] always talk of doubt and suspense of judgment, of danger in hasty determinations, of confining to very narrow bounds the enquiries of the understanding, and of renouncing all speculations which lie not within the limits of common life and practice. Nothing, therefore, can be more contrary than such a philosophy to the supine indolence of the mind, its rash arrogance, its lofty pretensions, and its superstitious credulity. Every passion is mortified by it, except the love of truth; and that passion never is, nor can be carried to too high a degree. (E 5.1, p. 41) Further evidence of Hume’s Sceptical credentials comes from his fondness for the Roman Sceptic Cicero.14 The four essays on happiness themselves, indeed, must have been written with Cicero’s De Finibus Bonorum et Malorum in mind. In this work, an Epicurean, a Stoic, and a Peripatetic (i.e. an Aristotelian) converse and argue about happiness or the ultimate goal of life, while Cicero himself narrates and takes on the role of the Sceptic, criticizing all three. There are certainly differences here: Hume wrote four separate monologues, rather than a dialogue; Cicero’s Peripatetic is replaced by a Platonist in Hume’s set; and Hume has his Sceptic present substantial positive views—albeit in a suitably undogmatic

22  Part I spirit—where Cicero, on behalf of that school, plays a purely critical role in his discussion. But the structural parallels are nevertheless clear and surely deliberate.15 Hume’s Dialogues concerning Natural Religion also has a Ciceronian model, namely De Natura Deorum, a dialogue in which an Epicurean, a Stoic, and a Sceptic debate the nature of the gods. Here the Sceptical side of the argument is held up by Cotta, with Cicero himself represented as an impartial observer; though Velleius (the Epicurean) remarks early on that Cotta and Cicero are both disciples of Philo, and have learned from him to be sure of nothing (Brookes 1896, p. 20). Philo was a Sceptic, and the last head of Plato’s Academy before its destruction in the first century BC.16 He is of course the namesake of the Sceptic in Hume’s Dialogues, and while it is controversial whether Hume agrees with everything this character says, there is no doubt that this is where his sympathies lie, at least broadly speaking. This is further confirmation of his close ties to this group. The other two characters in Hume’s Dialogues, Cleanthes and Demea, represent the Stoical and Platonist traditions respectively. Cleanthes is named after Zeno’s successor as head of the Stoic school in Athens. ­According to Mossner (1936), the early modern Stoic Butler was the model for Hume’s character. This seems plausible, as long as we understand ‘model’ suitably loosely; Cleanthes does not represent Butler’s views exactly, but there are many obvious—and obviously intentional—­ similarities, and more so than with any other prominent philosopher of the time. There is no obvious real-life candidate for Demea to be named after. Dorothy Coleman suggests etymology as a likelier inspiration for this name: ‘Demea’ comes from the Greek ‘demos’, meaning people, and Hume’s character represents a popular view of religion at the time (Coleman 2007, p. xi). For the modern counterpart, meanwhile, Mossner suggests Clarke, who certainly seems as good a candidate as any; though Malebranche strikes me as a better model for the Platonist in the essays on happiness, and it may be noted that Demea also quotes Malebranche in support of his mysticism (D 2.2, pp. 141–2). In current Hume scholarship, the question of whether or in what sense Hume was a Sceptic—notably with regard to his famous argument about induction—is a hot topic, with several people now favouring descriptive rather than normative readings (Garrett 1997, Noonan 1999, Owen 1999, Beebee 2006). As it happens, I favour a normative reading myself (in line with Millican 2007b, 2012), but my identification of Hume with the Sceptics here is not intended to foreclose that debate. It is perfectly consistent to suppose that Hume was a Sceptic in the present sense—i.e. an inheritor of the Sceptical tradition of Philo and his pupil Cicero— while denying that he was a Sceptic about induction in the now controversial sense. And surely it will not be denied, even by defenders of non-Sceptical readings, that Hume’s views on induction are to be situated

My Design in the Present Work  23 within the Sceptical tradition, for Hume himself explicitly presents them in this way, calling his account a ‘sceptical’ solution to ‘sceptical’ doubts, as already noted. Indeed, calling Hume a Sceptic in the present sense entails very little about his positive views, for it is the prerogative of the Sceptics to pick and choose doctrines from any of the other schools, depending on whoever seems to have the better of that particular argument. This is exactly what we find Hume doing. Though he was generally hostile to Platonism, as I have said, his positive views, where they have clear antecedents, are drawn from both the Epicurean and the Stoical traditions, as well as more directly from the Sceptics.17 The present identification therefore gives us a convenient label for Hume’s general outlook, without thereby shutting off the debate between the Hutchesonian and Mandevillean interpretations introduced at the end of the previous section. There is more driving my identification of Hume with the Sceptics than just the eclectic nature of his views, however, and this additional motivation provides a hint as to how this debate might profitably be taken forward. For it is also the prerogative of the Sceptics, crucially, to change their minds. Kemp Smith’s evidence for his Stoical interpretation of Hume drew predominantly on the moral Enquiry, and he read (or re-read) the earlier Treatise through this lens. Moore’s case for his Epicurean reading, in contrast, was based almost exclusively on Hume’s first publication. This suggests an interesting middle-ground hypothesis: that Hume started his career as more of an Epicurean Sceptic than a Stoical one, but that after the Treatise this balance shifted. This, in very broad outline, is precisely what I believe happened. In the case of the Treatise, then, my sympathies lie with Moore, Harris, and Tolonen. The differences between Hutcheson and Hume in this work strike me as much more substantial and important than the similarities. Harris and Tolonen are also right, I believe, in insisting on Mandeville (in Hutcheson’s place) as one of the most important positive influences on the early Hume. This last claim can initially sound quite implausible, given Mandeville’s reputation as perhaps the most full-blooded and uncompromising egoist of the period, and Hume’s reputation as one of the clearest and ablest opponents of this view. Tolonen attempts to resolve this dilemma by arguing that Mandeville, though undoubtedly an egoist in his early work, abandoned this view somewhere in the mid-1720s (Tolonen 2013, pp. 73–4); and it is Mandeville’s mature (and anti-egoist) philosophy, both he and Harris maintain, that was most significant in the formation of Hume’s early ideas. I do not find Tolonen’s case for the anti-egoist reading of Mandeville’s later work entirely convincing (see §2.3). However that may be, I resolve the immediate dilemma here in precisely the opposite direction. Though Hume was undeniably an anti-egoist when he wrote the moral Enquiry, I believe this was for him a later development; as I argue in Chapter 3,

24  Part I the Treatise itself defends a psychological egoism very much in line with Locke and Mandeville (or at least the early Mandeville). For me, then, the important change of mind in this context was Hume’s, rather than Mandeville’s. The anti-egoist interpretation of the Treatise comes to us from the Hutchesonian tradition of Hume scholarship. In accepting it, it seems to me, Harris and Tolonen are conceding too much to their opponents. In ostensibly reconstructing Hume’s view in the Treatise, Kemp Smith in fact quoted liberally from Hume’s moral Enquiry, without considering the possibility that Hume’s views were different when he wrote this later work. Admittedly there is some evidence of anti-egoism in the Treatise itself, and I do not claim that my interpretation of this work is simply obvious once the mere possibility of a change of mind has been raised; there is a controversial argument here to be won. But if I am right, then the seemingly wide gulf between Mandeville and (the early) Hume may be narrowed considerably, while that between Hume and Hutcheson widens accordingly. The last few paragraphs might suggest that, having sided more or less with Harris and Tolonen about the Treatise, I am now going to concur with Kemp Smith and Garrett about Hume’s later work. Here too, however, I continue to see Hutcheson’s influence as comparatively small. Both parties to the present debate, I suggest, have overlooked the considerable importance of Joseph Butler in the development of Hume’s thought. It is Butler, not Hutcheson, who deserves the credit for having persuaded Hume of the falsehood of psychological egoism. Thus I propose an important addition to the story Harris tells in chapter 1 of his intellectual biography of Hume. Having described Hume’s discovery, in the 1720s, of Shaftesbury ‘as an antidote to university’ (Harris 2015, pp. 38–51), he moves on to his engagement, in the 1730s, with Mandeville ‘as an antidote to Shaftesbury’ (2015, pp. 51–64). What is missing is the next episode in this sequence, namely Hume’s encounter with Butler as an antidote to Mandeville. Having introduced the relevant participants of the early modern egoism debate in Chapter 2, and developed my egoist interpretation of the Treatise in Chapter 3, I will conclude Part I of this book with a description, in Chapter 4, of this previously underexplored episode.

Notes   1 The material here presented discursively is summarized in tabular form in the two appendices. In Appendix 1, I lay out my reconstruction of Hume’s Treatise taxonomy, together with definitions of all his various terms for particular passions (‘pride’, ‘vanity’, ‘love’, ‘benevolence’, ‘anger’, etc.). In Appendix 2, I list all the sections and paragraphs of the Dissertation on the Passions alongside their likely ancestors from Treatise Book 2.   2 I am grateful to Peter Millican for drawing my attention to this point.   3 See, respectively, T 1.3.14.22 (pp. 165–6) and E 7.28 (pp. 75–6); T App.2–3 (pp.  623–5) and E  5.10–3 (pp.  47–50); and T  2.3.1.2 (p.  399), T  2.3.9.2 (p. 438), and T 2.3.9.4 (p. 439).

My Design in the Present Work  25   4 For a discussion of Hume on belief, see Broackes (2002). On the idea of necessary connection, see Millican (2007a), who offers a Lockean reconstruction of Hume’s account: ‘it is our reflexive awareness of making the inference that leads us to the very idea of connexion’ (p. 224; see also p. 249, n. 26). Regarding the impression of the will or volition, it is notable—and perhaps significant—­that it makes no reappearance anywhere in Hume’s later writing. See Millican: ‘[Hume’s] equation of the will with an internal impression, no doubt motivated by his Copy Principle, seems to be a slip, as it leaves no obvious mark on his treatment of the will elsewhere. A more charitable reading would be that Hume intends “the will” to refer to our faculty of knowingly—and “willingly”—giving rise to actions (of the mind and body), a faculty of which we become aware, and whose idea we thus acquire, through a corresponding internal impression’ (2009, p. 6).   5 To say that the moral sentiments are not passions proper does not mean— contra Loeb (1977)—that Hume’s theory of the passions has no bearing on his theory of morals. Hume himself describes the distinction between calm and violent emotions as ‘vulgar and specious’ (T 2.1.1.3, p. 276), adopting it merely for exegetical convenience. The similarities between the moral sentiments and the passions of pride, humility, love, and hatred are obvious and striking, and Hume himself draws the parallel explicitly towards the end of Book 3 (T 3.3.5.1, p. 614). On the relationship between the moral sentiments and these four passions, see e.g. Árdal (1966), Cohon (2008, 2010).   6 In one place Hume classifies benevolence and anger as direct passions (T 2.3.9.8, p. 439). For independent reasons, I believe this paragraph should not be taken as representative of Hume’s view in the Treatise; see §4.1. That paragraph aside, it seems to me more natural to view these passions as indirect in the Treatise (alongside love, hatred, pity, and malice, to which they are importantly related). To complicate matters further, Samuel Rickless has argued—­with some plausibility—that pity and malice should also count as direct passions, and that Hume’s classification of them as indirect should be treated as a mistake (Rickless 2013). I do not agree, but the issue has no bearing on my argument here. For present purposes, my reconstruction of the categories of the direct and indirect passions can be treated simply as an innocuous stipulation.   7 For an interesting (if somewhat speculative) reconstruction of the full story that Hume here alludes to, see E. C. Mossner (1950).   8 The same cannot be said of Hume’s first choices for filling the gap, Of Suicide and Of the Immortality of the Soul. These two suppressed essays were written some time earlier (perhaps around the same time as the Treatise), and were presumably offered for inclusion simply because something was needed, and they were to hand. Their irreligious content chimes with the Natural History in that one very general respect, and the defence of suicide would have struck a note of accord with the dedication that was ultimately prefixed to (some copies of) the Four Dissertations, which praised John Home’s tragedy Douglas, itself highly controversial for its sympathetic or even positive portrayal of the suicide of its main character. Home’s play did not premiere until 1756, however, after Hume had thought of including the scandalous essay in the set. In all, the specially written essay on taste is—or so I will argue—a better fit with the other three.   9 See Beauchamp (2007, pp. xx–ii) and Harris (2015, pp. 248–50, 289). Hume wrote to Andrew Millar on the 12th June 1755, asking him if he would like to publish ‘four short Dissertations, which I  have kept some years by me, in order to polish them as much as possible’ (HL 1, p. 223). These were the original four, and the letter was sent before Stanhope persuaded Hume to withdraw the dissertation on geometry. Of the Standard of Taste was therefore probably written the following year.

26  Part I 10 The link between Hume’s treatments of causation and of free-will is surely obvious. It has been insisted on particularly forcefully by Millican (2007a, 2007b, 2009, 2011), in opposition to the sceptical realist interpretation of Hume on causation (Millican argues that Hume’s defence of compatibalism depends crucially on a ‘thin’ account of necessity). 11 In the light of this conjecture, it is fair to ask how long the abandoned dissertation on geometry might have been. In the 1755 letter to Andrew Millar (mentioned in note 9 earlier), Hume wrote that the original four dissertations ‘wou’d make a Volume a fourth less than my Enquiry; as nearly as I can calculate’ (HL 1, p. 223). This must have been an underestimate, however. The ‘Enquiry’ in question was the moral Enquiry (at that time, the first Enquiry was still called Philosophical Essays). Its most recent edition (1753) was 275 pages long. The first three dissertations alone, as published in 1757, came to 209 pages in the same type, by themselves already more than three quarters the size of the moral Enquiry. A better way to estimate the original fourth dissertation’s size, I therefore suggest, is by comparison with what replaced it: first the two essays Of Suicide and Of the Immortality of the Soul (6,620 words between them), and later Of the Standard of Taste (7,759 words). Splitting the difference, we may suppose the dissertation on geometry was around 7,000 words. If this is right, then the original volume would still have been shorter than the moral Enquiry, explaining Hume’s estimate, but not by as much as a fourth (more like a sixth). Now, Treatise Book 1, part 2, is more than 15,000 words in total, so Hume evidently didn’t intend to revisit the whole thing. If the Book 2 material on geometry and the passions was cut to half the size, and included in the dissertation as I suggest, that would still have left Hume plenty of room to present his main arguments from T 1.2.1–2, and to answer a selection of the objections considered in T 1.2.4–5. 12 For a recent study of Epicureanism in the early modern period, see Wilson (2008); for Stoicism, see Brooke (2012); for Platonism, see Hedley and Hutton (2008); and for Scepticism, see Popkin (2003). 13 Perhaps the most notable such idea is the sceptical principle that I will discuss in Chapter 11, that ‘there is nothing, in itself, valuable or despicable, desirable or hateful, beautiful or deformed; but that these attributes arise from the particular constitution and fabric of human sentiment and affection’ (Sc  8, p. 162). 14 In his autobiographical My Own Life, Hume admits that it was Cicero he was secretly reading when his family thought he was studying to become a lawyer (MOL  3, p. xxxiii). For more on Hume’s Ciceronian influences, see Jones (1982). 15 The replacement of Cicero’s Peripatetic with a Platonist in Hume’s set calls for an explanation. Descartes and the other instigators of early modern philosophy were motivated in large part by a rejection of Aristotle and the Aristotelianism that had dominated the preceding centuries. By the time Hume came on the scene, therefore, the Peripatetic philosophy was essentially dead (certainly as far as Hume was concerned; see E 1.4, p. 7), while the Platonic tradition—thanks to Descartes and his followers—was still very much alive. 16 For the second half of its existence, the Academy had been run predominantly by Sceptics, hence the synonymy of ‘academical’ and ‘sceptical’ philosophy, as in, for example, the title of Hume’s first Enquiry, section 12. 17 In fact, Hume was not averse to drawing on the Platonists either, at least on matters not directly pertaining to religion or morals. George Berkeley was a positive influence in some respects, and he arguably counts as a Platonist (albeit a rather idiosyncratic one). Malebranche was a paradigmatic Platonist,

My Design in the Present Work  27 meanwhile, and many have seen him as having exerted an important positive influence. Peter Kail, for example, sees Hume as ‘using Malebranchean materials to arrive at conclusions that are squarely antithetical to those fundamental to the religious dimension of Malebranche’s philosophy’ (2008, p. 55). See also Wright (1983), Jones (1982), Lennon (1997), and Kail (2005, 2007a). On at least one point, it seems to me that the similarities between these two thinkers have been overstated (see §10.2). But it is likely that Malebranche’s ideas fed directly into Hume’s thoughts on sympathy and comparison; see James (2005) and Schmitter (2012).

Bibliography Árdal, Páll S. (1966). Passion and Value in Hume’s Treatise. Edinburgh: Edinburgh University Press. Beauchamp, Tom L., ed. (2007). A Dissertation on the Passions and the Natural History of Religion. Oxford: Clarendon Press. Beebee, Helen (2006). Hume on Causation. London: Routledge. Broackes, Justin (2002). ‘Hume, Belief, and Personal Identity’. In Reading Hume on Human Understanding. Ed. by Peter Millican. Oxford: Oxford University Press, pp. 187–210. Brooke, Christopher (2012). Philosophic Pride: Stoicism and Political Thought from Lipsius to Rousseau. Princeton: Princeton University Press. Brookes, Francis, ed. (1896). De Natura Deorum, by Cicero. London: Methuen & Co. Cohon, Rachel (2008). ‘Hume’s Indirect Passions’. In A Companion to Hume. Ed. by Elizabeth S. Radcliffe. Oxford: Blackwell, pp. 159–84. ——— (2010). ‘Hume’s Moral Sentiments as Motives’. Hume Studies 36(2), pp. 193–213. Coleman, Dorothy, ed. (2007). Dialogues concerning Natural Religion, and other writings, by David Hume. Cambridge: Cambridge University Press. Garrett, Don (1997). Cognition and Commitment in Hume’s Philosophy. New York & Oxford: Oxford University Press. ——— (2015). Hume. New York: Routledge. Harris, James (2009). ‘The Epicurean in Hume’. In Epicurus in the Enlightenment. Eds. by Neven Leddy and Avi S. Lifschitz. Oxford: Voltaire Foundation, pp. 161–81. ——— (2015). Hume: An Intellectual Biography. Oxford: Oxford University Press. Hedley, Douglas and Sarah Hutton, eds. (2008). Platonism at the Origins of Modernity. Dordrecht: Springer. Immerwahr, John (1994). ‘Hume’s Dissertation on the Passions’. The Journal of the History of Philosophy 32(2), pp. 225–40. Immerwahr, John and Jacob Sider Jost (2013). ‘Hume the Sociable Iconoclast: The Case of the Four Dissertations’. The European Legacy 18(5), pp. 603–18. James, Susan (2005). ‘Sympathy and Comparison: Two Principles of Human Nature’. In Impressions of Hume. Eds. by Marina Frasca-Spada and P. J. E. Kail. Oxford: Clarendon Press, pp. 107–24. Jones, Peter (1982). Hume’s Sentiments: Their Ciceronian and French Context. Edinburgh: Edinburgh University Press.

28  Part I Kail, P. J. E. (2005). ‘Hume’s Ethical Conclusion’. In Impressions of Hume. Eds. by Marina Frasca-Spada and P. J. E. Kail. Oxford: Clarendon Press, pp. 125–39. ——— (2007). Projection and Realism in Hume’s Philosophy. Oxford: Oxford University Press. ——— (2008). ‘On Hume’s Appropriation of Malebranche: Causation and Self’. European Journal of Philosophy 16(1), pp. 55–80. Kemp Smith, Norman (1941). The Philosophy of David Hume. Basingstoke: Palgrave Macmillan. Lennon, Thomas M. (1997). ‘Introduction’. In The Search After Truth, by Nicolas Malebranche. Trans. by Thomas M. Lennon and Paul J. Olscamp. Cambridge: Cambridge University Press, pp. vii–xxiii. Locke, John (1690). An Essay concerning Humane Understanding. London. Quotations and page references from Peter H. Nidditch, ed. (1975). An Essay concerning Human Understanding, by John Locke. Oxford: Clarendon Press. Loeb, Louis (1977). ‘Hume’s Moral Sentiments and the Structure of the Treatise’. The Journal of the History of Philosophy 15(4), pp. 395–403. Millican, Peter (2007a). ‘Against the “New Hume” ’. In The New Hume Debate. Eds. by Rupert Read and Kenneth Richman. 2nd edition. London: Routledge, pp. 211–77. ——— (2007b). ‘Hume’s Old and New: Four Fashionable Falsehoods, and One Unfashionable Truth’. Proceedings of the Aristotelian Society, Supplementary Volumes 81, pp. 163–99. ——— (2009). ‘Hume, Causal Reasoning, and Causal Science’. Mind 118(471), pp. 647–712. ——— (2011). ‘Hume, Causal Reasoning, and Free Will’. In Causation and Modern Philosophy. Eds. by Keith Allen and Tom Stoneham. New York  & Oxford: Routledge, pp. 123–65. ——— (2012). ‘Hume’s Scepticism about Induction’. In The Continuum Companion to Hume. Eds. by Alan Bailey and Dan O’Brien. London: Continuum, pp. 57–103. Moore, James (1995). ‘Hume and Hutcheson’. In Hume and Hume’s Connections. Eds. by M. A. Stewart and John P. Wright. Pennsylvania: Penn State University Press, pp. 23–57. Mossner, E. C. (1936). ‘The Enigma of Hume’. Mind, New Series 45(179), pp. 334–49. ——— (1950). ‘Philosophy and Biography: The Case of David Hume’. The Philosophical Review 59(2), pp. 184–201. Noonan, Harold (1999). Hume on Knowledge. London: Routledge. Norton, David Fate (1982). David Hume: Common-Sense Moralist, Sceptical Metaphysician. Princeton: Princeton University Press. Owen, David (1999). Hume’s Reason. New York: Oxford University Press. Popkin, Richard H. (2003). The History of Scepticism: From Savoranola to Bayle. New York: Oxford University Press. Rickless, Samuel (2013). ‘Hume’s Theory of Pity and Malice’. The British Journal for the History of Philosophy 21(2), pp. 324–44. Schmitter, Amy M. (2012). ‘Family Trees: Sympathy, Comparison, and the Proliferation of the Passions in Hume and his Predecessors’. In Emotion and Cognitive Life in Medieval and Early Modern Philosophy. Eds. by Martin Pickavé and Lisa Shapiro. Oxford: Oxford University Press, pp. 255–78.

My Design in the Present Work  29 Tolonen, Mikko (2013). Mandeville and Hume: Anatomists of Civil Society. Oxford: Voltaire Foundation. Wilson, Catherine, ed. (2008). Epicureanism at the Origins of Modernity. Oxford: Clarendon Press. Wright, John P. (1983). The Sceptical Realism of David Hume. Manchester: Manchester University Press.

2 Some Late Philosophers in England

In this chapter I offer a brief introduction to the relevant views of five of Hume’s most important influences: John Locke (§2.1), Anthony Ashley Cooper, the third Earl of Shaftesbury (§2.2), Bernard Mandeville (§2.3), Francis Hutcheson (§2.4), and Joseph Butler (§2.5). This selection would be a suitable one anyway, but the choice is made easy by the fact that it is these five writers that Hume himself approvingly refers to, in the introduction to the Treatise, as ‘some late philosophers in England, who have begun to put the science of man on a new footing’ (T 0.7, p. xvii).1 My focus will be on the debate about self-love, which divided the egoists Locke and Mandeville from the other three. This discussion is also intended to set the scene for the argument of the next two chapters: that, in between writing the Treatise and the Four Dissertations, Hume changed sides on precisely this controversy.

2.1.  Mr. Locke To many in the eighteenth century, it seemed obvious that all motivation is ultimately reducible to self-love, the desire for personal happiness, and consequently that the basis of morality is the prospect of rewards and punishments, either in this life or in the next. Two particularly forthright examples of this view can be found in John Clarke (1725, 1726) and Robert Clayton (1751, 1754). (The former is not to be confused with the much better-known Samuel Clarke, to whom he bore no relation.) This view, which we now call psychological egoism, is also closely connected to psychological hedonism, the view that all motivation ultimately depends on expectations of pleasure and pain. This latter doctrine can be understood in either a narrow or a broad sense. In the narrow sense, only expectations of our own pleasure and pain count, and hedonism thus construed is just a special case of egoism. In the broad sense, however, expectations of other people’s pleasure and pain may also directly motivate us, and hedonism understood in this way is logically independent of egoism. It is convenient here to use the term in the broad sense. This allows us to say that, as a matter of fact, all egoists of the time were also

Some Late Philosophers in England  31 hedonists (at least so far as I am aware), but that not all hedonists were egoists. Francis Hutcheson is the most notable exception to this latter rule; see §2.4. It is impossible to discuss the early modern egoism debate adequately without mentioning Thomas Hobbes, who was invariably named by anti-egoists as their principal target. It would take too long to reconstruct Hobbes’s views here with any degree of accuracy or certainty, however, and I will use the excuse of Hume’s footnote in the introduction of the Treatise to set this task to one side.2 In any case, a detailed examination of Hobbes would not add anything essential to the story I will be telling here, since the relevant ‘Hobbist’ ideas can all be found in either Locke or Mandeville.3 I  will say more about Hobbes, but in connection with Hume’s views on religion, in Chapter 5. If Hobbes was the first name in egoism in the early modern period, John Locke was the second: in Hume’s words, it was ‘Hobbes and Locke’, among the modern philosophers, ‘who maintained the selfish system of morals’ (M App2.3, p. 296). The two were a generation apart. Born in 1632, Locke was 44 years Hobbes’s junior, and his two major works—the Two Treatises of Government (1689) and the Essay concerning Human Understanding (1690)—appeared ten years after Hobbes’s death. Where Hobbes’s political philosophy was developed in the context of the English civil war (1642–60), Locke’s relates to the later revolution of 1688, in which King James II was overthrown. Locke’s Two Treatises were, respectively, an attack on absolute monarchy and the divine right of kings, and a defence of liberal democracy. These intriguing matters are not, however, my present concern. What chiefly interests me about Locke here is his psychological egoism and hedonism. In his Essay he maintains that desire is moved by ‘happiness and that alone’, with happiness being understood hedonistically: ‘Happiness then in its full extent is the utmost Pleasure we are capable of, and Misery the utmost pain’ (1690, p. 258). Indeed, the whole point of pleasure, according to Locke, is to excite us to action, and without it our wills would be entirely inert: [T]o excite us to these Actions of thinking and motion, that we are capable of, [the Author of our being] has been pleased to join to several Thoughts, and several Sensations, a perception of Delight. If this were wholly separated from all our outward Sensations, and inward Thoughts, we should have no reason to prefer one Thought or Action, to another; Negligence, to Attention; or Motion, to Rest. (1690, p. 129) Locke was a hedonist not only about motivation, in fact, but also about the passions more generally, maintaining that ‘Pleasure and Pain . . . are the hinges on which our Passions turn’ (1690, p. 229). In the paragraphs

32  Part I following this claim, he proceeds to explain how various passions are ‘moved by things, only as they appear to be the Causes of Pleasure and Pain, or to have Pleasure or Pain some way or other annexed to them’ (1690, p. 232). From the context it is clear that it is only our own pleasure and pain that he has in mind here, and consequently that his account is egoist as well as hedonist. Conformable to this egoism, Locke maintained that moral motivation was ultimately a matter of self-interest, ‘it being impossible to set any other motive or restraint to the actions of a free understanding agent but the consideration of good or evil; that is, pleasure or pain that will follow from it’ (c. 1686–8, p. 301). More than this, however, morality for Locke could only be grounded by appeal to each individual agent’s self-love, together with the rewards and punishments that are consequent upon virtue and vice respectively: ‘the true ground of Morality . . . can only be the Will and Law of a God, who sees Men in the Dark, has in his Hand Rewards and Punishments, and Power enough to call to account the Proudest Offender’ (1690, p. 69). Or again: ‘To establish morality . . . upon its proper basis, and such foundations as may carry an obligation with them, we must first prove a law, which always supposes a lawmaker: one that has a superiority and right to ordain, and also a power to reward and punish according to the tenor of the law established by him. This sovereign lawmaker who has set rules and bounds to the actions of men is God, their maker’ (c. 1686–8, p. 304). Psychological egoism is closely related to a view that we might call pessimism, according to which—to put it crudely—human beings are predominantly not very nice. It is important to be clear, however, that although egoism invites pessimism (and may be used by pessimists to support their view), it does not strictly entail it. Egoism entails that we do nice things for other people only with a view to getting something for ourselves in return, but what we want in return may simply be the pleasure of seeing that the other person is happy. Locke himself was open to the possibility of these innocent pleasures: [N]ature for wise ends of her own has made us so that we are delighted with the very being of our children. Some wise minds are of a nobler constitution, having pleasure in the very being and the happiness of their friends, and some yet of a more excellent make [are] delighted with the existence and happiness of all good men, some with that of all mankind in general, and this last may be said properly to love. (1676, p. 238) Egoism is a theoretical point about how motivation in general works, more than a practical observation about what particular motives people

Some Late Philosophers in England  33 have.4 And it is not obvious that Locke was a pessimist. When he emphasizes the prospect of some future reward for virtue, distinct from the pleasure of seeing other people happy, his concern is more with the foundation of morals than with the question of our immediate motives. Pessimist or not, it is clear and (so far as I am aware) uncontroversial that Locke was an egoist. There is however a puzzle in the interpretation of his moral philosophy, arising from the fact that he appears to have been a rationalist as well. ‘I am bold to think’, he wrote in the Essay, ‘that Morality is capable of Demonstration, as well as Mathematicks: Since the precise real Essence of the Things moral Words stand for, may be perfectly known; and so the Congruity, or Incongruity of the Things themselves, be certainly discovered, in which consists perfect Knowledge’ (1690, p. 516). A popular line of thought here is that, while egoism may be at the heart of his account of moral motivation, his account of moral obligation is rationalist; thus Colman (1983), Darwall (1995), and Sheridan (2007). I cannot do justice to this controversy here, and for present purposes I needn’t insist that Locke was an egoist about moral obligation as well as motivation. I may just remark in passing that the rationalism evident in Locke’s work may admit of a more innocuous interpretation. This interpretation is already suggested by the sentence just quoted, but is brought out more clearly in a manuscript draft of a section intended for the Essay: Were there no human law, nor punishment, nor obligation of civil or divine sanctions, there would yet still be such species of actions in the world as justice, temperance, and fortitude, drunkenness and theft. . .; there would be distinct notions of virtues and vices; for to each of these names there would belong a complex idea, or otherwise all these and the like words . . . would be empty insignificant sounds.  .  .  . But the knowledge of virtues and vices which a man attained to this way would amount to no more than taking the definitions or significations of the words of any language. . .; and so in effect would be no more but the skill how to speak properly. (c. 1686–8, p. 299) It is thus perhaps only in this rather trivial sense that Locke thought morality was capable of demonstration: moral truths simply follow from the definitions of moral terms. But these demonstrations, in taking such definitions as their starting point, cannot amount to a foundation of morality. They offer nothing that can move or oblige us either to speak in this way or to conform our actions to these ideas. For that, it would seem, we need to ‘deriv[e] these rules up to their original,  .  .  . urging them as the commands of the great God of heaven and earth, and such as according to which he would retribute to men after this life’ (ibid.).

34  Part I

2.2.  My Lord Shaftesbury In 1666, Locke met Anthony Ashley Cooper, one of the founding members of the Whig party, and later the first Earl of Shaftesbury. Cooper employed Locke the following year as a personal aide and physician, and Locke was closely associated with the family for the rest of his life. He found a suitable wife for the first Earl’s son, the second Earl, and was a tutor to the resulting third Earl. It is this third Earl, also called Anthony Ashley Cooper, but generally referred to now by his title, to whom we now turn. Shaftesbury’s political views—regarding freedom of speech, religious toleration, and constitutional government—were closely in line with Locke’s (and his own grandfather’s). Notwithstanding these agreements, however, and the close family association, the third Earl consciously modelled his life and thought on the ancient Stoics, and was a fierce opponent of Epicureanism. He published a few pieces individually at the start of the eighteenth century, soon collecting these together (and adding to them) in his three-volume Characteristicks of Men, Manner, Opinions, Times (1711). Hobbes is named as an explicit target in several places in the Characteristicks, but Shaftesbury was by no means ignorant of Locke’s similarly Epicurean views.5 Shaftesbury did not always write in a recognizably philosophical style (recognizable either to us or to his contemporaries), and this was intentional. His aims were perhaps as much therapeutic as they were theoretical. For brevity, however, I will focus here on his most conventionally philosophical work, the Inquiry concerning Virtue or Merit (in volume 2 of the Characteristicks). This Inquiry is divided into two books, in the second of which Shaftesbury attempts to prove that virtue is in every individual’s self-interest. Notwithstanding this pragmatic justification, his theoretical position—as stated in the first book—is anti-egoist through and through. In the first instance, Shaftesbury argues that every human being is part of the larger system of human society, and consequently that each of us may be considered good or ill according to whether our intentions and affections are for the good or ill of this system. This does not mean that self-love is necessarily a bad thing; to an extent, and properly directed, it is in fact a good thing. Like all affections, however, it is good only insofar as it tends to the good of the system, rather than to that of the individual. Where the good of others is directly concerned, moreover, it is positively bad: Whatsoever therefore is done which happens to be advantageous to the Species, thro an Affection merely towards Self-Good, does not imply any more Goodness in the Creature than as the Affection it-self is good. Let him, in any particular, act ever so well; if at the bottom, it be that selfish Affection alone which moves him; he is in himself

Some Late Philosophers in England  35 still vitious. Nor can any Creature be consider’d otherwise, when the Passion towards Self-Good, tho ever so moderate, is his real Motive in the doing that, to which a natural Affection for his Kind ought by right to have inclin’d him. (1711, p. 171) In sum, ‘[a] good Creature is such a one as by the natural Temper or Bent of his Affections is carry’d primarily and immediately, and not secondarily and accidentally, to Good, and against Ill’ (1711, p. 171). Having thus defined what it is to be good, Shaftesbury goes on to consider what it is to be virtuous, a characteristic he takes to be unique to human beings. Where any animal may be good, virtue depends on the specifically human ability to reflect on our own motives, and to feel ‘another kind of Affection towards those very Affections themselves, which have been already felt, and are now become the Subject of a new Liking or Dislike’ (1711, p. 172). Where the good person has good intentions, the virtuous person has, over and above these, the intention to be good: So that if a Creature be generous, kind, constant, compassionate; yet if he cannot reflect on what he himself does, or sees others do, so as to take notice of what is worthy or honest; and make that Notice or Conception of Worth and Honesty to be an Object of his Affection; he has not the Character of being virtuous: for thus, and no otherwise, he is capable of having a Sense of Right or Wrong; a Sentiment or Judgment of what is done, thro just, equal, and good Affection, or the contrary. (1711, p. 173) It is this affection for what is morally good that matters most in the assessment of the virtuous character. People may in fact have all manner of ill intentions, but so long as their inclination towards moral good prevails, they are still to be counted virtuous; indeed, the fact that these contrary inclinations need to be overcome shows that the principle of virtue is in their case all the stronger (1711, pp. 175–6). Shaftesbury’s Stoicism was not the only brand of anti-Epicurean thought to emerge in England at the start of the eighteenth century. The main alternative was the rationalism of Shaftesbury’s contemporary Samuel Clarke, as expounded in his two sets of Boyle Lectures delivered in 1704 and 1705, and each published a year later (Clarke 1705, 1706). Where Shaftesbury opposed self-love with benevolent affections, Clarke opposed it first and foremost with reason, and broadly speaking his views belong in the Cartesian or Platonic tradition that I introduced in ­Chapter  1. It may be noted, however, that Clarke’s Anglican brand of Platonism was perhaps somewhat less extreme than the Catholic Platonism of Frenchmen such as Nicolas Malebranche.

36  Part I I will say more about both Clarke and Malebranche in Chapter 8. For now, it is perhaps helpful to emphasize the similarities between Shaftesbury and Clarke. Though of course there were many important differences, and differences that were magnified as their respective traditions developed in later writers, there was at least a fundamental point of agreement at the outset: that Epicurus and Hobbes were dangerously wrong. Clarke and Shaftesbury were both sincere and devout Anglicans (with Clarke’s relative orthodoxy lying somewhere between Shaftesbury’s deistic and free-thinking tendencies, and what both would have recognized as the superstitious excesses of the Roman Catholics). More importantly for present purposes, while Clarke held that virtue required us to act on the rational perception of our duty, he did not deny the existence of natural inclinations to the same end: And as all Men are obliged to this [i.e. to embrace all Fellow-­ Creatures with universal Love, Charity, and Benevolence], by the necessary Law and Condition of their Being, and by all the outward Circumstances of the present State, wherein God has placed them; so they are also strongly prompted to it by the natural Inclinations of their own Minds, when not corrupted by the practice of Vice. For by Nature Men are plainly disposed to be kind, and friendly, and willing to do good; Nothing is naturally more agreeable and pleasant to the Mind of Man, than being helpful and beneficial one to another. (Clarke 1705, p. 4) While Shaftesbury insisted that a virtuous act could only stem from a virtuous sentiment, meanwhile, he took it for granted that reason could and should be used to justify and correct our affections, as, for example, in the following argument against partiality: But lest any shou’d imagine with themselves that an inferior Degree of natural Affection, or an imperfect partial Regard of this sort, can supply the place of an intire, sincere, and truly moral one  .  .  . we may consider first, That Partial Affection, or social Love in part, without regard to a compleat Society or Whole, is in it-self an Inconsistency, and implies an absolute Contradiction. . . . The Person, therefore, who is conscious of this Affection, can be conscious of no Merit or Worth on the account of it. It has no Foundation or Establishment in Reason. (Shaftesbury 1711, p. 205) Here Shaftesbury inconsistent, and remembered that moral philosophy

rejects a certain sentiment on the grounds that it is has no foundation in reason. And so it should be what we now call sentimentalism, the tradition in that runs from Shaftesbury, through Hutcheson and

Some Late Philosophers in England  37 Butler, and then—albeit substantially altered—to Hume himself, started life as an alternative to egoism rather than to rationalism.

2.3.  Dr. Mandeville The polite and honourable Stoicism of Lord Shaftesbury had its most striking contemporary foil in the wry and down-to-earth Epicureanism of Bernard Mandeville. In 1705, Mandeville published a short allegory in verse called The Grumbling Hive, or Knaves Turned Honest. The story begins by describing a large hive of industrious and prosperous bees, individually selfish, vain, and deceitful, but on the whole thriving. Greed fuelled the industry and trade of this hive, so that in time luxury rose ‘To such a Height, the very Poor/Liv’d better than the Rich before’ (1705, p. 11). But some particularly immoral bees then started to preach against all the dishonesty, and pray for virtue. Jove, angered by their hypocrisy, decided that the best punishment would be to grant them their wish. It was no sooner done than the hive collapsed: with everyone now a model of virtue, there was no more work for lawyers or prison guards; every bee settled its bar tab and resolved never to drink again, so the public houses closed; all being content with plain clothes, tailors were forced to shut up shop; and so on, the moral of the story being that vice is beneficial or even necessary for a great and prosperous state. In 1714, a year after Shaftesbury’s death (and coinciding with a posthumous second edition of the Characteristicks), Mandeville published The Grumbling Hive again, together with a short Enquiry into the Origin of Moral Virtue, and a lengthier set of remarks on the poem, all under the title of The Fable of the Bees: Or, Private Vices, Publick Benefits. In a second edition of 1723, he added two further essays, one on Charity and Charity-Schools, and another titled A Search into the Nature of Society. I will refer to these five works collectively as the Fable of the Bees. It is in this second edition, at the start of the Search into the Nature of Society, that Shaftesbury is first mentioned as an explicit target: The attentive Reader, who perused the foregoing part of this Book, will soon perceive that two Systems cannot be more opposite than his Lordship’s and mine. His Notions I confess are generous and refin’d: They are a high Compliment to Human-kind, and capable by the help of a little Enthusiasm of Inspiring us with the most Noble Sentiments concerning the Dignity of our exalted Nature: What Pity it is that they are not true. (1723, p. 372) Care is required in piecing together the moral and political theory Mandeville meant to put forward in his Fable of the Bees, for the fact is that he combined more abstract theoretical speculation with a sharp criticism

38  Part I of the morals of his day (i.e. the morals of noblemen such as Shaftesbury). The claim that private vices give rise to public benefits can be taken at face value, and I am sure that Mandeville often intended it to be. But it can also have a subversive implication: if a supposedly virtuous deed arises from questionable or hypocritical motives, one might wonder whether it really is a virtuous action at all. In the Enquiry into the Origin of Moral Virtue, Mandeville describes the moral virtues as ‘the Political Offspring which Flattery begot upon Pride’ (1714, p. 37), essentially the result of cunning politicians playing on the natural vanity of the populous, by praising public-spirited and self-sacrificing behaviour. It would be a mistake to think Mandeville was entirely opposed to this brain-child, and to the social and economic advantages it made possible. But it would also be a mistake to think he was entirely in favour of it. His criticism is implicit, or so it seems to me, in his discussions of lust and chastity (1714, pp. 59–65, 150–4), which are amusingly and refreshingly direct for the time. And it is explicit in his attack on the practice of posthumous endowments and the creation of Christian charity schools, both of which he saw as hideous exercises in vanity and self-aggrandizement, and a waste of time and money that could be put to much better use elsewhere (1723, pp. 285–370). Not to trouble ourselves too much with these difficulties, the Fable certainly endorses psychological egoism. In the first sentence of the Enquiry, Mandeville states that ‘[a]ll untaught Animals are only solicitous of pleasing themselves, and naturally follow the bent of their own Inclinations, without considering the good or harm that from their being pleased will accrue to others’ (1714, p.  27). The claim is reinforced throughout the work in too many places to cite. Concerning the affection of mothers towards their children, for example, which might be thought the least likely passion to arise from self-interest, the counterintuitive implication is overtly embraced: ‘All Mothers naturally love their Children: but as this is a Passion, and all Passions centre in Self-Love, so it may be subdued by any Superior Passion, to sooth that same SelfLove, which if nothing had interven’d, would have bid her fondle her Offspring’ (1714, p. 67). Mandeville defines virtue as applying ‘to every Performance, by which Man, contrary to the impulse of Nature, should endeavour the Benefit of others, or the Conquest of his own Passions out of a Rational Ambition of being good’ (1714, pp. 48–9). This definition is close to Shaftesbury’s account. For Shaftesbury, as we have seen, it is not enough just to endeavour to benefit others; we must reflect on the goodness of this endeavour, and make that goodness itself our end. For Mandeville, however, it is impossible to act other than from an impulse of nature, and our natural impulses are all selfish. Any attempt at a virtuous act in this sense, therefore, is necessarily hypocritical, and Shaftesbury’s desire of

Some Late Philosophers in England  39 being good is vanity in disguise, our natural desire for praise artificially moulded by politicians into something publicly useful. In 1729, Mandeville published a set of dialogues between two characters, Cleomenes and Horatio (a minor third character, Fulvia, leaves before the end of the first dialogue), under the title The Fable of the Bees, Part II. Cleomenes is essentially Mandeville’s mouthpiece in this work, being an explicit supporter of the views put forward in the original Fable. Horatio’s sympathies at the outset lie with Shaftesbury, and he is shocked by what he has heard of Mandeville’s ideas (though, like many of Mandeville’s contemporary critics, he has not read the book himself). As their discussion proceeds—over the course of several days—Horatio is persuaded to read the book, and is increasingly brought around to Mandeville’s and Cleomenes’s point of view. In an interesting recent study, Mikko Tolonen has argued that in Part II, and in the later Enquiry into the Origin of Honour (1732), Mandeville’s views are significantly different from those put forward in the Fable (Tolonen 2013). Alas, there is no space to pursue these matters in any detail here. I  am sympathetic to a lot of what Tolonen says about the development of Mandeville’s thought. Broadly speaking, my own sense is that, as time went by, Mandeville became less interested in criticizing Stoical morality, and more interested in serious speculation about how human society developed. Thus the crude story from the Fable—about how politicians invented virtue, apparently overnight, to flatter the population into self-denial and acts of public-spirited politeness, chastity, and generosity—is gone from Part II. In its place we have a much more sophisticated and realistic account, the details of which I won’t go into here, but which involves the confluence of several favourable environmental circumstances, a good deal of luck, and generations of trial and error. On at least one point, however, I am not persuaded by Tolonen’s interpretation. Tolonen argues that, from Part II onwards, Mandeville was no longer an egoist, but the evidence for this striking claim does not seem to me particularly strong. It is true, for example, that Mandeville does not repeat his explicit earlier assertion that the natural affection of parents to their children is, like every passion, a species of self-love; but nor, as far as I can see, does he ever explicitly deny it. Here are some of the things Cleomenes says about this passion: Natural Affection prompts all Mothers to take Care of the Off-spring they dare own; so far as to feed and keep them from Harm. (1729, p. 211) All Creatures naturally love their Offspring, whilst they are helpless, and so does Man. (1729, pp. 222–3)

40  Part I Natural Affection would prompt a wild Man to love, and cherish his Child; it would make him provide Food and other Necessaries for his Son, till he was ten or twelve years old, or perhaps longer. (1729, p. 225) Natural Affection would make wild Men, and Women too, sacrifice their Lives, and die for their Children. (1729, p. 277) ‘From his treatment of this original passion’, Tolonen writes, ‘it is difficult, or in fact ineffective, to claim that Mandeville was now arguing that man was, by nature, wholly incapable of other-regarding affection’ (2013, p. 74). I wouldn’t claim that Mandeville was arguing for egoism here, but the absence of any such argument does not entail the rejection of its conclusion. After all, an egoist need not contradict himself merely by acknowledging the existence of parental love, nor even by admitting that this passion can move us to give up our very lives for our children. All this shows, as far as the egoist is concerned, is that the pain we would suffer by not doing so—as it might be, the grief and the guilt—is more than we feel able to bear. According to Tolonen’s Mandeville, ‘ “natural affection” is such a powerful principle that it would force “a wild man to love, and cherish his child” without any concern for his own interest’ (2013, p. 73). But the crucial anti-egoist gloss here is Tolonen’s, not Mandeville’s (the full quotation is given earlier). And in Part II, Mandeville continues to insist, in the character of Cleomenes, that ‘[a]ll Men uninstructed, whilst they are let alone, will follow the Impulse of their Nature, without regard to others; and therefore all of them are bad, that are not taught to be good’ (1729, pp. 315–6). As with the question of Locke’s potentially rationalist attitude to moral obligation, however, I need not insist here on this controversial point. It suffices to note that both Locke and Mandeville were, at least at some point in their careers, psychological egoists, and that this view was explicitly targeted by many of their critics at the time.

2.4.  Mr. Hutcheson Mandeville and Shaftesbury were born just one year apart (in 1670 and 1671 respectively), but Shaftesbury died before his 42nd birthday, not long after his Characteristics were first published and just before the first edition of Mandeville’s Fable appeared. He thus had no chance to reply to Mandeville’s attack. The burden of this task fell instead to the younger Francis Hutcheson, who entered the lists in 1725 (two years after the second edition of the Fable), with his Inquiry into the Original of Our Ideas of Beauty and Virtue. In this work, according to the title page of its

Some Late Philosophers in England  41 first edition, ‘the principles of the late Earl of Shaftesbury are Explain’d and Defended, against the Author of the Fable of Bees’ (1725, p. 199). For Hutcheson, as for Shaftesbury, beauty and virtue are intimately related, and the Inquiry is a pair of treatises discussing these two subjects in turn. I will restrict my attention here to the second treatise, Concerning the Original of our Ideas of Virtue or Moral Good. In this work, according to its introduction, Hutcheson’s aim is to prove two related anti-egoist theses: I. ‘That some Actions have to Men an immediate Goodness; or, that by a superior Sense, which I call a Moral one, we perceive Pleasure in the Contemplation of such Actions in others, and are determin’d to love the Agent, (and much more do we perceive Pleasure in being conscious of having done such Actions our selves) without any View of further natural Advantage from them.’ II. It may perhaps also appear, ‘That what excites us to these ­Actions which we call Virtuous, is not an Intention to obtain even this sensible Pleasure; much less the future Rewards from Sanctions of Laws, or any other natural Good, which may be the Consequence of the virtuous Action; but an entirely different Principle of Action from Interest or Self-Love.’ (1725, pp. 86–7) This entirely different principle of action is a universal and disinterested benevolence: ‘some Determination of our Nature to study the Good of others; or some Instinct, antecedent to all Reason from Interest, which influences us to the Love of others; even as the moral Sense, above explain’d, determines us to approve the Actions which flow from this Love in our selves or others’ (1725, p. 122). Though ostensibly defending Shaftesbury, and obviously very much indebted to him, Hutcheson had many of his own ideas and arguments. Not least among these is the distinction—implicit in the two propositions just quoted from his introduction, and which runs through the whole of the treatise—between moral motivation and moral approbation. This distinction is a clear echo of Shaftesbury’s distinction between goodness and virtue, the latter depending on a reflective awareness of the former, but it is nevertheless importantly different. For Shaftesbury, virtuous motives (as opposed to merely good motives) necessarily involve this reflective awareness, and consequently involve the desire to do the right thing in part because it is the right thing to do. It was this motive that Mandeville diagnosed as vain and hypocritical. For Hutcheson, however, the immediate motive to virtue is benevolence, plain and simple. Our moral sense determines us to approve of that motive, and this in turn may give rise to an additional (and self-interested) reason to behave in a way that we reflectively approve of, and in a way that keeps us clear

42  Part I from the pain of guilt. But this derivative desire is in no way a part of benevolence itself. On the contrary, being self-interested, it is antithetical to that virtuous motive. Hutcheson’s distinction between motivation and approbation is not to be found in Mandeville any more than in Shaftesbury. And so, although the younger philosopher’s insistence on the existence of a genuinely disinterested benevolence (in section  2 of his second treatise) is in direct opposition to Mandeville, his parallel argument that moral approbation does not depend on self-love (in section  1) seems misplaced. It is very unclear that anyone before him had maintained that it did. Rather, the relevant opposition here would seem to be with the Epicurean view that our moral approbation is determined artificially, directed at whatever actions and motives our parents, our peers, or cunning politicians happen to encourage. Hutcheson certainly rejected this view: our reflective attitudes, he argues in sections  4 and 5, are determined by a natural moral sense that is common to all mankind. Perhaps the point of the argument in section  1, then, is just to show that even the derivative, self-interested motive to virtue itself depends on a principle distinct from self-love. In a series of letters to the London Journal in 1725, the same year that Hutcheson’s Inquiry first appeared, Gilbert Burnet (the younger) criticized Hutcheson for failing to provide ‘a sufficient Foundation’ for morality: ‘tho’ the Conclusions were generally True and Right in themselves’, Burnet wrote, ‘and were capable of Demonstrative Proof, yet he seemed to me to have left them unsupported’ (Burnet & Hutcheson 1735, p. iii).6 More specifically, Burnet raised a sceptical worry with regard to Hutcheson’s moral sense as the source of approbation: I saw indeed, there was some such thing in humane Nature. But . . . I could not be sure, it was not a deceitful and wrong Sense. The Pleasure arising from the Perceptions it afforded, did not seem sufficient to convince me that it was right. . . . I wanted therefore some further Test, some more certain Rule, whereby I  could judge whether my Sense, my moral Sense as the Author calls it, my Taste of Things, was right, and agreeable to the Truth of Things, or not. (Burnet & Hutcheson 1735, pp. 9–10) In order to answer this worry, Burnet appealed to the moral rationalism of Richard Cumberland (1672), Samuel Clarke (1706), and William Wollaston (1722), according to which the things approved of by Hutcheson’s moral sense are indeed reasonable or agreeable to truth. In 1728, Hutcheson published a second pair of treatises (which he referred to as his third and fourth), An Essay on the Nature and Conduct of the Passions and Illustrations on the Moral Sense. It was apparently Burnet’s attack that prompted Hutcheson to write the fourth of these

Some Late Philosophers in England  43 treatises. In the preface to the first edition, he claims that it ‘had never seen the light, had not some worthy Gentleman mistaken some things about the moral Sense alleg’d to be in Mankind’ (1728, p. 7), and in the third edition he added an explicit reference to the correspondence in the London Journal (1742, p. 10). It is in this treatise, perhaps, that the split between the Stoical and Platonic traditions of anti-egoist thought was first made official. I will return to this aspect of Hutcheson’s philosophy in Chapter 8. Though Hutcheson was vigorously opposed to psychological egoism, the third of his four treatises makes it clear that he did not reject psychological hedonism (in the broad sense defined in §2.1). On the contrary, he seems to have taken it for granted that pleasure and pain were the only ultimate objects of desire and aversion, insisting only that other people’s pleasures and pains could be ends in themselves alongside our own: ‘Desires arise in our Mind, from the Frame of our Nature, upon Apprehension of Good or Evil in Objects, Actions, or Events, to obtain for our selves or others the agreeable Sensation, when the Object or Event is good; or to prevent the uneasy Sensation, when it is evil’ (1728, p. 18). Where the egoist reduces all desires to just one, namely self-love, Hutcheson therefore reduces them to four: ‘Self-Love, Self-Hatred, or desire of private Misery, (if this be possible) Benevolence toward others, or Malice: All Affections are included under these’ (1728, p. 139). Of these four, however, it is self-love and benevolence that Hutcheson takes to underlie most of our actions. Self-hatred and malice are logical possibilities, but scarcely more than that; the former is never mentioned except in the one quotation just given, and Hutcheson describes the latter as ‘unnatural’ and rare, suggesting that cruel or unkind actions are more often the result of self-love when there is an opposition of interests (1725, p. 105).

2.5.  Dr. Butler The generation after Shaftesbury contained two notable inheritors of the Stoical tradition. Hutcheson was one, and Joseph Butler the other. A keen and extremely capable philosopher of religion, Butler read Clarke’s first set of Boyle Lectures with great interest, but was not persuaded by the attempted a priori demonstration of God that they contained. He took the liberty of writing to Clarke anonymously with some of his queries and objections, and a very amicable debate ensued. Clarke clearly recognized the intelligence of his unknown critic, and appended the correspondence (five letters each) to the fourth edition of his lectures (1716). Twenty years later, Butler published his own attempt at proving the existence of God, in his Analogy of Religion (1736). In Stoical fashion, his case was a version of the a posteriori design argument, and he remained convinced that no a priori demonstration was possible. The book was enormously successful, running to a second edition in the very same year.

44  Part I In the intervening period, Butler had been ordained into the Church of England, and given the job of preacher at the Rolls Chapel in London. Most of his sermons do not survive, but in 1726, a year after the appearance of Hutcheson’s Inquiry, he published a set of Fifteen Sermons Preached at the Rolls Chapel which included (among other things) an attack on Epicureanism and psychological egoism in a somewhat similar spirit to Hutcheson’s. The Sermons were not perhaps quite as well received as the Analogy at first, but a second edition (incorporating a new preface) was published in 1729, and a third in 1736. I presume this last was a response to demand generated by the success of the Analogy. Like Hutcheson, Butler insists that we have an abundance of genuinely disinterested and benevolent desires, and that these are the source of our virtuous behaviour. Furthermore, his appeal to what he calls the principle of conscience, at the heart of his moral theory, is by his own (later) admission very nearly just Hutcheson’s moral sense by another name: [W]e have a Capacity of reflecting upon Actions and Characters, and making them an Object to our Thought: And on doing this, we naturally and unavoidably approve some Actions, under the peculiar view of their being virtuous and of Good-desert; and disapprove others, as vicious and of Ill-desert.  .  .  . It is manifest great Part of common Language, and of common Behaviour over the World, is formed upon Supposition of such a Moral Faculty; whether called Conscience [the term Butler himself typically prefers], moral Reason, moral Sense, or divine Reason; whether considered as a Sentiment of the Understanding, or as a Perception of the Heart, or, which seems the Truth, as including both. (1736, p. 309) What this quotation also illustrates, however, is that Butler’s sentimentalism—­ unlike Hutcheson’s—was not intended as a rival to Clarke’s moral rationalism, but as a complement to it. Indeed, Butler pointedly switches the ordinary use of the terms, speaking of a sentiment of the understanding, and a perception of the heart, while being indifferent between calling the moral faculty ‘moral Reason’ or a ‘moral Sense’. More explicitly, in the preface added to the second edition of his Sermons, Butler distinguishes ‘two Ways in which the Subject of Morals may be treated’: ‘One begins from inquiring into the abstract Relations of things: the other from a Matter of Fact, namely, what the particular Nature of Man is, its several parts, their Oeconomy or Constitution’ (1729, p. 37). He says that both methods lead to the same thing, and has nothing bad to say about the former (which is Clarke’s approach); he merely remarks that his own work proceeds mainly in the latter way (ibid.). A second significant difference between Butler and Hutcheson, and a more important one in the present context, is that Butler argued against psychological hedonism as well as egoism, by exposing a fundamental

Some Late Philosophers in England  45 error in both views. The error common to Locke and Hutcheson is the assumption that, because all desires terminate in happiness or misery (our own or someone else’s), this happiness or misery must also be their ultimate end. Butler concedes that our own pleasure or happiness is indeed the ultimate end of self-love: ‘[t]he object [this passion] pursues is somewhat internal, our own Happiness, Enjoyment, Satisfaction’ (1726, 1729, p. 111). But he goes on to distinguish self-love from all our other ‘particular’ appetites and passions, whose ultimate ends are many and various: That all particular Appetites and Passions are towards external Things themselves, distinct from the Pleasure arising from them, is manifest from hence; that there could not be this Pleasure, were it not for that prior suitableness between the Object and the Passion: There could be no Enjoyment or Delight from one thing more than another, from eating Food more than from swallowing a Stone, if there were not an Affection or Appetite to one thing more than another. (1726, p. 111) Two things follow from this simple but very important point. First, none of these ‘particular’ appetites or passions is reducible to self-love, on pain of circularity. Second, there is no tension between self-love and benevolence, any more than between self-love and vanity, or self-love and resentment. Self-love is satisfied when any of our ‘particular’ passions is satisfied, however selfish or selfless (in the everyday sense of these terms) that passion may be. To put it another way, Butler’s central insight is that egoism and hedonism both get things fundamentally back to front: objects are not desired because they are pleasant (at least not in the first instance); rather, they are pleasant because we desire them. And so, where Locke had maintained that without pleasure there could be no desire, Butler insists that without desire there could be no pleasure: [T]he very idea of an interested Pursuit, necessarily pre-supposes particular Passions or Appetites; since the very Idea of Interest or Happiness consists in this, that an Appetite or Affection enjoys its Object. . . . Take away these Affections, and you leave Self-love absolutely nothing at all to employ itself about; no End or Object for it to pursue, excepting only that of avoiding Pain. (1729, p. 20) I will refer to this view as motivational pluralism, in virtue of the fact that it involves the postulation of a large number of primitive and irreducible desires. This is in contrast both with egoism, which allows for just one primitive desire, and with Hutcheson’s anti-egoist hedonism, which allows for at most four.

46  Part I Butler’s many primitive or ‘particular’ passions will feature heavily in the discussion of Hume to follow, and it will be convenient to have a name for them. Rather than adopt Butler’s terminology, which is not very descriptive, I will refer to them instead as primary or original desires, in contrast to the secondary or interested desires that depend on self-love. The point of this terminology is that original desires are explanatorily and causally basic. Not to get ahead of ourselves, this terminology also agrees with Hume’s later appropriation of Butler’s theory, as we will see in Chapter 4. Before we can appreciate the considerable change of mind this appropriation represents, however, we must first examine the hedonistic and egoistic theory of motivation that Hume put forward in his Treatise. This is the subject to which I now turn.

Notes 1  Hutcheson was of Scottish descent, born in Ireland, and he spent significant periods of time in both countries. Mandeville was Dutch, though he emigrated to England and passed most of his life in London. The other three were indeed English. 2 It is interesting that Hobbes is not included in Hume’s footnote. Paul Russell has hypothesized that the Treatise was modelled on Hobbes’s Human Nature and De Corpore Politico (Hobbes 1650a,b), and conjectured that Hobbes is not on the list precisely because his influence on Hume was so great, and Hume thought it unwise to draw attention to that fact (Russell 2008, pp. 61–7). For scepticism regarding this conjecture, see Harris, who suggests instead that ‘Hobbes is not included on the list because he was, precisely, not one of those who put the science of man on a new footing—that is, because his science of man is grounded on an a priori commitment to materialism and to mechanistic physics, and not the Baconian inductivism that Newton had had such success with’ (Harris 2009, p. 40). 3 In recent times, Bernard Gert has argued that Hobbes was not an egoist in any important sense (Gert 1967, 2006). I do not agree, but my disagreement is as much with Gert’s understanding of egoism as with his understanding of Hobbes (see note 4). 4 As Hume puts it: ‘An Epicurean or a Hobbist readily allows, that there is such a thing as friendship in the world, without hypocrisy or disguise; though he may attempt, by a philosophical chymistry, to resolve the elements of this passion, if I may so speak, into those of another, and explain every affection to be self-love, twisted and moulded, by a particular turn of imagination, into a variety of appearances’ (M  App2.4, pp.  296–7). Gert, while denying that Hobbes maintained ‘psychological egoism’, allows that he did endorse ‘tautological egoism’, ‘a view which sounds like psychological egoism but which has no empirical consequences’ (1967, p. 507). Gert seems to imply that ‘tautological’ egoism is true but uninteresting, and that Hobbes’s commitment to it does not justify the standard egoist interpretation (though it may, together with Hobbes’s acknowledged pessimism, go a long way towards explaining it). Like Butler and the later Hume, however, I consider this view to be neither tautological nor uninteresting; it is a significant and superficially attractive position that, on reflection, turns out to be false. And while it has, strictly speaking, no immediate empirical consequences, it certainly encourages the pessimism for which Hobbes is so notorious. In freeing us of it, meanwhile, Butler made the corresponding optimism seem much less naïve.

Some Late Philosophers in England  47 5 In a letter written to a student whom he had taken under his wing, published posthumously in 1716, Shaftesbury wrote: ‘In general truly it has happened, that all those they call Free-Writers now-a-days, have espoused those Principles, which Mr. Hobbes set a foot in this last Age. Mr. Locke, as much as I honour him on account of other Writings (viz. on Government, Policy, Trade, Coin, Education, Toleration, &c.) and as well as I knew him, and can answer for his Sincerity as a most zealous Christian and Believer, did however go in the self same Track’ (1716, p. 38). 6 This book was a reprint of the letters, together with a preface and postscript written by Burnet ‘some time before his Death’ (title page). He died in 1726. See Peach (1970) for a confirmation of the date of the original correspondence, concerning which there was previously some confusion, apparently started by Hutcheson himself.

Bibliography Burnet, Gilbert and Francis Hutcheson (1735). Letters concerning the True Foundation of Virtue or Moral Goodness. London. Butler, Joseph (1726, 1729). Fifteen Sermons preached at the Rolls Chapel. London. Quotations from the 1729 edition; page references from David E. White, ed. (2006). The Works of Bishop Butler. Rochester, NY: University of Rochester Press. ——— (1736). The Analogy of Religion, Natural and Revealed, to the Constitution and Course of Nature. London. Quotations from the second (corrected) 1736 edition; page references from David E. White, ed. (2006). The Works of Bishop Butler. Rochester, NY: University of Rochester Press. Clarke, John (1725). An Examination of the Notion of Moral Good and Evil. London. ——— (1726). The Foundation of Morality in Theory and Practice. York. Clarke, Samuel (1705). A Demonstration of the Being and Attributes of God. London. ——— (1706). A Discourse Concerning the Unchangeable Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation. London. ——— (1716). A Discourse Concerning the Being and Attributes of God, the Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation, 4th edition. London. Clayton, Robert (1751). An Essay on Spirit. London. ——— (1754). Some Thoughts on Self-Love, Innate-Ideas, Free-Will, Taste, Sentiment, Liberty, and Necessity, &c. London. Colman, John (1983). John Locke’s Moral Philosophy. Edinburgh: Edinburgh University Press. Cumberland, Richard (1672). De Legibus Naturae. London. Darwall, Stephen (1995). The British Moralists and the Internal ‘Ought’: 1640– 1740. Cambridge: Cambridge University Press. Gert, Bernard (1967). ‘Hobbes and Psychological Egoism’. The Journal of the History of Ideas 28(4), pp. 503–20. ——— (2006). ‘Hobbes’s Psychology’. In The Cambridge Companion to Hobbes. Ed. by Tom Sorell. Cambridge: Cambridge University Press, pp. 157–74. Harris, James (2009). ‘Of Hobbes and Hume: A  Review of Paul Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion’. Philosophical Books 50(1), pp. 38–46.

48  Part I Hobbes, Thomas (1650a). Humane Nature: Or, the Fundamental Elements of Policy. London. ——— (1650b). De Corpore Politico: Or, the Elements of Law. London. Hutcheson, Francis (1725). An Inquiry into the Original of our Ideas of Beauty and Virtue; in Two Treatises. London. Quotations and page references from Wolfgang Leidhold, ed. (2004). An Inquiry into the Original of our Ideas of Beauty and Virtue in Two Treatises, by Francis Hutcheson. Indianapolis: Liberty Fund. ——— (1728, 1742). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense. London. Quotations and page references from Aaron Garrett, ed. (2002). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense, by Francis Hutcheson. Indianapolis: Liberty Fund. Locke, John (1676). ‘Pleasure, Pain, the Passions’. Quotations and page references from Mark Goldie, ed. (2000). Locke: Political Essays. Cambridge: Cambridge University Press, pp. 237–45. ——— (c. 1686–8). ‘Of Ethic in General’. Quotations and page references from Mark Goldie, ed. (2000). Locke: Political Essays. Cambridge: Cambridge University Press, pp. 297–304. ——— (1689). Two Treatises on Government. London. ——— (1690). An Essay concerning Humane Understanding. London. Quotations and page references from Peter H. Nidditch, ed. (1975). An Essay concerning Human Understanding, by John Locke. Oxford: Clarendon Press. Mandeville, Bernard (1705). The Grumbling Hive: Or, Knaves Turned Honest. London. Quotations and page references from F. B. Kaye, ed. (1924). The Fable of the Bees, by Bernard Mandeville. Oxford: Clarendon Press. ——— (1714, 1723). The Fable of the Bees: Or, Private Vices, Publick Benefits. London. Quotations and page references from F. B. Kaye, ed. (1924). The Fable of the Bees, by Bernard Mandeville, Volume 1. Oxford: Clarendon Press. ——— (1729). The Fable of the Bees, Part II. London. Quotations and page references from F. B. Kaye, ed. (1924). The Fable of the Bees, by Bernard Mandeville, Volume 2. Oxford: Clarendon Press. ——— (1732). An Enquiry into the Origin of Honour. London. Peach, Bernard (1970). ‘The Correspondence Between Francis Hutcheson and Gilbert Burnet: The Problem of the Date’. The Journal of the History of Philosophy 8(1), pp. 87–91. Russell, Paul (2008). The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion. Oxford: Oxford University Press. Shaftesbury, Earl of [Anthony Ashley Cooper] (1711, 1714). Characteristicks of Men, Manners, Opinions, Times. Quotations from the 1714 edition; page references from Lawrence E. Klein, ed. (2000). Characteristics of Men, Manners, Opinions, Times. Cambridge: Cambridge University Press. ——— (1716). Several Letters Written by a Noble Lord to a Young Man at the University. London. Sheridan, Patricia (2007). ‘Pirates, Kings and Reasons to Act: Moral Motivation and the Role of Sanctions in Locke’s Moral Theory’. The Canadian Journal of Philosophy 37(1), pp. 35–48. Tolonen, Mikko (2013). Mandeville and Hume: Anatomists of Civil Society. Oxford: Voltaire Foundation. Wollaston, William (1722). The Religion of Nature Delineated.

3 Founded on Pain and Pleasure

In this chapter I  present my argument for the egoist and hedonist interpretation of Treatise Book 2. I begin with Hume’s account of desire and aversion in general (§3.1), before offering some preliminary remarks about how this account combines with other principles in the Treatise to produce a more complete theory of motivation (§3.2). I then proceed to the examination of various motivational indirect passions: pity, generosity, and vanity (§3.3); envy, malice, and ambition (§3.4); and finally benevolence and anger (§3.5). In each case I argue that Hume’s account entails the reduction of these desires to self-love, repeatedly confirming the general conclusion established in §3.1.

3.1.  The Prospect of Pain or Pleasure The distinction between direct and indirect passions, as I said in Chapter 1, was new with Hume. Nevertheless, its basis was essentially a Lockean hedonism: By direct passions I understand such as arise immediately from good or evil, from pain or pleasure. By indirect such as proceed from the same principles, but by the conjunction of other qualities. (T 2.1.1.4, p. 276) In this passage right at the start of Book 2, Hume states plainly that every passion arises from pleasure or pain in some way or other. The difference is that some do so ‘immediately’, while others do so ‘by the conjunction of other qualities’. I will discuss some of these other associative qualities (as I dubbed them in §1.1), and some of the indirect passions on which they depend, later in this chapter. Meanwhile, it is convenient to begin my reconstruction of Hume’s theory of motivation with the simpler case of the direct passions. Hume begins part 3 of Book 2, Of the will and direct passions, with a limited restatement of his hedonism:

50  Part I WE come now to explain the direct passions, or the impressions, which arise immediately from good or evil, from pain or pleasure. Of this kind are, desire and aversion, grief and joy, hope and fear. (T 2.3.1.1, p. 399) As I discussed in §1.3, Hume then turns to the questions of liberty and necessity, the combat of passion and reason, and—with ‘reason’ revealed in such contexts as really ‘calm desires’ in disguise—to the causes of the violent passions. In the course of his argument concerning the combat of passion and reason, Hume’s hedonism is once again evident: ‘’Tis obvious, that when we have the prospect of pain or pleasure from any object, we feel a consequent emotion of aversion or propensity, and are carry’d to avoid or embrace what will give us this uneasiness or satisfaction. . . ’Tis from the prospect of pain or pleasure that the aversion or propensity arises towards any object’ (T 2.3.3.3, p. 414). Following these various related matters, Hume’s official account of the direct passions then comes in the penultimate section of Book 2. Yet again, this section opens with a clear statement of hedonism: ’TIS easy to observe, that the passions, both direct and indirect, are founded on pain and pleasure, and that in order to produce an affection of any kind, ’tis only requisite to present some good or evil. Upon the removal of pain and pleasure there immediately follows a removal of love and hatred, pride and humility, desire and aversion, and of most of our reflective or secondary impressions. (T 2.3.9.1, p. 438) The restriction here to most of our reflective impressions is curious, and Hume doesn’t specify the exceptions that he had in mind. If Hume is here assuming that the three peculiar impressions highlighted in §1.1— namely those of necessary connection, belief, and volition—are reflective impressions, then that might explain the qualification. In any case, nothing here implies that any passions are exempt from this hedonistic basis, and desire and aversion are explicitly affirmed not to be. Strictly speaking, none of these assertions of hedonism necessarily entails egoism, if we allow that Hume could have been assuming—along with Hutcheson—that other people’s pleasure and pain are operative here in addition to our own. On the face of it, however, the egoist interpretation seems much more likely, given the clear echoes of Locke, and the notable absence of any explicit claim or argument to the contrary. Egoism is also strongly suggested by the confidence with which Hume puts forward these hedonist claims, describing them as ‘obvious’ (T  2.3.3.3, p.  414) and ‘easy to observe’ (T 2.3.9.1, p. 438). To many of his readers, religious and sceptical alike, egoistic hedonism was indeed considered obvious. On the other hand, it is not at all obvious that desire and aversion can arise

Founded on Pain and Pleasure  51 from considerations of other people’s pleasure or pain, and this was of course a point that Hutcheson argued for at length. If Hume agreed with him when he wrote these words, it is hard to see why he would not have been clearer on the point. Following the paragraph just quoted, Hume has this to say, in general, about the direct passions: The impressions, which arise from good and evil most naturally, and with the least preparation are the direct passions of desire and aversion, grief and joy, hope and fear, along with volition. The mind by an original instinct tends to unite itself with the good, and to avoid the evil, tho’ they be conceiv’d merely in idea, and be consider’d as to exist in any future period of time. (T 2.3.9.2, p. 438) Though it isn’t named as such, it is most likely that the original instinct in question is the principle of self-love, understood in hedonistic terms. ‘Good’ and ‘evil’ are just synonyms for ‘pleasure’ and ‘pain’ here, as the two paragraphs quoted earlier demonstrate. And, for the reasons given a moment ago, if Hume had thought that other people’s pleasure and pain could trigger these desires directly, he would surely have said so. Furthermore, Hume’s various accounts of individual desires are, as we will see in the course of this chapter, overtly egoist in nature. Hume goes on to state how this principle gives rise to the various direct passions, without any assistance from the associative qualities such as sympathy or comparison: When good is certain or probable, it produces joy. When evil is in the same situation there arises grief or sorrow. When either good or evil is uncertain, it gives rise to fear or hope, according to the degrees of uncertainty on the one side or the other. Desire arises from good consider’d simply, and aversion is deriv’d from evil. The will exerts itself, when either the good or the absence of the evil may be attain’d by any action of the mind or body. (T 2.3.9.5–7, p. 439) Hume’s original list of direct passions at the start of Book 2 also included despair and security, though these are not discussed in this section. They feature briefly in his discussion of the causes of the violent passions (T 2.3.4.7–10, pp. 421–2), but he doesn’t give us precise definitions. Presumably they are the contraries of hope and fear respectively: when some pleasure is unlikely, we feel despair; and when some pain is unlikely, we feel security.

52  Part I I will have more to say about joy, sorrow, hope, and fear in Chapter 6 (mainly §6.2). For now, it may just be noted that all four passions arise from ideas of pleasure and pain. Joy and sorrow, though they may be felt in response to currently experienced pleasures or pains as well as anticipated ones, still depend on our reflecting on that experience. And the same is true of desire and aversion, which arise (as earlier) from the consideration of pleasure and pain; or equivalently (as in the earlier argument concerning the combat of passion and reason) from the prospect of pleasure and pain (T 2.3.3.3, p. 414). This point was also clear at the start of Book 1, when Hume introduced the general category of reflective impressions, saying by way of example that the ‘idea of pleasure and pain, when it returns upon the soul, produces the new impressions of desire and aversion, hope and fear’ (T 1.1.2.1, p. 8; my emphasis).

3.2.  The Chief Spring and Moving Principle For those who read Hume as a psychological egoist in the Treatise, the evidence examined in the previous section has—understandably— seemed pretty decisive; see Darwall (1993, p. 423) and Karlsson (2006, pp.  246–7). In opposition to Darwall’s egoist interpretation, however, Don Garrett has claimed that: [A]lthough Hume clearly holds that prospective pleasure and pain can be sufficient to move the will, he never claims that it is necessary. (Nor does he treat motivation by the prospect of pleasure and pain egoistically—Appendix 2 of An Enquiry concerning the Principle of Morals, entitled ‘Of Self-Love,’ argues directly against psychological egoism). (Garrett 2007, p. 273) In response to Garrett’s parenthetical remark, first, I insist that the nature of Hume’s view in the Treatise cannot be established by pointing to what he says in the moral Enquiry. No one denies that Hume was an anti-egoist in this later work; the question is not whether but when he acquired this Stoical view. To be clear, I am not suggesting Garrett would deploy such an obviously question-begging move in this connection; the previous quotation is taken from the context of a different debate, and Darwall did not claim that Hume later changed his mind. Nevertheless, if anti-egoist interpreters of the Treatise have hitherto drawn much of their evidence from the moral Enquiry, their case must be substantially weakened in the light of my proposals about how Hume’s thought changed on precisely this matter. Garrett’s first remark is more slippery. Strictly speaking, the assertion that ‘Desire arises from good consider’d simply, and aversion is deriv’d from evil’ (T 2.3.9.7, p. 439) can be read as a sufficiency claim, rather

Founded on Pain and Pleasure  53 than a necessity claim, as Garrett maintains. In context, however, the implication of necessity seems to me quite clear. This is after all Hume’s official statement of his position on the origin of desire and aversion; if he thought that these passions could arise from other sources, he would surely have written that ‘Desire arises either from good consider’d simply, or . . . ’. But he does not. Indeed, any disjunctive definition of this sort, if the second disjunct had anti-hedonist implications, would contradict his definition of the direct passions—which include desire and aversion—­as those that arise immediately from good and evil. To complicate matters, however, Hume does in fact contradict his definition in precisely this way, and in the very next paragraph, saying that ‘the direct passions frequently arise from an original instinct, which is perfectly unaccountable’, and that ‘[t]hese passions, properly speaking, produce good and evil, and proceed not from them, like the other affections’ (T 2.3.9.8, p. 439). Garrett, not unreasonably, takes this paragraph at face value, and accordingly tries to interpret the previous definition as consistent with it. I will argue in §4.1, however, that we should not take this paragraph at face value. Unfortunately I  cannot present this argument, or at least not as convincingly as I would like, until we have gone further with our examination of the rest of Book 2. I must ask the reader to give me the benefit of the doubt until then. This one paragraph aside, the most natural reading of the evidence considered so far is that Hume was, when he wrote the Treatise, a straightforward Lockean hedonist and egoist. He maintained that desire and aversion arise from the prospect of (our own) pleasure and pain, and more generally that pleasure and pain are the foundation of all our passions. If we can forget for a moment Hume’s later arguments against this view, or else can remember the simple point that people sometimes change their minds, then there should be nothing inherently unlikely or surprising about this interpretation. On the contrary, given the importance of Locke’s influence on Hume in general, and the fact that the alternatives to egoism in Hume’s day were all bound up with religious views of which he was highly sceptical, this is exactly the view we should expect him to have held. Add to this the strong impression that Mandeville made on him—something for which we will see ample evidence in the next two sections—and the truly surprising fact is not that Hume was ever an egoist, but that he ever abandoned this view. In addition to his alternative interpretations of the apparent statements of egoism in Book 2, Garrett also offers the following passage from Book 1 as evidence for his view that prospective pleasure and pain are, in the Treatise, merely sufficient rather than necessary for motivation: There is implanted in the human mind a perception of pain and pleasure, as the chief spring and moving principle of all its actions. (T 1.3.10.2, p. 118)

54  Part I He might also have quoted this very similar claim from Book 3: The chief spring or actuating principle of the human mind is pleasure or pain; and when these sensations are remov’d, both from our thought and feeling, we are, in a great measure, incapable of passion or action, of desire or volition. (T 3.3.1.2, p. 574) To my mind, the echo of Locke rings loud and clear in these passages (recall §2.1). Where Locke maintained that the perception of pleasure and pain was the sole moving principle, however, Hume here describes it as the chief spring, implying that there are others. Garrett takes this to mean that these other principles can move us without the aid of self-love. But this reading is not inevitable. The alternative is that self-love sometimes moves us with the aid of other principles. Hume doesn’t specify, either in Book 1 or in Book 3, which other principles he had in mind here. I  suggest that he was thinking of precisely those ‘other qualities’ that are responsible, in conjunction with pleasure and pain, for the generation of the indirect passions: the associative qualities such as the double relation of impressions and ideas, sympathy, and comparison. As the argument of this chapter proceeds, I will explain how several of these associative qualities combine with self-love to produce various particular desires and aversions. What these other principles explain, however, is why the ends of these desires and aversions please or displease us; self-love is always necessary to generate the actual motive out of this hedonistic prospect. This analysis, I submit, makes much better sense of Hume’s claim that pleasure and pain are the chief moving principle; on Garrett’s reading, it is simply one principle among many, leaving it very unclear why Hume should have placed it in charge. In the particular case of sympathy, Hume claims not only that this principle directs our self-love, but also that it augments all of our passions in general, adding further motivational force to the fundamental hedonic principle: We can form no wish which has not a reference to society. A perfect solitude is, perhaps, the greatest punishment we can suffer. Every pleasure languishes when enjoy’d a-part from company, and every pain becomes more cruel and intolerable. Whatever other passions we may be actuated by; pride, ambition, avarice, curiosity, revenge or lust; the soul or animating principle of them all is sympathy; nor wou’d they have any force, were we to abstract entirely from the thoughts and sentiments of others. (T 2.2.5.15, p. 363) To say these passions would not have any force without sympathy is surely an overstatement, and I assume Hume would have retracted it if

Founded on Pain and Pleasure  55 challenged. Still, bearing in mind passages like these, it is not difficult to see why he might have refrained from describing self-love as the only moving principle in the two passages quoted earlier, even if he thought (as he clearly said elsewhere) that the prospect of pleasure and pain is the cause of all desire.

3.3.  A Communication of Sentiments It would be an understatement to say that Hume was intrigued by the principle of sympathy in Books 2 and 3 of the Treatise; one might rather say that he was obsessed. ‘No quality in human nature is more remarkable’, he wrote when first introducing it, ‘both in itself and in its consequences’ (T 2.1.11.2, p. 316). Following this introduction, it features in the explanations of many more phenomena than I have the space (or the need) to go into here. I will focus on just the two most pertinent, namely vanity and compassion. Hume’s explanation of sympathy divides into two stages. First, we form an idea of another’s passion, as it might be, by observing ‘those external signs [of it] in the countenance or conversation’ (T  2.1.11.3, p. 317). Next, this idea is converted into the passion itself, by a transfusion of force and vivacity from the ever-present impression of self, and thanks to the relation of ideas between them and us (T 2.1.11.4–5, pp. 317–8). Thus ‘sympathy depends on the relation of objects to ourselves’ (T 2.1.11.15, p. 322). The details of this account raise many questions that I will not go into here, as they are not directly relevant to my argument, and have been discussed extensively by others.1 What matters for now are certain uses that Hume makes of his principle. In his later work, it may be noted, he continued to appeal to the principle on occasion, but his sketches of how it worked were scarcely more detailed than the one I have just given (see P 3.4, p. 18; P 3.7, p. 19). I do not assume that this was because he lost confidence in his associationist reduction; it may simply have been that he thought his readers would find such details tedious.2 Be that as it may, the headline application of sympathy in the Treatise, given in the section in which it is first introduced and unpacked, is to the indirect passion of vanity, the desire for praise or fame. That this is Hume’s first illustration of the principle itself speaks to the Epicurean nature of Book 2, because this was (as we saw in §2.3) the very passion at the heart of Mandeville’s egoistic account of moral motivation. When he wrote the Treatise, Hume himself did not think moral motivation was reducible to vanity (see §3.4 and §3.5). But his interest in this passion, as well as a part of his explanation of it, almost certainly came from this source. The passions of pride and humility, Hume maintains, do not arise only from our own apprehension of our good and bad qualities, but also ‘from praise and blame, from reputation and infamy’ (T 2.1.11.9, p. 320). The

56  Part I love and hatred of others have a bearing on our own sense of self-worth largely because of sympathy, by means of which these foreign passions are communicated to our own minds. In this particular case, however, the transmission is also facilitated by an awareness of a strong partiality in our own favour, and an argument from authority: [N]othing is more natural than for us to embrace the opinions of others . . . both from sympathy, which renders all their sentiments intimately present to us; and from reasoning, which makes us regard their judgment, as a kind of argument for what they affirm. These two principles of authority and sympathy influence almost all our opinions; but must have a peculiar influence, when we judge of our own worth and character. . . . [B]eing conscious of great partiality in our own favour, we are peculiarly pleas’d with any thing, that confirms the good opinion we have of ourselves, and are easily shock’d with whatever opposes it. (T 2.1.11.10, p. 321) Though the appeal to sympathy here is not to be found in Mandeville, the application of the principle of authority was lifted—almost word for word—from the Fable of the Bees, Part II: Nature has given [Creatures] an Instinct, by which every Individual values itself above its real Worth; this in us, I mean, in Man, seems to be accompany’d with a Diffidence, arising from a Consciousness, or at least an Apprehension, that we do over-value ourselves: It is this that makes us so fond of the Approbation, Liking and Assent of others; because they strengthen and confirm us in the good Opinion we have of ourselves. (1729, p. 134) There can be little doubt that Hume had this passage in mind when writing his own account of the phenomenon. In any case, as sympathy and authority make us pleased by the respect of others, and pained by their contempt, so self-love then interests us in seeking this pleasure and avoiding the corresponding pain. Hume’s account of vanity is thus egoistic. The next most important application of sympathy in the Treatise, at least for present purposes, is to the passion of pity or compassion. Hume writes: ’Twill be easy to explain the passion of pity, from the precedent reasoning concerning sympathy. We have a lively idea of every thing related to us. All human creatures are related to us by resemblance. Their persons, therefore, their interests, their passions, their pains and pleasures must strike upon us in a lively manner, and produce an

Founded on Pain and Pleasure  57 emotion similar to the original one; since a lively idea is easily converted into an impression. If this be true in general, it must be more so of affliction and sorrow. These have always a stronger and more lasting influence than any pleasure or enjoyment. (T 2.2.7.2, p. 369) Add to this, that pity depends, in a great measure, on the contiguity, and even sight of the object; which is a proof, that ’tis deriv’d from the imagination. Not to mention that women and children are most subject to pity, as being most guided by that faculty. The same infirmity, which makes them faint at the sight of a naked sword, tho’ in the hands of their best friend, makes them pity extremely those, whom they find in any grief or affliction. (T 2.2.7.4, p. 370) As with vanity, the emphasis on sympathy in this context, and also the associationist reduction of it, are not to be found in Mandeville. But many of the basic ideas resonate strongly with his. For example: This Virtue [i.e. Charity] is often counterfeited by a Passion of ours, call’d Pity or Compassion, which consists in a Fellow-feeling and Condolence for the Misfortunes and Calamities of others: all Mankind are more or less affected with it; but the weakest Minds generally the most. It is raised in us, when the Sufferings and Misery of other Creatures make so forcible an Impression upon us, as to make us uneasy. It comes in either at the Eye or Ear, or both; and the nearer and more violently the Object of Compassion strikes those Senses, the greater Disturbance it causes in us, often to such a Degree as to occasion great Pain and Anxiety. (1723, p. 287) Hume likewise writes that charity—his word for it is ‘benevolence’—is ‘counterfeited’ by pity (T  2.2.7.1, pp.  368–9). His account of benevolence, it should be acknowledged, is much less pessimistic than Mandeville’s account of charity, as we will see in §3.5. Still, the similarities with their accounts of pity, both in word and in substance, are striking. In the first instance, pity is a sympathetically communicated sorrow, ‘an uneasiness  .  .  . arising from the misery of others’ (T  2.2.9.1, p.  381). In the second place, however, it is also ‘a desire of happiness to another, and aversion to his misery’ (T 2.2.9.3, p. 382). This desire arises, straightforwardly enough, through an application of self-love: sympathy makes us pained by the troubles of others, whereupon selflove motivates us to alleviate their suffering. Thus Hume’s explanation of the motivating power of pity, just like that of vanity, is fundamentally egoistic. It is optimistically egoistic, and hence importantly distinct from

58  Part I Hobbes’s pessimistic account of pity as arising ‘from the imagination that the like calamity may befall [the person feeling pity] himselfe’ (Hobbes 1651, p. 43). Indeed, Hume explicitly opposes ‘[t]hose philosophers, who derive this passion from I know not what subtile reflections on the instability of fortune, and our being liable to the same miseries we behold’ (T 2.2.7.4, p. 370). But it is egoistic nonetheless. Contrasting Hume with Hutcheson here helps to reinforce this interpretation. The hedonist in Hutcheson was moved to posit a ‘Publick Sense’, i.e. a ‘Determination to be pleased with the Happiness of others, and to be uneasy at their Misery’ (Hutcheson 1728, p. 17). Functionally, this public sense plays the same role in Hutcheson’s psychological theory that sympathy plays in Hume’s. The anti-egoist in Hutcheson, however, then felt the need to argue at great length that our public desires or affections are not simply the result of self-love applied to the anticipation of the pleasures and pains of this public sense (1728, pp. 23–9). For the egoist, the existence of something functionally equivalent to a public sense or sympathy affords an obvious and straightforward explanation of (superficially) benevolent or disinterested behaviour, and so it is no wonder that Hutcheson should have been so concerned with refuting this possibility. With this in mind, however, the complete absence of any such argument in Hume’s Treatise is very telling. The clear implication of Hume’s discussion of sympathy and pity is that the postulation of any genuinely selfless desires in explaining the relevant phenomena is simply unnecessary. Though the current prominence of the anti-egoist interpretation of the Treatise owes much to Kemp Smith, the interpretation itself goes back further, to an old article by E. B. McGilvary (1903). Kemp Smith explicitly leaned on the arguments in this article (1905, p. 336; 1941, pp. 140– 2), as did at least one other defender of the anti-egoist interpretation at the time (Chapman Sharp 1921, p. 43). McGilvary, for his part, came up with an alternative reading of Hume on the nature of pity, which neither Kemp Smith nor more recent anti-egoist interpreters have explicitly endorsed, but which perhaps ought to be addressed in case the memory of it lingers. He suggested that it was not only other people’s pains and sorrows that are communicated to us by sympathy, but also their desires. These desires, meanwhile, even if they are egoist to begin with, are no longer egoist when transferred to the minds of other people. You want your pleasure, but when this desire is communicated to me it remains a desire for your pleasure, which, when I  have it, is plainly not egoist (McGilvary 1903, pp. 291–4). It seems to me that McGilvary focuses his argument here in exactly the wrong place. He spends a great deal of time defending an interpretation of the sympathy mechanism according to which a sympathetically communicated desire, though it is egoist in the original mind, is not egoist in the mind to which it has been communicated. But this point

Founded on Pain and Pleasure  59 is surely obvious. The substantial claim that needs defending, but for which McGilvary offers literally no support, is that Hume thought of pity as a sympathetically communicated desire, rather than a sympathetically communicated sorrow (which then generates a self-interested desire in the obvious way). If this were so, then Hume’s treatment of pity would indeed be anti-egoist. But it simply isn’t what Hume says. Pity is directed at those who are suffering from ‘grief or affliction’ (T 2.2.7.4, p. 370); it makes us ‘first conceive a lively idea of [another’s] sorrow, and then feel an impression of it’ (T 2.2.7.5, p. 370); it is ‘an uneasiness . . . arising from the misery of others’ (T 2.2.9.1, p. 381); and so on. Not once does Hume say that we are actuated directly by a sympathetically communicated desire. There is no conceptual difficulty with McGilvary’s suggestion; the problem is simply that it is McGilvary’s rather than Hume’s.

3.4.  To Reap a Pleasure From the Comparison Sympathy is not the only principle in the Treatise that combines with self-love to stimulate and direct our desires. Another very important such principle is that of comparison, according to which we ‘always judge more of objects by comparison than from their intrinsic worth and value’ (T 2.2.8.2, p. 372). The details of how this principle works—which are quite involved and not, perhaps, entirely clear—need not concern us here. What matters for my present argument are just the uses that Hume makes of it. As with sympathy, though some of these uses remain in his later work (see P 3.8, p. 19), the detailed attempt at explaining the principle itself is dropped.3 Hume’s main use of comparison is in explaining the passions of malice and envy. In a nutshell, these arise in us when we seem either greater or less in our own eyes, by virtue of a comparison with others: ’Tis evident we must receive a greater or less satisfaction or uneasiness from reflecting on our own condition and circumstances, in proportion as they appear more or less fortunate or unhappy, in proportion to the degrees of riches, and power, and merit, and reputation, which we think ourselves possest of. Now as we seldom judge of objects from their intrinsic value, but form our notions of them from a comparison with other objects; it follows, that according as we observe a greater or less share of happiness or misery in others, we must make an estimate of our own, and feel a consequent pain or pleasure. The misery of another gives us a more lively idea of our happiness, and his happiness of our misery. The former, therefore, produces delight; and the latter uneasiness. (T 2.2.8.8, p. 375)

60  Part I In the first instance, envy and malice are the pain and pleasure that arise from this comparison. In the second instance, however, self-love then operates on these feelings to produce the malicious desire: ‘malice is the unprovok’d desire of producing evil to another, in order to reap a pleasure from the comparison’ (T 2.2.8.12, p. 377). In his treatment of envy and malice Hume is at odds with Hutcheson, who thought malice in this sense was rare and ‘unnatural’ (Hutcheson 1725, p. 105; recall §2.4). At the same time, he is firmly in line with the Epicurean tradition. The idea here goes back to a famous passage from Lucretius, whose poem De Rerum Natura is the best source we have of Epicurus’s ideas: ’Tis sweet, when on the mighty sea the storm winds rouse the main, To watch from shore another toil with all his might in vain: Not that the hurt of others can to us delightful be, But that we like to look on ills from which ourselves are free. Sweet is it too to view in line the mighty strife of war Arrayed across the plains when we from danger stand afar. (Baring 1884, p. 55) Unsurprisingly, there is also a more immediate precedent for Hume’s account of malice in Mandeville: As every Body would be happy, enjoy Pleasure and avoid Pain if he could, so Self-love bids us look on every Creature that seems satisfied, as a Rival in Happiness; and the Satisfaction we have in seeing that Felicity disturb’d, without any Advantage to our selves but what springs from the Pleasure we have in beholding it, is call’d loving Mischief for Mischief’s sake; and the Motive of which that Frailty is the Result, Malice, another Offspring derived from the same Original. (1714, p. 139) The ‘same Original’ here referred to is the passion of envy, which Mandeville had examined in the preceding paragraphs. Mandeville’s account of envy is superficially different from Hume’s; he defines it as a mixture of grief and anger (1714, p. 136). But the differences here do not seem to run very deep. Hume seems to use the word ‘envy’ exclusively for the grief component of this mixture, while reserving ‘malice’—instead of ‘anger’—for the related desire (and also for the related pleasure when the comparison goes in our favour). Furthermore, though Mandeville does not explicitly appeal to principles of association, he nevertheless provided Hume with a phenomenon in this context that those principles might be used to explain. According to Mandeville, ‘[i]f one, who is forced to walk on Foot envies a great Man for keeping a Coach and Six, it will never be with that Violence, or

Founded on Pain and Pleasure  61 give him that Disturbance which it may to a Man, who keeps a Coach himself, but can only afford to drive with four Horses’ (1714, p. 136). Likewise, according to Hume: ’Tis worthy of observation concerning that envy, which arises from a superiority in others, that ’tis not the great disproportion betwixt ourself and another, which produces it; but on the contrary, our proximity. A common soldier bears no such envy to his general as to his sergeant or corporal; nor does an eminent writer meet with so great jealousy in common hackney scriblers, as in authors, that more nearly approach him. (T 2.2.8.13, p. 377) Hume then accounts for this by saying that the dissimilarity cuts off the relation of ideas, thereby either preventing us from making the comparison at all or, failing that, at least diminishing the effects of it. In part 1 of Book 2, Hume refers ahead to his part 2 account of comparison, while deriving ambition from the same origin: Comparison is in every case a sure method of augmenting our esteem of any thing. A rich man feels the felicity of his condition better by opposing it to that of a beggar. But there is a peculiar advantage in power, by the contrast, which is, in a manner, presented to us, betwixt ourselves and the person we command. The comparison is obvious and natural: The imagination finds it in the very subject: The passage of the thought to its conception is smooth and easy. And that this circumstance has a considerable effect in augmenting its influence, will appear afterwards in examining the nature of malice and envy. (T 2.1.10.12, pp. 315–6) Though the word ‘ambition’ does not appear in this passage, this is the name that Hume gives to the desire for power.4 And in referring back to this passage later on (in the discussion of pity and malice), the word is explicit: ‘I have observ’d in considering the nature of ambition, that the great feel a double pleasure in authority from the comparison of their own condition with that of their slaves’ (T 2.2.8.14, p. 378). The passions of pity, malice, and envy set up an obvious pattern. Pity is a sympathetically communicated sorrow in the sufferings of another, malice is a comparative joy in another’s misery, while envy is a comparative grief in another’s happiness. The pattern would be completed by a sympathetically communicated pleasure in the good fortunes of others. There is no reason why positive passions should not be sympathetically communicated in just the same way as negative ones. For the most part, Hume does focus on the negative cases, even saying that the effect is more pronounced in the case of ‘affliction and sorrow’, because ‘[t]hese have

62  Part I always a stronger and more lasting influence than any pleasure or enjoyment’ (T 2.2.7.2, p. 369). But he does give at least one example of the happy counterpart to pity: ‘when a person obtains any honourable office, or inherits a great fortune, we. . . [rejoice] for his prosperity’ (T 2.2.7.5, p. 370). He also maintains that we sympathize with ‘the fictitious joy as well as every other passion’ at the theatre (T 2.2.7.3, p. 369), and the communication of a positive feeling is important in accounting for our respect for the rich (T 2.2.5.1–14, pp. 357–62). While there is thus no doubt that Hume believed in the positive counterpart to pity, it is less obvious what name should be given to this passion, since Hume never labels it explicitly. Happily, however, there is only one plausible candidate, namely ‘generosity’. This word features alongside pity, malice, and envy in Hume’s original list of indirect passions (T  2.1.1.4, p.  277), and again in a later list of positive passions related to love (T 2.1.4.3, p. 283). The implicit idea, then, is that generosity motivates acts of kindness to others grounded in the pleasure we receive from the sympathetic communication of their happiness. Just like all the other passions examined so far, therefore, its motivational force has an egoistic basis. This understanding of generosity also helps to refute another old argument from McGilvary in favour of the anti-egoist interpretation of the Treatise. Hume’s account of moral motivation in Book 3 is not egoist in the previously familiar ways: he does not appeal to the prospect of rewards and punishments in the afterlife, like Locke, nor does he reduce our (seemingly) virtuous motives to vanity, as Mandeville does. But his account still depends on self-love. The original motive to justice, first of all, is for Hume essentially an enlightened self-interest, when we realize that a mutual respect of property is the best way for all of us to get (and trust that we may keep) what we want. Explicitly, he attributes it to a combination of ‘selfishness and limited generosity’ (T 3.2.2.16, p. 494). The dependence of the former on self-love is obvious; the dependence of the latter on self-love is established by the argument I  gave in §3.3 concerning pity, and the observation just made about the most likely interpretation of ‘generosity’ as likewise being derived from sympathy. McGilvary glosses Hume’s account of the motive to justice as depending on selfishness and limited benevolence, interpreting the latter as an antiegoist desire (1903, pp. 294–7). In fact, however, Hume almost always refers to this second component of his account as ‘generosity’ rather than ‘benevolence’ (T 3.2.2.6, p. 487; T 3.2.2.16, p. 494; T 3.2.2.18, p. 495; T 3.2.2.20, p. 496; T 3.2.2.24, p. 498; T 3.2.5.8, p. 519). I take it that this was intentional. The motive to justice needs to apply to strangers as well as friends, and to this end sympathy is the only thing that can do the job; benevolence, as we will see in the next section, is by definition directed only towards our friends. Admittedly Hume’s use of these terms is not entirely consistent, and there are two places where he reduces the

Founded on Pain and Pleasure  63 motive in question to ‘benevolence’ rather than ‘generosity’ (T 3.2.2.13, p. 492; T 3.2.2.20, p. 496). In the first of these cases, however, Hume’s precise words are ‘[b]enevolence to strangers’, confirming that—strictly speaking—what he really has in mind is generosity. If the original motive to justice depends on sympathy and self-love, however, there is still the matter of the natural virtues, the paradigm being benevolence itself. While Hume opposed Hutcheson by insisting that justice is an artificial virtue, his treatment of the natural virtues, at least, is much closer to Hutcheson’s. For both writers, in particular, these virtues are their own motives. And so we come to the question of whether benevolence itself had an egoist or an anti-egoist basis in Hume’s Book 2 account of motivation.

3.5.  Benevolence Is an Original Pleasure Whereas pity and malice arise, as we have seen, through sympathy and comparison, benevolence and anger are produced by love and hatred: Love is always follow’d by a desire of the happiness of the person belov’d, and an aversion to his misery: As hatred produces a desire of the misery and an aversion to the happiness of the person hated. (T 2.2.6.3, p. 367) Benevolence and anger are therefore directed only towards those we love or hate (our friends or enemies), as I noted of the former passion at the end of the previous section. Pity and malice, by contrast, may be directed towards anyone: ‘Pity is a concern for, and malice a joy in the misery of others, without any friendship or enmity to occasion this concern or joy’ (T 2.2.7.1, p. 369; last emphasis mine). Having pointed out the ‘conjunction of this desire and aversion with love and hatred’, Hume proceeds to consider ‘two different hypotheses’ that may account for it (T  2.2.6.4, p.  367). The first is that love and hatred are themselves desires of the relevant kind; the second is that ‘benevolence and anger are passions different from love and hatred, and only conjoin’d with them, by the original constitution of the mind’ (T 2.2.6.6, p. 368). He argues for the latter hypothesis, on the grounds that ‘these desires arise only upon the ideas of the happiness or misery of our friend or enemy being presented by the imagination’, whereas love and hatred ‘may express themselves in a hundred ways, and may subsist a considerable time, without our reflecting on the happiness or misery of their objects’ (ibid.). There are echoes of Hutcheson here, but with notable differences. Hutcheson divides love into ‘Love of Complacence or Esteem, and Love of Benevolence’, and hatred into ‘Hatred of Displicence or Contempt, and Hatred of Malice’ (1725, p. 102). Hume’s term for the last of these is

64  Part I ‘anger’ rather than ‘malice’ (his account of ‘malice’ chiming instead with Mandeville’s use of that word), but the basic quartet of passions is essentially the same. In insisting on his second hypothesis about benevolence and anger, however, Hume is distancing himself from Hutcheson. Love and hatred, for Hume, are the love of esteem and hatred of contempt only, and their connection with the corresponding desires is simply a contingent fact of human nature. ‘I see no contradiction’, he writes, ‘in supposing a desire of producing misery annex’d to love, and of happiness to hatred’ (T 2.2.6.6, p. 368). For present purposes, a different pair of conflicting hypotheses regarding benevolence and anger is pertinent. The first is that these are primary desires for the happiness or misery of others, independently of any pleasure we may derive from those ends. The second is that they are secondary desires for the happiness or misery of others, desires that arise because the happiness of our friends and the misery of our enemies are sources of pleasure. According to the first hypothesis, benevolence and anger are genuinely disinterested desires; according to the second, their motivational force depends, like the other desires examined in this chapter, on the principle of self-love. Anti-egoist interpreters of the Treatise have favoured the first hypothesis, and with some reason. Benevolence and anger are certainly the most plausible candidates for genuinely antiegoist desires in the Treatise, and we will see at the start of the next chapter that there is some direct textual evidence in support of this view (in the paragraph whose examination I already postponed in §3.2). Before turning to the assessment of this evidence, however, I will consider and dismiss two other possible arguments for the anti-egoist interpretation of these passions, in each case offering a corresponding and much more compelling reason for favouring the alternative. The first specious argument for interpreting benevolence and anger as genuinely anti-egoist desires in the Treatise is based on Hume’s definitions of them as desires for the happiness of our friends or misery of our enemies, and not as desires for the pleasure that these things bring. While superficially suggestive of the anti-egoist hypothesis, these definitions really establish nothing. The happiness of our friends is the immediate end of benevolence, and it is by virtue of having this immediate end that benevolence distinguishes itself from every other desire. It does not follow that this is its ultimate end, however, the end that fuels our desire. On the contrary, given Hume’s general account of desire, he must—if he is to be consistent—suppose that the ultimate end of benevolence is the pleasure of knowing that our friends are happy. The case is the same, after all, with every other desire examined in this chapter. Vanity is the desire for praise, whose ultimate end is the pleasure praise gives us through sympathy. Pity is the desire for the happiness of another (not necessarily a friend), whose ultimate end is the pleasure—or at least the absence of the pain—that is communicated to us from them, again through sympathy.

Founded on Pain and Pleasure  65 Malice is the desire for the misery of another, whose ultimate end is the pleasure we reap from comparing their situation with our own. Nothing in Hume’s treatment of benevolence and anger in T  2.2.6 entails that these passions are exceptional in this regard. Everything in Hume’s general account of desire, meanwhile, entails that they are not. The second specious argument is based on the fact that benevolence and anger are joined to love and hatred by an original instinct, which invites the thought that they are themselves original desires. But there is no reason why an original instinct must necessarily result in an original desire. It might instead give rise to an original pleasure (as it might be, a pleasure in the happiness of our friends), which pleasure then grounds the corresponding self-interested desire. In T 2.2.6 itself, there seems to be nothing to settle this question either way. We need not trouble ourselves long with speculation, however, for happily Hume himself settled the matter decisively in the text, in the later section on the mixture of benevolence and anger with pity and malice: [B]enevolence is an original pleasure arising from the pleasure of the person belov’d, and a pain proceeding from his pain: From which correspondence of impressions there arises a subsequent desire of his pleasure, and aversion to his pain. (T 2.2.9.15, p. 387) If the argument from consistency with Hume’s general account of desire doesn’t establish my second hypothesis, this explicit endorsement of it from Hume’s own pen surely does. To the best of my knowledge, antiegoist interpreters of the Treatise have hitherto overlooked this passage, and it is difficult to see how they can possibly square it with their view. Armed with the knowledge that benevolence is an original pleasure, rather than an original desire, we can also nullify the evidence that antiegoist interpreters have thought to garner from a later passage in Book 2. Following his argument that reason alone cannot motivate us, Hume claims that his opponents have confused reason with ‘certain calm desires and tendencies, which, tho’ they be real passions, produce little emotion in the mind, and are more known by their effects than by the immediate feeling or sensation’ (T 2.3.3.8, p. 417). He goes on: These desires are of two kinds; either certain instincts originally implanted in our natures, such as benevolence and resentment, the love of life, and kindness to children; or the general appetite to good, and aversion to evil, consider’d merely as such. (ibid.) Kemp Smith takes this to imply that benevolence, resentment, the love of life, and kindness to children are primary desires, and hence genuinely

66  Part I disinterested (1941, p. 141). It is difficult to make out, in his discussion, any compelling case with regard to the Treatise itself, since as I  have already said Kemp Smith liberally throws in quotations from Hume’s later work to back up his interpretation, without considering the possibility that Hume might have changed his mind. In the later work, as we will see in the next chapter, Hume certainly did believe that original instincts give rise to original and disinterested desires. In the Treatise, however, the balance of evidence—both from his general remarks about desire and from his particular claims about benevolence—shows that he thought they give rise to original pleasures, and that the desires then arise secondarily, from the application of self-love. Notice also the clear Lockean precedent for this view, in a passage I already quoted in §2.1: ‘[N]ature for wise ends of her own has made us so that we are delighted with the very being of our children. Some wise minds are of a nobler constitution, having pleasure in the very being and the happiness of their friends’ (Locke, 1676, p. 238). Hume contrasts these calm instinctive desires with ‘certain violent emotions of the same kind’ (T 2.3.3.9, p. 417), using resentment in its violent form as an example. What he says here, it must be admitted, has anti-hedonist implications: When I receive any injury from another, I often feel a violent passion of resentment, which makes me desire his evil and punishment, independent of all considerations of pleasure and advantage to myself. (ibid.; my emphasis) However, there is reason to think Hume was overstating when he said that resentment makes us desire the punishment of an enemy independent of all considerations of personal pleasure. For there is still the original pleasure we get from seeing our enemies suffer, and to be consistent Hume would surely have insisted that this was the ultimate driving factor. His real point in this passage, I take it, is that the strong desire for this pleasure takes us over, in the moment, causing us to forget all our other interests. Either way, this claim was most likely intended as a criticism of Hutcheson, who as I have already noted thought human nature was scarcely capable of such ‘malicious disinterested Hatred’ (Hutcheson 1725, p. 105); recall §2.4. Hume was not nearly as pessimistic as Mandeville. Though he shared the Dutchman’s interest in explaining the darker aspects of human nature, his descriptions were never as harsh as Mandeville’s, and moreover he believed there were lighter aspects as well. Benevolence is perhaps the clearest case in point. When we love someone, Hume maintained, this makes us pleased by their happiness, and this innocent pleasure is then what moves us to promote their well-being. We are not looking for any act of kindness from them in return, still less for any heavenly

Founded on Pain and Pleasure  67 reward for our beneficent behaviour. Nor—as in Mandeville’s account of charity—­are we secretly or hypocritically motivated by the desire for praise. Humean benevolence is not reducible to vanity, but rather arises because the happiness of our friends is, given irreducible facts of human nature, a source of pleasure. While showing that he wasn’t a pessimist, however, Hume’s account of benevolence in the Treatise doesn’t show that he wasn’t an egoist; on the contrary, just like his account of every other desire in this early work, it confirms it. Both in the choice of passions discussed and in the accounts given of them, Book 2 of the Treatise is a thoroughly Epicurean work.

Notes 1 See, for example, Árdal (1966, ch. 3), Mercer (1972, chs. 2 and 3), Vitz (2004), Bohlin (2009), Waldow (2012), and Taylor (2015a). Commentators have tended to focus on the second stage of the process; for a discussion of the first, see Pitson (1996). 2 More interesting than the question of whether Hume came to doubt his associationist reduction of sympathy is the question of whether he continued to believe the principle itself was of central importance in the explanation of morals, and to what extent the arguably distinct principle of humanity took its place in the moral Enquiry. For arguments that not much changed, see Abramson (2001), Vitz (2004), and Debes (2007a, 2007b); on the other side, see Taylor (2013, 2015b), Arnold (2018), and Millican (forthcoming). I incline towards thinking there was a significant change here, and suspect that Hume’s conversion to Butler’s motivational pluralism may also have been a factor behind it. Humanity, unlike sympathy, appears in the moral Enquiry to be an original desire for the happiness of others. But the examination of this large matter will have to await another occasion. 3 Comparison has received less attention from commentators than sympathy, which is understandable since Hume himself devotes less time to it. For an examination of both principles in Hume’s Treatise see Postema (2005). Postema combines his discussion of Hume with reference to an impressive number of other key figures, including Mandeville, but Malebranche is notably absent. This lack is made up for by James (2005) and Schmitter (2012). 4 E.g. ‘Strength is a kind of power; and therefore the desire to excel in strength is to be consider’d as an inferior species of ambition’ (T 2.1.8.4, p. 300); ‘If I be void of ambition, power gives me no enjoyment’ (M App2.12, p. 301).

Bibliography Abramson, Kate (2001). ‘Sympathy and the Project of Hume’s Second Enquiry’. Archiv für Geschichte der Philosophie 83(1), pp. 45–80. Árdal, Páll S. (1966). Passion and Value in Hume’s Treatise. Edinburgh: Edinburgh University Press. Arnold, James (2018). David Hume on the Reality of Moral Distinctions: Scepticism and Anti-Scepticism in Hume’s Moral Philosophy. Doctoral Dissertation, King’s College London. Baring, Thomas Charles (1884). The Scheme of Epicurus: A Rendering into ­English Verse of the Unfinished Poem of Lucretius, Entitled ‘De Rerum N ­ atura’. ­London: Kegan Paul, Trench, & Co.

68  Part I Bohlin, Henrik (2009). ‘Sympathy, Understanding, and Hermeneutics in Hume’s Treatise’. Hume Studies 35(1&2), pp. 135–70. Broackes, Justin (2002). ‘Hume, Belief and Personal Identity’. In Reading Hume on Human Understanding. Ed. by Peter Millican. Oxford: Oxford University Press, pp. 187–210. Chapman Sharp, Frank (1921). ‘Hume’s Ethical Theory and Its Critics’. Mind, New Series 30(117), pp. 40–56. Darwall, Stephen (1993). ‘Motive and Obligation in Hume’s Ethics’. Noûs 27(4), pp. 415–48. Debes, Remy (2007a). ‘Has Anything Changed? Hume’s Theory of Association and Sympathy after the Treatise’. The British Journal for the History of Philosophy 15(2), pp. 313–38. ——— (2007b). ‘Humanity, Sympathy and the Puzzle of Hume’s Second Enquiry’. The British Journal for the History of Philosophy 15(1), pp. 27–57. Garrett, Don (2007). ‘The First Motive to Justice: Hume’s Circle Argument Squared’. Hume Studies 33(2), pp. 257–88. Hutcheson, Francis (1725). An Inquiry into the Original of our Ideas of Beauty and Virtue; in Two Treatises. London. Quotations and page references from Wolfgang Leidhold, ed. (2004). An Inquiry into the Original of our Ideas of Beauty and Virtue in Two Treatises, by Francis Hutcheson. Indianapolis: Liberty Fund. ——— (1728, 1742). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense. London. Quotations and page references from Aaron Garrett, ed. (2002). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense, by Francis Hutcheson. Indianapolis: Liberty Fund. James, Susan (2005). ‘Sympathy and Comparison: Two Principles of Human Nature’. In Impressions of Hume. Eds. by Marina Frasca-Spada and P. J. E. Kail. Oxford: Clarendon Press, pp. 107–24. Karlsson, Mikael M. (2006). ‘Reason, Passion, and the Influencing Motives of the Will’. In The Blackwell Guide to Hume’s Treatise. Ed. by Saul Traiger. Oxford: Blackwell, pp. 235–55. Kemp Smith, Norman (1905). ‘The Naturalism of David Hume (II)’. Mind, New Series 14(55), pp. 335–47. ——— (1941). The Philosophy of David Hume. Basingstoke: Palgrave Macmillan. Locke, John (1676). ‘Pleasure, Pain, the Passions’. Quotations and page references from Mark Goldie, ed. (2000). Locke: Political Essays. Cambridge: Cambridge University Press, pp. 237–45. ——— (1690). An Essay concerning Humane Understanding. London. Quotations and page references from Peter H. Nidditch, ed. (1975). An Essay concerning Human Understanding, by John Locke. Oxford: Clarendon Press. Mandeville, Bernard (1714, 1723). The Fable of the Bees: Or, Private Vices, Publick Benefits. London. Quotations and page references from F. B. Kaye, ed. (1924). The Fable of the Bees, by Bernard Mandeville, Volume 1. Oxford: Clarendon Press. ——— (1729). The Fable of the Bees, Part II. London. Quotations and page references from F. B. Kaye, ed. (1924). The Fable of the Bees, by Bernard Mandeville, Volume 2. Oxford: Clarendon Press. McGilvary, E. B. (1903). ‘Altruism in Hume’s Treatise’. The Philosophical ­Review 12(3), pp. 272–98.

Founded on Pain and Pleasure  69 Mercer, Philip (1972). Sympathy and Ethics. Oxford: Clarendon Press. Millican, Peter (2007). ‘Against the “New Hume” ’. In: The New Hume Debate. Eds. by Rupert Read and Kenneth A. Richman. 2nd edition. London: Routledge, pp. 211–77. ——— (forthcoming). ‘The Relation Between Hume’s Two Enquiries’. In Reading Hume on the Principles of Morals. Ed. by Jacqueline Taylor. Oxford: Oxford University Press. Pitson, Tony (1996). ‘Sympathy and Other Selves’. Hume Studies 22(2), pp. 255–72. Postema, Gerald J. (2005). ‘ “Cemented with Diseased Qualities”: Sympathy and Comparison in Hume’s Moral Psychology’. Hume Studies 31(2), pp. 249–98. Schmitter, Amy M. (2012). ‘Family Trees: Sympathy, Comparison, and the Proliferation of the Passions in Hume and his Predecessors’. In Emotion and Cognitive Life in Medieval and Early Modern Philosophy. Eds. by Martin Pickavé and Lisa Shapiro. Oxford: Oxford University Press, pp. 255–78. Taylor, Jacqueline (2013). ‘Hume on the Importance of Humanity’. Revue Internationale de Philosophie 263, pp. 81–97. ——— (2015a). ‘Sympathy, Self, and Others’. In The Cambridge Companion to Hume’s Treatise. Eds. by Donald C. Ainslie and Annemarie Butler. Cambridge: Cambridge University Press, pp. 188–205. ——— (2015b). Reflecting Subjects: Passion, Sympathy, and Society in Hume’s Philosophy. New York: Oxford University Press. Vitz, Rico (2004). ‘Sympathy and Benevolence in Hume’s Moral Psychology’. The Journal of the History of Philosophy 42(3), pp. 261–75. Waldow, Anik (2012). ‘Sympathy and the Mechanics of Character Change’. Hume Studies 38(2), pp. 221–42.

4 A Considerable Adjustment

Having argued that Hume was an egoist in the Treatise, I turn now to chart the emergence of his later anti-egoism. Though the fact of this antiegoism is uncontroversial, there are several important points of detail to establish. First, and perhaps most strikingly, I will argue that Hume’s conversion to Butler’s way of thinking in fact began just before the Treatise was published, and in time to insert one incongruous paragraph anticipating his later view (§4.1). This was the paragraph briefly noted in §3.2 and §3.5; my examination of it here thus completes the argument started in the previous chapter. My second point is that the egoism controversy was, at least after 1740, a matter of considerable importance to Hume, although it has often been side-lined in recent discussions (§4.2 and §4.3). This sizable shift in interests, even if it didn’t coincide with a change of opinion on the matter, would still constitute a significant difference between the Treatise and the later works. Third, I will insist that it was indeed Butler, and not Hutcheson, who was responsible for converting Hume to anti-egoism (§4.4). And finally, by way of setting up the discussion to come in Part II, I will explain how the Dissertation on the Passions differs markedly from Book 2 of the Treatise on this point, confirming that it is not merely a précis of the earlier work (§4.5).

4.1.  Beside Good and Evil . . . Books 1 and 2 of the Treatise were mostly written between 1735 and 1737, during Hume’s stay at the Jesuit college of La Flèche in France. No doubt ideas had been forming before this retreat, and for all we know some parts may even have been drafted; but in any case, when he arrived at London in August of 1737 it was with a complete manuscript and an intent to publish (see Harris 2015, pp. 80, 116). He secured a contract towards the end of the following year, and the two volumes were printed in January 1739. The work did not remain untouched in the intervening 16 months, however, and perhaps the most significant influence on Hume’s thought during this final pre-publication phase was Joseph Butler, whose Analogy of Religion had been published in London in 1736.

A Considerable Adjustment  71 I don’t imagine Hume would have read the Analogy while he was in France, but he cannot have failed to notice it on his return to Britain. He had certainly read it before the year was out, and been suitably impressed, for in December  1737 he wrote thus to his friend Henry Home (later Lord Kames): Your Thoughts & mine agree with Respect to Dr Butler, and I wou’d be glad to be introduc’d to him. I am at present castrating my Work, that is, cutting off its noble Parts, that is, endeavouring it shall give as little Offence as possible; before which I cou’d not pretend to put it into the Drs hands. (NHL, p. 2) At about this time, Kames wrote a letter of introduction to Butler on Hume’s behalf. Hume had hoped to present the letter to Butler before asking his opinion of his own manuscript, but alas the meeting never took place, as we learn from a surviving letter from Hume to Kames in March 1738: I shall not trouble you with any formal compliments or thanks, which would be but an ill return for the kindness you have done me in writing in my behalf, to one you are so little acquainted with as Dr Butler; and I am afraid, stretching the truth in favour of a friend. I have called upon the Doctor, with a design of delivering him your letter, but find he is at present in the country. I am a little anxious to have the Doctor’s opinion. My own I dare not trust to; both because it concerns myself, and because it is so variable, that I know not how to fix it. (HL 1, p. 25) Though Hume failed to get Butler’s opinion, however, the doctor’s ideas still had a direct influence on the final text of Books 1 and 2. If there was not much time for any substantial rewriting, there was at the very least time to remove some of the more overtly anti-religious arguments from Book 1 (the ‘noble Parts’ referred to in the letter to Kames). There was also time to add some passages to Book 2. For example, at the end of T 2.3.5 Hume refers to ‘a late eminent philosopher’ who has observed that ‘custom increases all active habits, but diminishes passive’ (T 2.3.5.5, p.  424). The philosopher in question was almost certainly Butler, who made this point in the Analogy (1736, p. 83); I do not know of anyone else who had made it. I conjecture that it was also around this time that Hume first read Butler’s Sermons, being prompted to do so by his good opinion of the Analogy. The first edition of the Sermons appeared in 1726, with a second in 1729, so it is possible that Hume had come across them earlier. But I can

72  Part I think of little reason why a set of sermons by a young English clergyman would have struck him at this time as a particularly worthwhile read.1 A  third edition appeared alongside the Analogy in 1736, and I  expect that this is what first brought them to Hume’s attention. If this is right, then we have a plausible explanation for this particularly curious paragraph in the penultimate section of Book 2, the discussion of which I postponed in the previous chapter: Beside good and evil, or in other words, pain and pleasure, the direct passions frequently arise from a natural impulse or instinct, which is perfectly unaccountable. Of this kind is the desire of punishment to our enemies, and of happiness to our friends; hunger, lust, and a few other bodily appetites. These passions, properly speaking, produce good and evil, and proceed not from them, like the other affections. (T 2.3.9.8, p. 439) Though Hume doesn’t mention Butler here explicitly—nor even refer obliquely to any ‘late eminent philosopher’ or the like—this is a clear statement of Butler’s crucial point about the priority of desire over pleasure. As with the observation about active and passive habits mentioned a moment ago, I do not know of anyone else who had made this point. In any case, because I will want to refer to this passage often, it will be convenient to label it ‘the Butler paragraph’. Since the 1737 manuscript of Books 1 and 2 does not survive, we have no way of knowing for certain which passages of the 1739 publication were late additions. In the case of the Butler paragraph, however, there are extremely compelling reasons for thinking that this was one of them. The evidence external to the text is circumstantial, and Hume might have read the Sermons some time earlier; though the balance of probability, I suggest, favours my conjecture. Be that as it may, the evidence internal to the text is considerably weightier, and derives from the fact that what Hume says in this paragraph spectacularly contradicts the rest of the text, in at least five different ways. It is possible, I suppose, that Hume was simply confused. But it seems to me overwhelmingly more likely that this was a last-minute attempt to accommodate Butler’s anti-hedonist insight, within a previously constructed hedonist framework that had no room for it. In the Butler paragraph, Hume commits himself to the existence of desires that satisfy the following five conditions: (i) they are ‘direct passions’; (ii) they ‘arise from a natural impulse or instinct, which is perfectly unaccountable’; (iii) they include benevolence and anger (‘the desire of punishment to our enemies, and of happiness to our friends’); (iv) they also include ‘hunger, lust, and a few other bodily appetites’; and (v) they ‘produce good and evil, and proceed not from them’. Now let us consider how these claims relate to the rest of the Treatise. First, (i) and (v)

A Considerable Adjustment  73 together contradict Hume’s definition of the direct passions, as those that ‘arise immediately from good and evil’ (T  2.1.1.4, p.  276). Second, (i) and (ii) contradict the definition of the passions as a whole as a subset of the secondary or reflective impressions; it is the original impressions or impressions of sensation that are said to arise ‘in the soul originally, from unknown causes’ (T  1.2.1.1, p.  7). Third, (iii) and (v) contradict Hume’s claim that ‘benevolence is an original pleasure arising from the pleasure of the person belov’d’, from which ‘there arises a subsequent desire of his pleasure’ (T 2.2.9.15, p. 387). Fourth, (i) and (iv) contradict the identification of bodily appetites such as ‘thirst or hunger’ as original impressions rather than passions (T 1.1.2.1, p. 8; see also T 2.1.1.1, p. 275). Finally, (v) by itself contradicts Hume’s repeated assertions that ‘the passions’—without exception—‘are founded on pain and pleasure’ (T 2.3.9.1, p. 438; see also T 2.1.1.4, p. 276; T 2.3.1.1, p. 399; T 2.3.3.3, p. 414). I would also add that (iii) and (v) contradict Hume’s claim, in the immediately preceding paragraph, that desire and aversion arise from good and evil considered simply. Given Garrett’s alternative interpretation of that previous paragraph, however, this might seem in the present context like begging the question. I am tempted to say, further, that the classification of benevolence and anger as direct passions, entailed by (i) and (iii), goes against the grain of the rest of the Book, since these passions are discussed in part 2 alongside the other indirect passions, and since, also like those passions, they involve ‘other qualities’ beyond self-love, namely the two original principles connecting love with benevolence and hatred with anger. This point is less forceful, however, since Hume never explicitly says that these are indirect passions. Their absence from the original lists of passions at the start of Book 2, both direct and indirect, is itself rather striking. I suppose it is possible they were initially in the list of indirect passions, and removed at the same time as the Butler paragraph was added; but this is highly speculative. Perhaps it is just coincidence that they aren’t mentioned. Or perhaps their absence is rather a symptom of Hume’s generally Epicurean or Mandevillean outlook when he wrote this work. In any case, five contradictions are ample for my case, without the need to insist on a sixth or a seventh. If the interpretation of a text on some point is controversial, it is of course no good defending one side of the argument by simply dismissing any evidence on the other side as a late addition. Let me be clear, then, that my case for the egoist interpretation of the Treatise in no way depends on my conjecture that the Butler paragraph wasn’t part of Hume’s original manuscript. My case for the egoist interpretation is based on the overwhelming evidence examined in the previous chapter, against which this one isolated paragraph seems to me scant opposition. And my basis for conjecturing that it was a late addition—aside from the favourable (albeit circumstantial) external evidence—is how astonishingly contradictory it

74  Part I is with the rest of the text, on matters not limited to egoism. No doubt anti-egoist interpreters will seek to downplay some of the contradictions just highlighted, by offering different interpretations of the relevant (apparently) egoist passages; but then their argument is with my previous chapter, not with the present section.

4.2.  To Remove This Contradiction Terence Penelhum, exhibiting what I take to be the majority view of the Treatise on the matter of motivation, writes: The glib apposition of ‘from good or evil, from pain or pleasure’ carries a strong and unfortunate suggestion of psychological hedonism. . . . But it is plain from another passage of great importance that Hume is anxious to distance himself from psychological hedonism. He complicates his classification of the passions by saying that ‘the direct passions frequently arise from a natural impulse or instinct, which is perfectly unaccountable’ and lists ‘the desire of punishment to our enemies, and of happiness to our friends; hunger, lust, and a few other bodily appetites’ (T 2.3.9.8, SBN 439). He says that these passions ‘produce good and evil, and proceed not from them.’ (Penelhum 2015, p. 212) In opposition to this common trend, I insist that the Treatise doesn’t carry an unfortunate suggestion of psychological hedonism; it contains clear and repeated assertions of it. And the Butler paragraph doesn’t complicate Hume’s classification of the passions; it contradicts it. My argument for these claims has been given at length in the previous chapter and the previous section, but the evidence can perhaps be summed up or made more striking simply by quoting the most relevant passages once again, side by side and in the order they appear in the text: By direct passions I understand such as arise immediately from good or evil, from pain or pleasure. By indirect such as proceed from the same principles, but by the conjunction of other qualities. . . . WE come now to explain the direct passions, or the impressions, which arise immediately from good or evil, from pain or pleasure. . . . ’Tis obvious, that when we have the prospect of pain or pleasure from any object, we feel a consequent emotion of aversion or propensity, and are carry’d to avoid or embrace what will give us this uneasiness or satisfaction. . . . ’Tis from the prospect of pain or pleasure that the aversion or propensity arises towards any object. . . . THE impressions, which arise from good and evil most naturally, and with the least preparation are the direct passions of desire and aversion, grief and joy, hope and fear, along with volition. . . . Desire arises from

A Considerable Adjustment  75 good consider’d simply, and aversion is deriv’d from evil. The will exerts itself, when either the good or the absence of the evil may be attain’d by any action of the mind or body. . . . Beside good and evil, or in other words, pain and pleasure, the direct passions frequently arise from a natural impulse or instinct, which is perfectly unaccountable.  .  .  . These passions, properly speaking, produce good and evil, and proceed not from them, like the other affections. (T 2.1.1.4, p. 276; T 2.3.1.1, p. 399; T 2.3.3.3, p. 414; T 2.3.9.2, p. 438; T 2.3.9.7–8, p. 439) As this juxtaposition shows, the Butler paragraph constitutes a sudden and wholly unexpected volte-face on Hume’s part. Without warning or even acknowledgement, he is suddenly telling us that his previous definitions are all wrong, and his earlier generalizations are all false. Of course, texts should be interpreted charitably, at least within reason, and so I understand the impetus behind the standard view, which seeks to smooth over the Treatise’s rough edges to present a coherent, polished product. But the reality of the Treatise is that it wasn’t a polished product. It was a first edition, rushed through the press while its author’s ideas were still forming, and with a haste that he later repented ‘a hundred, & a hundred times’ (HL 1, p. 158). There is nothing uncharitable about seeing Hume as someone whose views changed and developed over time. And when we allow that some of these developments were happening even as Books 1 and 2 of the Treatise were going to the press, the published text takes on a considerably more interesting aspect. Glossing over its inconsistencies is no kindness to Hume, and is a great disservice to Hume scholarship, since it robs us of hugely valuable evidence concerning the evolution of his ideas. As for the contradiction itself, the appropriate way to remove it is not by fudging our interpretation of the Treatise, but by looking at what Hume wrote in his later works. As I will now show, all hints of hedonism and egoism are systematically exorcized from these works, while Butler’s motivational pluralism is increasingly evident in their place. The first sign of this change came as soon as 1741, in the essay Of the Dignity of Human Nature (now known as Of the Dignity or Meanness of Human Nature; Hume expanded the title in 1770). In this essay, Hume distinguishes between two ‘sects . . . in the learned world’ (DM 1, p. 80), one consisting of those ‘who are inclined to think favourably of mankind’, the other of those who prefer to ‘give us a mean opinion of our nature’ (DM 2, p. 81). In opposition to this second group, Hume appeals to Butler’s point about the priority of desire over pleasure: In my opinion, there are two things which have led astray those philosophers, that have insisted so much on the selfishness of man. In

76  Part I the first place, they found, that every act of virtue or friendship was attended with a secret pleasure; whence they concluded, that friendship and virtue could not be disinterested. But the fallacy of this is obvious. The virtuous sentiment or passion produces the pleasure, and does not arise from it. I  feel a pleasure in doing good to my friend, because I love him; but do not love him for the sake of that pleasure. (DM 10, pp. 85–6) Since the paragraph just quoted mentions the first of two points against the egoists, it would be remiss to ignore the second, which is a particular objection to Mandeville (at least the early Mandeville). The argument is that vanity (the desire at the heart of Mandeville’s account of moral motivation in the Fable of the Bees) cannot prompt us to virtuous action unless we already have a love of virtue for its own sake: [V]anity is so closely allied to virtue, and to love the fame of laudable actions approaches so near the love of laudable actions for their own sake, that these passions are more capable of mixture, than any other kinds of affection; and it is almost impossible to have the latter without some degree of the former. Accordingly, we find, that this passion for glory is always warped and varied according to the particular taste or disposition of the mind on which it falls. Nero had the same vanity in driving a chariot, that Trajan had in governing the empire with justice and ability. To love the glory of virtuous deeds is a sure proof of the love of virtue. (DM 11, p. 86) Of the Dignity or Meanness of Human Nature is a short essay intended for lighter reading than Hume’s more substantial philosophical texts. In the first three editions of this essay (1741, 1742, and 1748), however, Hume mentions his desire to pursue these matters in greater depth: I may, perhaps, treat more fully of this Subject in some future Essay. In the mean Time, I  shall observe, what has been prov’d beyond Question by several great Moralists of the present Age, that the social Passions are by far the most powerful of any, and that even all the other Passions receive from them their chief Force and Influence. Whoever desires to see this Question treated at large, with the greatest Force of Argument and Eloquence, may consult my Lord Shaftsbury’s Enquiry concerning Virtue. (1741, p. 620) In the fourth edition (1753), this promissory note was deleted. The key event between 1748 and 1753 that prompted the deletion, of course, was

A Considerable Adjustment  77 the publication (in 1751) of the first edition of Hume’s Enquiry concerning the Principles of Morals, the very work that he had promised in the earlier essay. While the Treatise had espoused a Lockean and Mandevillean egoism, Hume changed his mind very soon afterwards. In the 1741 set of Essays, Moral and Political, he was instead agreeing with Butler and Shaftesbury, and explicitly anticipating the extended arguments of the moral Enquiry to this effect. I will turn to the moral Enquiry in the next section. Before this work appeared, however, there were, in 1748 and 1750, two editions of the Enquiry concerning Human Understanding. In section  1 of these two editions, there was a lengthy footnote intended to illustrate the possibility of establishing clear and decisive results in the science of human nature. Two such discoveries were chosen. The first, credited to Hutcheson, was that moral distinctions are founded on sentiment rather than reason; the second, now explicitly credited to Butler for the first time, was the crucial point about the priority of desire over pleasure: It had been usual with Philosophers to divide all the Passions of the Mind into two Classes, the selfish and benevolent, which were suppos’d to stand in constant Opposition and Contrariety; nor was it thought that the latter could ever attain their proper Object but at the Expense of the former . . . Philosophers may now [note: see Butler’s Sermons] perceive the Impropriety of this Division. It has been prov’d, beyond all Controversy, that even the Passions, commonly esteem’d selfish, carry the Mind beyond Self, directly to the Object; that tho’ the Satisfaction of these Passions gives us Enjoyment, yet the Prospect of this Enjoyment is not the Cause of the Passion, but on the contrary the Passion is antecedent to the Enjoyment, and without the former, the latter could never possibly exist; that the Case is precisely the same with the Passions, denominated benevolent, and consequently that a Man is no more interested when he seeks his own Glory than when the Happiness of his Friend is the Object of his Wishes; nor is he any more disinterested when he sacrifices his Ease and Quiet to public Good than when he labours for the Gratification of Avarice and Ambition. Here therefore is a considerable Adjustment in the Boundaries of the Passions, which had been confounded by the Negligence or Inaccuracy of former Philosophers. (E 1.na)2 This note was deleted in the third edition (1756). The obvious explanation for its removal is the same as for the removal of the promissory note in Of the Dignity or Meanness of Human Nature, namely the appearance of the Enquiry concerning the Principles of Morals, which examined both anti-rationalism and anti-egoism in considerably more depth.

78  Part I In the light of these remarks, my conjecture that Hume read Butler’s Sermons only after he had written the Treatise (albeit just before it was published) seems increasingly plausible. For it is hardly likely that someone who already thought Butler was responsible for ‘a considerable Adjustment in the Boundaries of the Passions, which had been confounded by the Negligence or Inaccuracy of former Philosophers’, could have written a large volume on the passions that not only fails to mention this adjustment, but on the contrary seems guilty—except perhaps in one isolated paragraph—of the very negligence or inaccuracy it is supposed to fix. In any case, even if it can be maintained that Hume was already a motivational pluralist when he wrote the Treatise, defenders of this interpretation can hardly deny that the view is extremely well hidden, behind many pronouncements that could easily be mistaken for assertions of egoism. In the later work, in striking contrast, Hume brings Butler’s view explicitly to the foreground of his philosophy, and gives it a ringing endorsement.

4.3.  Without Any Regard to Interest The student who approaches Hume’s moral philosophy today is likely to imagine that rationalism was a more important target for Hume than egoism. For one thing, current moral philosophers seem to be overwhelmingly concerned with Hume’s anti-rationalist arguments, and this bias has perhaps affected Hume scholarship as well.3 For another, the texts themselves may seem to support this view: students are presented with a copy of the Treatise, and a copy of the Enquiries; the former is overtly anti-rationalist from the start, with hardly any hint of antiegoism (indeed, if I am right it is even an egoist work itself), while the latter likewise begins with the debate between the sentimentalists and the rationalists, reserving its anti-egoist arguments for an appendix at the end. This reaction, however forgivable, is quite mistaken. Egoism was at least as important a target for Hume as rationalism, if not more important. To get a more accurate reflection of its significance, we must do several things. First, we must increase the weighting for the anti-egoist work: Hume only ever published one edition of the Treatise, which he later renounced, but proudly published ten editions of the moral Enquiry. Second, we must include the additional material: the deleted note from the early editions of the first Enquiry, and the essay Of the Dignity or Meanness of Human Nature. Last but not least, it must be noted that what we now know as the second appendix of the moral Enquiry, which contains Hume’s most sustained attack on egoism, was relegated to this position only at the very end of his life, for the posthumous 1777 edition. In all other editions—which is to say, in every edition published

A Considerable Adjustment  79 in Hume’s lifetime—it had pride of place at the very start of section 2, Of Benevolence. The first appendix, in contrast, which contains Hume’s anti-rationalist arguments (where section 1 merely introduces and postpones the debate), was always an appendix.4 The facts, therefore, are as follows: After one edition of the Treatise, a publication which the author himself sorely regretted, Hume wrote an essay against egoism, not rationalism, using this as the occasion to promise a fuller examination of morals. In an early footnote in the first Enquiry (deleted after the moral Enquiry appeared), Hutcheson’s anti-rationalism and Butler’s anti-egoism were mentioned side by side. They were side by side again in the moral Enquiry, the second section of which, before this was moved to an appendix, began as follows: There is a principle, supposed to prevail among many, which is utterly incompatible with all virtue or moral sentiment. . . . This principle is, that all benevolence is mere hypocrisy, friendship a cheat, public spirit a farce, fidelity a snare to procure trust and confidence; . . . There is another principle, somewhat resembling the former; which has been much insisted on by philosophers, and has been the foundation of many a fair system; that, whatever affection one may feel, or imagine he feels for others, no passion is, or can be disinterested; that the most generous friendship, however sincere, is a modification of self-love. (M App2.1–2, pp. 295–6) After a number of preliminary objections to this egoist view, Hume once again appeals to Butler’s crucial point in his defence of the alternative: There are bodily wants or appetites, acknowledged by every one, which necessarily precede all sensual enjoyment, and carry us directly to seek possession of the object. Thus, hunger and thirst have eating and drinking for their end; and from the gratification of these primary appetites arises a pleasure, which may become the object of another species of desire or inclination, that is secondary and interested. In the same manner, there are mental passions, by which we are impelled immediately to seek particular objects, such as fame, or power, or vengeance, without any regard to interest; and when these objects are attained, a pleasing enjoyment ensues, as the consequence of our indulged affections. . . . In all these cases, there is a passion, which points immediately to the object, and constitutes it our good or happiness; as there are other secondary passions, which afterwards arise, and pursue it as a part of our happiness, when once it is constituted such by our original affections. (M App2.12, p. 301)

80  Part I Just like Butler, Hume makes clear that our pleasure in the important cases is a consequence of our desires being satisfied, and hence that those desires cannot themselves be hedonistic, on pain of circularity. After the Treatise, Hume apparently still maintained that pity and malice were secondary desires, the upshot of self-love combined with sympathy and comparison respectively. At least, he never explicitly denied this, and his summaries of the Book 2 accounts given in the Dissertation on the Passions, while lacking most of the earlier details, suggest a clear continuity: Compassion . . . is an uneasiness in the sufferings of another. It seems to spring from the intimate and strong conception of his sufferings; and our imagination proceeds by degrees, from the lively idea to the real feeling of another’s misery. (P 3.7, p. 19) The comparison of ourselves with others seems to be the source of envy and malice. The more unhappy another is, the more happy do we ourselves appear in our own conception. (P 3.8, p. 19) The passage from the moral Enquiry partially quoted a moment ago, however, shows a change of mind on all of the other desires discussed in the previous chapter, namely vanity, ambition, benevolence, and anger. These are all now primary and disinterested: Nature must, by the internal frame and constitution of the mind, give an original propensity to fame, ere we can reap any pleasure from that acquisition, or pursue it from motives of self-love, and desire of happiness. If I have no vanity, I take no delight in praise: If I be void of ambition, power gives me no enjoyment: If I be not angry, the punishment of an adversary is totally indifferent to me. . . . Now where is the difficulty in conceiving, that this may likewise be the case with benevolence and friendship, and that, from the original frame of our temper, we may feel a desire of another’s happiness or good, which, by means of that affection, becomes our own good, and is afterwards pursued, from the combined motives of benevolence and self-enjoyment? (M App2.12–3, pp. 301–2) Kemp Smith, who maintained that benevolence and anger were already original desires in the Treatise, casually remarked that in this section of the moral Enquiry ‘Hume adds to the list love of fame or power’ (1941, p. 141). He seems to have missed that the love of fame and the love of power are two distinct passions (namely vanity and ambition).

A Considerable Adjustment  81 More importantly, he ignores the fact that the Treatise had offered egoistic reductions of these two desires based on sympathy and comparison respectively. In the moral Enquiry Hume isn’t simply expanding his account of motivation; he is rewriting it. This being said, the case of vanity is in fact somewhat complicated by the Dissertation on the Passions, for in this work Hume reverts to his Treatise account, apparently forgetting what he had written in the moral Enquiry. In the Treatise, when deriving vanity from the principle of sympathy, Hume was explicit in denying that original instincts played any role: Among these phænomena we may esteem it a very favourable one to our present purpose, that tho’ fame in general be agreeable, yet we receive a much greater satisfaction from the approbation of those, whom we ourselves esteem and approve of, than of those, whom we hate and despise. In like manner we are principally mortify’d with the contempt of persons, upon whose judgment we set some value, and are, in a great measure, indifferent about the opinions of the rest of mankind. But if the mind received from any original instinct a desire of fame, and aversion to infamy, fame and infamy wou’d influence us without distinction; and every opinion, according as it were favourable or unfavourable, wou’d equally excite that desire or aversion. (T 2.1.11.11, p. 231) In the moral Enquiry, as we have just seen, he changed his mind, insisting that the mind does contain ‘an original propensity to fame’ (M App2.12, p. 301). But then in the Dissertation on the Passions, when he was obviously looking at Book 2 again and revising this material, he seems to have forgotten what he wrote in the moral Enquiry, for he reverts to his former denial of the originality of this passion, and repeats the Mandevillean explanation: Our opinions of all kinds are strongly affected by society and sympathy, and it is almost impossible for us to support any principle or sentiment, against the universal consent of every one, with whom we have any friendship or correspondence. But of all our opinions, those, which we form in our own favour; however lofty or presuming; are, at bottom, the frailest, and the most easily shaken by the contradiction and opposition of others. . . . Hence that strong love of fame, with which all mankind are possessed. It is in order to fix and confirm their favourable opinion of themselves, not from any original passion, that they seek the applauses of others. And when a man desires to be praised, it is for the same reason, that a beauty is pleased with surveying herself in a favourable looking-glass, and seeing the reflection of her own charms. (P 2.33, p. 14)

82  Part I It is difficult to escape the conclusion that Hume was simply careless on this point: in the Treatise he explicitly denies that vanity arises from an original instinct (preferring Mandeville’s explanation); in the moral Enquiry he explicitly says that it does (preferring Butler’s story); and in the Dissertation he again explicitly denies it (reverting once more to the Mandevillean account). Fortunately, this oscillation is not a devastating problem, either for Hume’s own mature view or for my hypothesis concerning the development of his thought. We need only remember that a commitment to an original desire for fame is not itself inconsistent with Mandeville’s idea. There is no reason why we should not have both an original desire for praise and a secondary desire for the same, as a self-interested means of bolstering our own pride. Hume doesn’t actually say this; he seems to assume that it must be one or the other, and to oscillate back and forth between the two. Taking into account both the moral Enquiry and the Dissertation, however, and doing our best to find in them a coherent view, we may surmise that this hybrid account was, or at least ought to have been, his considered view. And regarding my developmental story, it suffices that the original desire for praise is an additional feature of Hume’s later thinking, doubtless inspired by Butler, that was nowhere to be found in the Treatise.

4.4.  See Butler’s Sermons I conjectured in §4.1 that what I  have called the Butler paragraph (T 2.3.9.8, p. 439) was directly inspired by Butler. Whether or not this is so, there can be little doubt that Hume’s later motivational pluralism owed much to this source. Not only does Hume’s account intrinsically agree with the substance of Butler’s, but it is Butler that Hume himself credits with the discovery in the footnote subsequently deleted from the first two editions of the first Enquiry (discussed in §4.2). This is one of the very few places where Hume explicitly references his sources. Surprisingly, however, it has been suggested twice before—by Kemp Smith, and more recently by the Nortons—that the main inspiration for Hume’s anti-egoism was in fact Hutcheson. In addition to the direct evidence that Hume got his motivational pluralism from Butler, there is a very simple reason why he can’t have got it from Hutcheson. This is because, as I already explained in Chapter 2, Hutcheson wasn’t a motivational pluralist. Though he was of course vehemently anti-egoist, he nevertheless agreed with the basic hedonist principle that desire is founded on pleasure and pain, rather than the other way around: Our Sense of Pleasure is antecedent to Advantage or Interest, and is the Foundation of it. We do not perceive Pleasure in Objects,

A Considerable Adjustment  83 because it is our Interest to do so; but Objects or Actions are Advantageous, and are pursu’d or undertaken from Interest, because we receive Pleasure from them. Our Perception of Pleasure is necessary, and nothing is Advantageous or naturally Good to us, but what is apt to raise Pleasure mediately, or immediately. Such Objects as we know, either from Experience or Sense, or Reason, to be immediately, or mediately Advantageous, or apt to minister Pleasure, we are said to pursue from Self-Interest, when our Intention is only to enjoy this Pleasure, which they have the Power of exciting. (Hutcheson 1725, p. 86) Where Hutcheson differed from the egoists was simply in insisting that other people’s pleasure and pain were the object of some of our desires as well.5 Why, then, would someone think that Hume got his anti-egoism— hinted at in the Butler paragraph, and developed more fully in his later works—from Hutcheson? Norton and Norton, in their annotations to the Treatise, seem to hold that he did. In their annotation to the Butler paragraph, they make no mention of Butler, claiming instead that ‘Hume’s account is . . . similar to that of Hutcheson, who said that the “Appetites”, unlike the other “Desires and Aversions”, produce good and evil or pleasure and uneasiness’ (2007, vol. 2, p. 877). In support of this highly questionable interpretation of Hutcheson, they quote the following passage from his Essay on the Nature and Conduct of the Passions: But we must here observe an obvious Difference among our Desires, viz. that ‘some of them have a previous, painful, or uneasy Sensation, antecedently to any Opinion of Good in the Object; nay, the Object is often chiefly esteemed good, only for its allaying this Pain or Uneasiness; or if the Object gives also positive Pleasure, yet the uneasy Sensation is previous to, and independent of this Opinion of Good in the Object.’ These Desires we may call Appetites. ‘Other Desires and Aversions necessarily presuppose an Opinion of Good and Evil in their Objects; and the Desires or Aversions, with their concomitant uneasy Sensations, are produced or occasioned by this Opinion or Apprehension.’ (Hutcheson 1728, p. 67) I can only reply that this, though perhaps superficially similar, is nevertheless a quite different point from Butler’s (and Hume’s in the Butler paragraph). Here Hutcheson is simply noting that some desires are for the removal of a present pain (e.g. hunger and thirst, the examples that he goes on to give a little way down the page), while others are for positively

84  Part I pleasant objects. This distinction is perfectly consistent with hedonism. Indeed, Locke himself was master of it: Desire . . . is increased and varied by divers considerations; for instance, when it is in pursuit of a positive good, the first consideration that sets it on work or at least quickens it is the possibility, for we have little desire for what we once conceive impossible. . . . But in the desire of the removal of some present evil it is quite otherwise, for there, the evil causing a constant pain, there is a constant desire to be eased of it, whether it be considered as possible or no. (Locke 1676, pp. 242–3) A more sustained case for the importance of Hutcheson’s influence rather than Butler’s on this matter was made by Kemp Smith (1941). Kemp Smith was generally very keen on emphasizing the importance of Hutcheson’s influence, something that he took to have been with Hume right from the start; it was his overarching hypothesis ‘That Hume, under the Influence of Hutcheson, entered into his Philosophy through the Gateway of Morals’ (1941, p. 12). This hypothesis does not seem to me particularly plausible, but I will not criticize it in detail here, deferring simply to Millican (2016). What I want to take issue with at present is a particular proposal made in the context of this general claim: that the existence of original desires, and consequently the denial of egoism, was one of Hume’s very first philosophical commitments; and, of course, in keeping with the general claim, that Hume got this idea from Hutcheson.6 Kemp Smith claims that Hume distinguished, in this connection, between ‘primary’ and ‘secondary’ desires, and that these terms came from Hutcheson (1941, p. 165). No references are given, and Fieser has objected that Hume himself did not use the terms (1992, p. 9). Fieser was half right: Hume doesn’t use them in the Treatise, but he does use them in the moral Enquiry, in a passage I quoted in §4.3 (M App2.12, p. 301). This is in the context of stating precisely Butler’s anti-egoist point; contrary to Kemp Smith, it has nothing to do with Hutcheson, and it postdates the Treatise by more than ten years. There is no evidence that Hume got the terms from Hutcheson while he was writing his early work. Kemp Smith doesn’t say where Hutcheson uses these terms any more than where Hume does, but my best guess is that he had in mind the following passage from Hutcheson’s Essay: Now since we are capable of Reflection, Memory, Observation, and Reasoning about the distant Tendencies of Objects and Actions, and not confined to things present, there must arise, in consequence of our original Desires, ‘secondary Desires of every thing imagined useful to gratify any of the primary Desires, with strength proportioned to the several original Desires, and the imagined Usefulness,

A Considerable Adjustment  85 or Necessity, of the advantageous Object.’ Hence it is that as soon as we come to apprehend the Use of Wealth or Power to gratify any of our original Desires, we must also desire them. (Hutcheson 1728, p. 19) Here, however, Hutcheson is once again making a different point. Hutcheson’s original or primary desires are all directly for pleasure or the avoidance of pain (either for ourselves or for others). His secondary desires, meanwhile, are for the means to pleasant ends, where the means themselves may be indifferent or even painful. This is no doubt a perfectly respectable distinction; but it is not Butler’s distinction between the desire for pleasure on the one hand, and desires for objects, independently of the pleasure they bring, on the other.

4.5.  Conformable or Contrary to Passion It is the standard view that, when writing the Dissertation on the Passions, Hume did nothing more than select a few passages from Book 2 of the Treatise, reorder them, and make some minor stylistic changes. My rival position, of course, is that at least some of Hume’s revisions were more substantial. It is true that only a very small proportion of the Dissertation contains entirely new material, and that many of the modifications to the old material are purely stylistic. But though they are small in number, the few additions and non-stylistic changes that Hume did make are enormous in implication. Hume’s account of desire and aversion is an excellent case in point. Let us start by looking again at what Hume wrote in the Treatise concerning the direct passions, including desire and aversion: ’TIS easy to observe, that the passions, both direct and indirect, are founded on pain and pleasure, and that in order to produce an affection of any kind, ’tis only requisite to present some good or evil. . . When good is certain or probable, it produces Joy. When evil is in the same situation there arises Grief or Sorrow. When either good or evil is uncertain, it gives rise to Fear or Hope, according to the degrees of uncertainty on the one side or the other. Desire arises from good consider’d simply, and Aversion is deriv’d from evil. The Will exerts itself, when either the good or the absence of the evil may be attain’d by any action of the mind or body. Beside good and evil, or in other words, pain and pleasure, the direct passions frequently arise from a natural impulse or instinct, which is perfectly unaccountable. Of this kind is the desire of punishment to our enemies, and of happiness to our friends; hunger, lust,

86  Part I and a few other bodily appetites. These passions, properly speaking, produce good and evil, and proceed not from them, like the other affections. (T 2.3.9.1, 2.3.9.5–8, pp. 438–9) Recall, in particular, that ‘good’ and ‘evil’ are used in the Treatise simply as synonyms for ‘pleasure’ and ‘pain’, and thus that the penultimate paragraph just quoted constitutes an endorsement of hedonism; this is an account which, as I have already discussed at length, is then promptly contradicted in the Butler paragraph. With this in mind, let us now take a look at what happened to these paragraphs when Hume came to rewrite them in the Dissertation on the Passions: SOME objects produce immediately an agreeable sensation, by the original structure of our organs, and are thence denominated Good; as others, from their immediate disagreeable sensation, acquire the appellation of Evil. Thus moderate warmth is agreeable and good; excessive heat painful and evil. Some objects again, by being naturally conformable or contrary to passion, excite an agreeable or painful sensation; and are thence called Good or Evil. The punishment of an adversary, by gratifying revenge, is good; the sickness of a companion, by affecting friendship, is evil. . . When good is certain or probable, it produces Joy. When evil is in the same situation, there arises Grief or Sorrow. When either good or evil is uncertain, it gives rise to Fear or Hope, according to the degrees of uncertainty on the one side or the other. Desire arises from good considered simply, and Aversion is derived from evil. The Will exerts itself, when either the good or the absence of the evil may be attained by any action of the mind or body. (P 1.1–2, 4–6, p. 3) To a careless eye, not much has changed. The last three paragraphs are word for word repeats (spot the extra comma), and the second paragraph seems to be a reworded version of the Butler paragraph from the Treatise, moved a little way up the page. But if we look again more closely, we can see that, with a few carefully crafted alterations, Hume has effected a considerable change in meaning. The clues are in the first two paragraphs just quoted. In the first, Hume defines good and evil as pertaining to the objects that give rise to agreeable and disagreeable sensations; as the external causes of pleasure and pain, then, rather than as identical to these feelings. In the second, similarly,

A Considerable Adjustment  87 objects are said to be good or evil, when they excite pleasure or pain (by virtue of satisfying or thwarting a primary desire). This seemingly small terminological change has an enormously important consequence: the very same words that constituted, in the Treatise, an endorsement of hedonistic egoism no longer carry that meaning. In both the Treatise and the Dissertation, Hume maintains that ‘Desire arises from good considered simply, and Aversion is derived from evil’. But where this formerly meant that desire arises from the consideration of pleasure, and aversion from the consideration of pain, it now means that these passions arise from the consideration of objects, objects which excite pleasure and pain in us either ‘by the original structure of our organs’ or ‘by being naturally conformable or contrary to passion’. Notice, moreover, the clear echo of Butler in this last phrase, who talked of pleasure arising from the ‘suitableness between the Object and the Passion’ (Butler 1726, p. 111). The author of the Dissertation was not a hedonist or an egoist. He was, as he had been since 1741, a clear and consistent supporter of Butler’s pluralist motivational psychology. As noted in the introduction, Selby-Bigge dismissed the Dissertation on the Passions long ago as containing nothing more than ‘verbatim extracts from Bk. II of the Treatise, with some trifling verbal alterations’ ( 1975, pp. xx–xxi). As the present example illustrates, this diagnosis is certainly understandable. But this hasty reaction has apparently discouraged commentators from looking more closely at the detail of what Hume says in the earlier and the later works. When we do this, we can see that the differences are in fact considerable. By altering just a few words, Hume changed sides on one of the major debates at the time, abandoning his hedonism in favour of Butler’s pluralist view. Or rather (as we saw earlier in this chapter), the change of sides had already taken place. But the first four paragraphs of the Dissertation, when contrasted with their corresponding Treatise passages, provide a clear and independent corroboration of the developmental story that I have told, and a refutation of the view that there are no substantial differences between these two texts; and all this just on the first page of the later work. And so I conclude Part I of this book, and my first argument for substantial change between Treatise Book 2 and the Four Dissertations. As we will see in Part II, this was by no means the only change that Hume made in his philosophy of emotion. Even by itself, however, I hope it may suffice to motivate a closer look at what Hume had to say concerning the passions following his first examination of them in 1739.

Notes 1 Hume was obviously familiar with Clarke’s Boyle Lectures, and hence potentially also with Butler’s correspondence with Clarke, which was appended to the fourth edition of these lectures (1716). This correspondence was anonymous on Butler’s side, however, so even if Hume had read this exchange and

88  Part I been impressed by Clarke’s antagonist, he presumably would not have known that the unnamed ‘Gentleman from Glocestershire’ was also the author of the Sermons. A more likely route into Hume’s consciousness would be through the approving mention of the Sermons themselves in the preface to Hutcheson’s Essay and Illustrations, first published in 1728; but even this is just a passing remark, noting merely that Butler ‘has done so much Justice to the wise and good Order of our Nature’ (1728, p. 9). Hume’s nephew, who inherited his uncle’s library, had a copy of the 1736 edition of the Analogy, but apparently no edition of the Sermons (Norton & Norton 1996, p. 80). And there is no mention of the Sermons in M. A. Stewart’s essay on Hume’s early intellectual development (Stewart 2005). I take it, therefore, that my conjecture is consistent with the known facts. 2 This note is not reproduced in Selby-Bigge’s edition of the two Enquiries; it can be found, however, in Beauchamp (2000, p.  232) and Millican (2007, p. 177). I am indebted to Peter Millican for drawing my attention to it, and for first suggesting to me the importance of Butler as a source of Hume’s later anti-egoism. 3 In Radcliffe’s Companion to Hume (2008a)—a 528-page volume including eight essays on the passions and morals—the index entry for egoism points to one solitary page. The page in question is not even from one of the essays on the passions or morals, but from Buckle’s introductory article putting Hume in his historical context, in which Hume’s opposition to Hobbes, in agreement with Shaftesbury and Hutcheson, is briefly noted (Buckle 2008, p. 34). Further inspection reveals just two other brief mentions of Hume’s anti-egoism (Cohon 2008, p. 173; Baier 2008, p. 298). This is in contrast to three articles devoted to Hume’s critique of moral rationalism (Brown 2008), the ‘Humean’ theory of motivation (Radcliffe 2008b), and moral noncognitivism (Sturgeon 2008). Norton and Taylor’s Cambridge Companion to Hume (2009) does a little better, mostly thanks to their own contributions. Norton’s, on the foundations of morality in Hume’s Treatise, begins by summarizing the debates leading up to that work, including in particular the debate between the sentimentalists and the egoists (Norton 2009, pp. 271–84). When we come to Hume himself, however, this introduction seems strangely out of place, for that debate has suddenly been replaced with the more parochial debate between Hutcheson and the rationalists (2009, pp. 286–90). Norton’s article focuses exclusively on the Treatise, which focus, being shared by most Hume scholars, presumably helps to explain the corresponding focus on anti-rationalism at the expense of antiegoism. Taylor is the exception that proves the rule: her article spends more time on this aspect of Hume’s moral philosophy (Taylor 2009, pp. 316–8), but then she has long been urging a preference for the moral Enquiry over Book 3 of the Treatise. I suggest she would do well to emphasize this issue even more. Penelhum’s article in the same collection, on Hume’s moral psychology, notes Hume’s anti-egoism just once in passing, 23 pages in (Penelhum 2009, p. 260). 4 This fact, incidentally, explains a curious remnant of the earlier structure in the 1777 edition: in a footnote in the second appendix, Hume writes that ‘we shall have occasion frequently to treat of [the sentiment of benevolence] in the course of this enquiry’ (M App2.n60, p. 298); this is an odd thing to say in an appendix, but of course an entirely natural thing to say right at the start of a work. The remnant suggests that Hume did not fully work out this late rearrangement. 5 The quotation just given is from Hutcheson’s Inquiry (1725), which predated Butler’s Sermons by a year. We can surmise that Hutcheson read the Sermons soon after, for they are mentioned in the preface to his Essay and Illustrations (1728, p.  9). But I  can find no evidence that this later work took on board

A Considerable Adjustment  89 Butler’s crucial point about the priority of desire over pleasure. The passage just quoted was not changed in any subsequent editions. 6 Part 3 of Kemp Smith’s book is titled a ‘Detailed Consideration of [Hume’s] Central Doctrines, Taken in What May Be Presumed to Have Been the Order of Their First Discovery’ (p. 157); following some introductory remarks, original desires are then one of the first things discussed (pp. 163–5).

Bibliography Baier, Annette C. (2008). ‘Enquiry concerning the Principles of Morals: Incomparably the Best?’. In A Companion to Hume. Ed. by Elizabeth S. Radcliffe. Oxford: Blackwell, pp. 293–320. Beauchamp, Tom L. (2000). An Enquiry Concerning Human Understanding, by David Hume. Oxford: Oxford University Press. Brown, Charlotte R. (2008). ‘Hume on Moral Rationalism, Sentimentalism, and Sympathy’. In A Companion to Hume. Ed. by Elizabeth S. Radcliffe. Oxford: Blackwell, pp. 219–39. Buckle, Stephen (2008). ‘Hume in the Enlightenment Tradition’. In A Companion to Hume. Ed. by Elizabeth S. Radcliffe. Oxford: Blackwell, pp. 21–37. Butler, Joseph (1726, 1729, 1736). Fifteen Sermons preached at the Rolls Chapel. London. ——— (1736). The Analogy of Religion, Natural and Revealed, to the Constitution and Course of Nature. London. Clarke, Samuel (1716). A Discourse on the Being and Attributes of God, the Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation, 4th edition. London. Cohon, Rachel (2008). ‘Hume’s Indirect Passions’. In A Companion to Hume. Ed. by Elizabeth S. Radcliffe. Oxford: Blackwell, pp. 159–84. Fieser, James (1992). ‘Hume’s Classification of the Passions and Its Precursors’. Hume Studies 18(1), pp. 1–17. Harris, James (2015). Hume: An Intellectual Biography. Oxford: Oxford University Press. Hutcheson, Francis (1725). An Inquiry into the Original of our Ideas of Beauty and Virtue; in Two Treatises. London. Quotations and page references from Wolfgang Leidhold, ed. (2004). An Inquiry into the Original of our Ideas of Beauty and Virtue in Two Treatises, by Francis Hutcheson. Indianapolis: Liberty Fund. ——— (1728). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense. London. Kemp Smith, Norman (1941). The Philosophy of David Hume. Basingstoke: Palgrave Macmillan. Locke, John (1676). ‘Pleasure, Pain, the Passions’. Quotations and page references from Goldie, Mark, ed. (2000). Locke: Political Essays. Cambridge: Cambridge University Press, pp. 237–45. Millican, Peter, ed. (2007). An Enquiry concerning Human Understanding, by David Hume. Oxford: Oxford University Press. ——— (2016). ‘Hume’s Chief Argument’. In The Oxford Handbook of Hume. Ed. By Paul Russell. Oxford: Oxford University Press, pp. 82–108.

90  Part I Norton, David Fate (2009). ‘The Foundations of Morality in Hume’s Treatise’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 270–310. Norton, David Fate and Mary J. Norton (1996). The David Hume Library. Edinburgh: Edinburgh Bibliographical Society with the National Library of Scotland. ———, eds. (2007). A Treatise of Human Nature. Volume 2: Editorial Material. Oxford: Clarendon Press. Norton, David Fate and Jacqueline Taylor (2009). The Cambridge Companion to Hume. Cambridge: Cambridge University Press. Penelhum, Terrence (2009). ‘Hume’s Moral Psychology’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 238–69. ——— (2015). ‘The Indirect Passions, Myself, and Others’. In The Cambridge Companion to Hume’s Treatise. Eds. by Donald C. Ainslie and Annemarie Butler. Cambridge: Cambridge University Press, pp. 206–29. Radcliffe, Elizabeth S., ed. (2008a). A Companion to Hume. Oxford: Blackwell. ——— (2008b). ‘The Humean Theory of Motivation and Its Critics’. In A Companion to Hume. Oxford: Blackwell, pp. 477–92. Selby-Bigge, L. A. (1975). Enquiries Concerning Human Understanding and Concerning the Principles of Morals. 3rd edition, revised by P. H. Nidditch. Oxford: Oxford University Press. Stewart, M. A. (2005). ‘Hume’s Intellectual Development, 1711–1752’. In Impressions of Hume. Eds. by Marina Frasca-Spada and P. J. E. Kail. Oxford: Clarendon Press, pp. 11–58. Sturgeon, Nicholas L. (2008). ‘Hume’s Metaethics: Is Hume a Moral Noncognitivist?’. In A Companion to Hume. Oxford: Blackwell, pp. 513–28. Taylor, Jacqueline (2009). ‘Hume’s Later Moral Philosophy’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 311–40.

Part II

5 The Religious Passion

Taking my project to be sufficiently motivated by the discussion of Part I, I turn now to the study of Hume’s mature philosophy of emotion as a whole, as presented in the Four Dissertations. I will devote roughly two chapters in turn to each of these dissertations, beginning, in this chapter and the next, with the Natural History of Religion and section 1 of the Dissertation on the Passions. The main substance of these texts is the topic of the next chapter; in this chapter I am concerned with some preliminary dialectical and contextual matters. I begin with an overview of Hume’s philosophy of religion, rejecting the popular framework drawn in terms of the distinction between reasons and causes (§5.1), in favour of an alternative based on the distinction between superstition and true religion (§5.2). The point is of general importance, but is particularly relevant for our understanding of Hume’s argument in the Natural History, as we will see more fully in the next chapter. I then proceed to explicate the idea of a ‘natural history’, with reference to Thomas Hobbes’s views in the Leviathan (§5.3), John Trenchard’s argument in his Natural History of Superstition (§5.4), and Lord Bolingbroke’s essay On the Folly and Presumption of Philosophers, to which the Warburtonians compared Hume’s dissertation (§5.5).

5.1.  Two Questions in Particular It used to be thought that Hume’s interest in religion derived mainly from a desire for notoriety, and that his post-Treatise attacks on Christian orthodoxy were intended chiefly to excite some controversy and to boost book sales (see, for example, Selby-Bigge 1975, pp. xi–xii). We now know better. Hume’s interest in religion was with him from the start, and if his sceptical views were downplayed in the Treatise, this was only from prudence or a desire not to offend. This interest, furthermore, is no small or detachable aspect of his thought; on the contrary, a plausible case can be made for it being one of the driving forces behind his whole philosophy; see Millican (2002, 2016) and Russell (2008).

94  Part II Precisely because of its centrality, however, not to mention the sensitive nature of the topic, there is no single work laying out Hume’s views on religion systematically, and students of this aspect of his philosophy are obliged to garner evidence from several different texts. Keith Yandell lists the Natural History of Religion, the Dialogues concerning Natural Religion, sections  10 and 11 of the first Enquiry (Of Miracles and Of a Particular Providence and of a Future State), and the three essays Of Superstition and Enthusiasm, Of Suicide, and Of the Immortality of the Soul (Yandell 1990, p. 4). John Gaskin also includes—‘less obviously’— the moral Enquiry and the History of England for their moral attacks on religion, and of course the Treatise (Gaskin 2009, pp.  484–5). Gaskin notes that the Treatise ‘does not seem to us much concerned with religion [because] our sensitivities regarding what would constitute an attack on religion are much weaker than those of Hume’s contemporaries’ (ibid.). Those wanting to enhance their sensitivities in this regard should read Russell (2008). Given the scattered nature of this material, it is desirable to have some framework in which to place Hume’s philosophy of religion, some way both of bringing the relevant texts together and of classifying their distinct but complementary contributions to Hume’s overall critique. We owe the standard framework—indeed, the only framework, so far as I  am aware—to Gaskin (1988, 2009). On Gaskin’s view, the key distinction is that between reasons for belief and causes of belief (where the latter may happen to be good reasons, but need not be): Suppose we put the fundamental question thus: Why does anyone believe in God or gods, or cleave to the teachings of such theistic religions as Christianity or Islam? The answer may be given (nonexclusively) in terms of either reasons or causes, and it is under this division that Hume’s examination of religion begins to look like a comprehensive critique rather than a collection of challenging but discrete sections. (2009, p. 485) On this understanding, Hume’s explicit discussions of religion can be divided into those examining the causes of religious belief (primarily the Natural History of Religion, but also perhaps the earlier essay Of Superstition and Enthusiasm), and those criticizing the arguments in support of it (sections 10 and 11 of the first Enquiry, the posthumous essays Of Suicide and Of the Immortality of the Soul, and of course the Dialogues concerning Natural Religion). The distinction between reasons and causes was certainly familiar to Hume, and he even draws it himself in the context of religious belief, in the introduction to the Natural History:

The Religious Passion  95 As every enquiry, which regards religion, is of the utmost importance, there are two questions in particular, which challenge our attention, to wit, that concerning its foundation in reason, and that concerning its origin in human nature. (N 0.1, p. 33) He goes on to restrict his attention in this work to the latter question: ‘What those principles are, which give rise to the original belief, and what those accidents and causes are, which direct its operation, is the subject of our present enquiry’ (ibid.). On the face of it, this is clear support for Gaskin’s framework. As the reader will probably have guessed, however, I have been setting this framework up with a view to knocking it down. The distinction between reasons and causes, between philosophy and psychology, seems to me an unhelpful one to use in carving up Hume’s critique of religion. I do not believe it structured Hume’s own thoughts on the matter, and in the next section I will present an alternative distinction that I  believe did. Immediately, however, what troubles me most about Gaskin’s framework is that it encourages a blinkered interpretation of the Natural History. This work obviously does provide (the main part of) Hume’s account of the causes of religious belief, and so it certainly is a contribution to the study of religious psychology. In fact, Gaskin if anything understates the importance of the psychological context of the Natural History; it seems to me (in keeping with the central claim of this book) that Hume’s account in this first of his four dissertations is essentially an extension of his treatment of the passions, a point I will be pressing at several points in this chapter and the next. Preliminary support for this claim was already given in §1.3, when I  noted the central importance of the passions of hope and fear in Hume’s account, and used this to help explain the rearrangement of material in the Dissertation on the Passions. Additional evidence can be found in the essay Of Superstition and Enthusiasm, in which Hume briefly anticipates the argument of the Natural History by attributing these two ‘corruptions of true religion’ (SE 1, p. 73) to negative and positive passions respectively: The mind of man is subject to certain unaccountable terrors and apprehensions . . . infinite unknown evils are dreaded from unknown agents; and where real objects of terror are wanting, the soul  .  .  . finds imaginary ones, to whose power and malevolence it sets no limits. . . . Weakness, fear, melancholy, together with ignorance, are, therefore, the true sources of Superstition.   But the mind of man is also subject to an unaccountable elevation and presumption . . . the imagination swells with great, but confused conceptions, to which no sublunary beauties or enjoyments can correspond . . . a full range is given to the fancy in the invisible regions

96  Part II or world of spirits, where the soul is at liberty to indulge itself in every imagination, which may best suit its present taste and disposition. . . . Hope, pride, presumption, a warm imagination, together with ignorance, are, therefore, the true sources of Enthusiasm. (SE 2–3, pp. 73–4) But though the emotional causes of belief are the primary concern of the Natural History, it should not therefore be assumed that the question of reasons has been taken off the table. That, unfortunately, appears to be Gaskin’s conclusion: ‘the account is of the causes and conditions that “naturally” produce religion (as, for example, the presence of air and water “naturally” produces rust on iron) without reference to any reasons that can be produced in favour of or against the religion in question’ (Gaskin 2009, p.  483; my emphasis). Gaskin is not alone in this onesided interpretation of Hume’s dissertation; on the contrary, this seems to be quite a common view.1 In opposition to this trend, however, I will urge that the Natural History is a work of psychology and philosophy, containing important parts of Hume’s overall critique of religious belief in addition to the discussion of its causes. Though my focus here is on the Natural History, the point generalizes. Hume was well aware of the distinction between reasons and causes (as we have seen), but his first dissertation is by no means unique among his works in running the discussion of these two things together. I am thinking here particularly of Hume’s famous argument concerning induction, but also—more relevantly in the present context—section 10 of the first Enquiry, Of Miracles. This essay suffers from precisely the opposite problem facing the Natural History: being placed on the side of reasons rather than causes, its part 1 argument tends to overshadow the psychological discussion of part 2. This latter, as well as being relevant to part 1, complements Hume’s account of the causes of religious belief given in the Natural History. In §6.3, I will also point out how the argument of the Natural History complements the argument in Of Miracles. It might be thought that there is a tension in the interpretation of the Natural History I am putting forward here. On the one hand, I insist— perhaps even more forcibly than Gaskin and others—on the psychological dimension of Hume’s discussion, by positioning his argument squarely in the context of his philosophy of emotion. But on the other hand, I argue that this psychological preoccupation does not exclude more paradigmatically philosophical concerns. The tension is superficial, however, for in my view these philosophical concerns are essential to Hume’s treatment of the passions, which was philosophically (and religiously) loaded from the start, although it is only in his later work that these implications are brought out more clearly. In support of my central contention that the Four Dissertations can be seen as Hume’s Enquiry concerning the

The Religious Passion  97 Passions, it may be noted that this is also precisely the pattern we observe when moving from Books 1 and 3 of the Treatise to the two Enquiries.

5.2.  My Veneration for True Religion Having challenged the framework based on the distinction between reasons and causes, I  now propose an alternative distinction in its place. A more helpful dichotomy, I suggest, is that between true religion and its false corruptions, especially superstition (I will touch on enthusiasm as well before the end of this section). These labels were a standard part of seventeenth- and eighteenth-century Christian apologetics, with writers using the former for their own view, and the latter for whatever alternative was the present object of their scorn or ridicule. Hume latched onto this practice, openly criticizing ‘superstition’ on both moral and rational grounds, while ostensibly endorsing ‘true religion’: ‘In proportion to my veneration for true religion’, says Philo in the Dialogues, ‘is my abhorrence of vulgar superstitions’ (D 12.9, p. 219). But for Hume and Philo, of course, the idea of superstition had a much broader application, while the extension of true religion became vanishingly small.2 Whether or not Hume intended Philo’s endorsements of true religion to be sincere, it is very doubtful that his own endorsements were. In any case, given how vague and thin his notion was, it is unclear how much he would be committed to even if these assertions were to be taken at face value. It is possible, however, to place some definite upper and lower bounds on the concept. The least that Hume’s true religion involves, first, is the belief in an intelligent author of nature; or, at the very least, the belief that, as Philo puts it, ‘the cause or causes of order in the universe probably bear some remote analogy to human intelligence’ (D  12.33, p.  227). As an upper bound, meanwhile, what it presumably cannot include is the belief that this author of nature ever intervenes in the natural order once it has been set up. That true religion for Hume precludes all such particular divine intervention is suggested by the main argument he offers in support of it in the Natural History, which is premised precisely on the absence of any such interference: Many theists, even the most zealous and refined, have denied a particular providence, and have asserted, that the Sovereign mind or first principle of all things, having fixed general laws, by which nature is governed, gives free and uninterrupted course to these laws, and disturbs not, at every turn, the settled order of events by particular volitions. From the beautiful connexion, say they, and rigid observance of established rules, we draw the chief argument for theism; and from the same principles are enabled to answer the principal objections against it. (N 6.2, p. 52)

98  Part II Within these two constraints—the belief in invisible, intelligent power, but the denial that it is among the immediate causes of particular events— there remains much room for manoeuvre. Cleanthes’s quite precise statement of (what he takes to be) true religion falls within these bounds: [It] represents us as the workmanship of a Being perfectly good, wise, and powerful; who created us for happiness, and who, having implanted in us immeasurable desires of good, will prolong our existence to all eternity, and will transfer us into an infinite variety of scenes, in order to satisfy those desires, and render our felicity compleat and durable. (D 12.24, p. 224) But, for all I have said so far, true religion might also be consistent with the author of nature being indifferent to good and evil (which Philo maintains to be the most probable hypothesis; D 11.15, p. 212), and with the mortality of the soul and the absence of any future state (as argued for in Of the Immortality of the Soul and section 11 of the first Enquiry respectively). On top of this, there is also the question, prompted by Gaskin’s rather rich interpretation of the notion (1988, p. 188), about the ‘proper office of religion’, which, according to Cleanthes, ‘is to regulate the heart of men, humanise their conduct, infuse the spirit of temperance, order, and obedience’ (D 12.12, p. 220). Should this ‘proper office of religion’ be built into Hume’s concept as well, as Gaskin maintains? It is wrong, I think, to expect a definite answer to the question of how thick or thin Hume’s notion of true religion is. For when Hume is attacking superstition, he is deliberately vague about the alternative; as well he might be, since true religion is in this context the face-saving escape route that he is strategically offering to his orthodox readers. The less he says about it here, the more likely such readers are to align themselves with it, and therefore to acquiesce in his anti-superstitious arguments. And when he is not attacking superstition, meanwhile, but directly considering how much should go into true religion, the concept will necessarily be subject to much give and take, along with the flow of the arguments. This is particularly clear in the Dialogues, which is after all a debate between three characters who all have very different opinions on what counts as true in matters of theology.3 However that may be, the two limitations that we can confidently place on Hume’s true religion—that it includes the belief in (something like) an intelligent author of nature, but one who does not interfere with creation once it is set up—are really all we need to make the concept a useful one. For they enable us to explain both why true religion and superstition are classified together under the broader heading of religious belief and how they chiefly differ from one another. Superstition is like true religion inasmuch as the superstitious also believe in invisible, intelligent power. It is unlike true religion, however, in that the superstitious believe that

The Religious Passion  99 this invisible power is among the immediate causes of certain natural phenomena, and not just (or not even, depending on the superstition) the initial intelligent cause of the whole frame of nature. If we replace Gaskin’s distinction between reasons and causes with this distinction between superstition and true religion, a different way of framing Hume’s discussions of religion emerges that is, I suggest, both less distorting of what he says and closer to his own way of seeing things. On this suggestion, Hume’s discussions of religion can be divided into those explicitly attacking superstition on both moral and intellectual grounds (Of Superstition and Enthusiasm, Enquiry section 10, and the Natural History of Religion),4 and those examining true religion with a view to seeing how much should go into it (Enquiry section 11, the Dialogues). The difference in content between these sets of works is also reflected by a striking difference in style: when discussing true religion, Hume is notably more cautious and guarded, using the dialogue form to distance himself from the debate and from his own controversial opinions; while in the other texts he speaks his mind directly, in his own voice.5 While Hume’s opposition to superstition is clear, his attitude to enthusiasm is harder to gauge. In the essay Of Superstition and Enthusiasm, he subjects the former to his characteristically uncompromising criticism, while the latter ends up shining by comparison.6 In the Natural History itself, meanwhile, enthusiasm hardly features at all, and though the influence of the positive passions is briefly acknowledged, it is the negative ones that are emphasized: ‘Any of the human affections may lead us into the notion of invisible, intelligent power; hope as well as fear, gratitude as well as affliction: But if we examine our own hearts, or observe what passes around us, we shall find, that men are much oftener thrown on their knees by the melancholy than by the agreeable passions’ (N 3.4, p. 42; see also Philo’s similar remarks at D 12.29–30, pp. 255–6). As I noted in §1.4, furthermore, enthusiasm is a corruption of Stoicism, whereas superstition is a corruption of Platonism, and Hume was generally more amenable to the former school of thought than to the latter, certainly on matters pertaining to religion. For a full account of Hume’s attitude to enthusiasm, however, one would also need to examine his History of England, and such an examination is beyond the scope of this book. To return to the main issue, the only direct textual evidence in favour of Gaskin’s framework is the introduction to the Natural History, which on the surface does seem to carve things up in the same way that Gaskin does. But Hume’s apparent setting aside of the question of reasons here was most likely just a device to disarm his orthodox readers. Reading between the lines, we can see that the distinction between superstition and true religion is once again at play: As every enquiry, which regards religion, is of the utmost importance, there are two questions in particular, which challenge our

100  Part II attention, to wit, that concerning its foundation in reason, and that concerning its origin in human nature. Happily, the first question, which is the most important, admits of the most obvious, at least, the clearest solution. The whole frame of nature bespeaks an intelligent author; and no rational enquirer can, after serious reflection, suspend his belief a moment with regard to the primary principles of genuine Theism and Religion. But the other question, concerning the origin of religion in human nature, is exposed to some more difficulty. (N 0.1, p. 33) The discussion that Hume is setting aside here is not in fact the rational basis of religion in general, but only the rational basis of ‘genuine Theism’. And while he allows that the basis of this belief is sound, nothing at all has been said about the reasonableness of superstition. As we will see in the next chapter, his argument touches on the rationality of this corruption of true religion as well as its psychological causes.

5.3.  Unaccountable Terrors and Apprehensions Before turning to the details of Hume’s argument, I have more contextual matters to pursue. In §1.4, I noted the Epicurean nature of Hume’s views on religion, but without going into any details. In my survey of British philosophers in Chapter 2, meanwhile, I mentioned the very important Epicurean figure of Thomas Hobbes, but for want of space said very little about him. Having left the egoism debate behind, I will now take a brief look at some of Hobbes’s remarks concerning religion, in part 1 of his Leviathan (1651). Hobbes’s account provides some important context for Hume’s Natural History of Religion, and discussing it here may partly make up for the two gaps just mentioned. The title of Hume’s Natural History, though there are other possible sources, might well have been conceived with Hobbes in mind. In his Leviathan part 1, chapter  9, Hobbes distinguishes natural from civil history: The Register of Knowledge of Fact is called History. Whereof there be two sorts: one called Natural History; which is the History of such Facts, or Effects of Nature, as have no Dependance on Mans Will; Such as are the Histories of Metalls, Plants, Animals, Regions, and the like. The other, is Civill History; which is the History of the Voluntary Actions of men in Common-wealths. (1651, p. 60) Hobbes doesn’t attempt to give an explicit ‘natural history of religion’, but in chapter 12 he ventures an account of ‘the Naturall seed of Religion’ in man (1651, p. 79), which as we will see is a clear precursor to

The Religious Passion  101 Hume’s own natural history; and since this account is given in part 1, before the discussion of the commonwealth in part 2, it implicitly belongs on the side of natural rather than civil history. It is a part of Hobbes’s account of man as an animal, rather than as a member of a society. Officially, Hobbes reduces the natural seed of religion to four things: ‘Opinion of Ghosts, Ignorance of second [i.e. immediate] causes, Devotion towards what men fear, and Taking of things Casuall for Prognostiques’ (1651, p. 79). The first of these is the most important, however, and receives the most attention; the other three explain, relatively briefly, some of the particulars of our religious practices. As for the belief in ghosts or invisible power, it arises through a combination of curiosity, ignorance, and fear. Early on in part 1, in the section on the imagination, Hobbes briefly attributes ‘the greatest part of the Religion of the Gentiles in time past’ to the ‘ignorance of how to distinguish Dreams, and other strong Fancies, from Vision and Sense’ (1651, p. 18). His point, simply enough, is that dreams and vivid imaginings may pass for reality, giving rise to a belief in all manner of fanciful things. In chapter 12, however, which officially takes up this subject in detail, this one point becomes the conclusion of a longer story. This longer story begins with two features of human nature: curiosity about the causes of the events that concern us, and the need to satisfy this curiosity with something, even when the truth itself is not to be found. Thus men are ‘curious in the search of the causes of their own good and evill fortune’, but when a man ‘cannot assure himselfe of the true causes of things, (for the causes of good and evill fortune for the most part are invisible,) he supposes causes of them, either such as his own fancy suggesteth; or trusteth to the Authority of other men, such as he thinks to be his friends, and wiser than himselfe’ (1651, p. 76). Curiosity then begets fear or anxiety: For being assured that there be causes of all things that have arrived hitherto, or shall arrive hereafter; it is impossible for a man, who continually endeavoureth to secure himselfe against the evill he feares, and procure the good he desireth, not to be in a perpetuall solicitude of the time to come. . . . So that man, which looks too far before him, in the care of future time, hath his heart all the day long, gnawed on by feare of death, poverty, or other calamity; and has no repose, nor pause of his anxiety, but in sleep. (ibid.) Just as curiosity must be satisfied, so too this ‘perpetuall feare . . . must needs have for object something. And therefore when there is nothing to be seen, there is nothing to accuse, either of their good, or evil fortune, but some Power, or Agent Invisible: In which sense perhaps it was, that

102  Part II some of the old Poets said, that the Gods were at first created by humane Feare’ (ibid.). Though fear is not the sole cause of religious belief, for Hobbes, it is certainly the single most important factor. So much so, in fact, that in the earlier section on the passions, he simply defines religion as a kind of fear: ‘Feare of power invisible, feigned by the mind, or imagined from tales publiquely allowed, Religion; not allowed, Superstition. And when the power imagined, is truly such as we imagine, True Religion’ (1651, p. 42). For Hobbes, then, mere belief in invisible power was not enough to make someone religious; the fear of that power was also necessary. But to return to chapter 12, and to the conclusion of the story explaining our belief in—and fear of—these things, Hobbes accounts for our concept of ‘the matter, or substance of the Invisible Agents, so fancyed’ (1651, p. 77) by means of that confusion between imagination and reality already mentioned: ‘they could not by natural cognition, fall upon any other conceipt, but that it was the same with that of the Soule of man; and that the Soule of man, was of the same substance, with that which appeareth in a Dream . . . which, men not knowing that such apparitions are nothing else but creatures of the Fancy, think to be reall, and externall Substances’ (ibid.). In the course of presenting this account, Hobbes makes a point of restricting it to polytheism, to the belief in ‘the many Gods of the Gentiles’ (1651, p.  76). For the belief of ‘one God, Eternall, Infinite, and Omnipotent’, he has a much simpler explanation, namely ‘the desire men have to know the causes of naturall bodies, and their severall vertues, and operations’, satisfied by the conclusion that there must be a first cause: For he that from any effect hee seeth come to passe, should reason to the next and immediate cause thereof, and from thence to the cause of that cause, and plonge himselfe profoundly into the pursuit of causes; shall at last come to this, that there must be (as even the Heathen Philosophers confessed) one First Mover; that is, a First, and an Eternall cause of all things; which is that which men mean by the name of God: And all this without thought of their fortune; the solicitude whereof, both enclines to fear, and hinders them from the search of the causes of other things; and thereby gives occasion of feigning of as many Gods, as there be men that feigne them. (1651, p. 77) In this, as we will see in the next chapter, Hume himself would go much further. Not only did he reject the argument for a necessary first mover, but the argument that he did officially endorse for monotheism—the design argument—is explicitly discounted as an explanation for the origin of the first monotheistic beliefs. Here too it is our passionate rather

The Religious Passion  103 than our rational nature that provides Hume with the main materials for his naturalistic account.

5.4.  The Subject of Our Present Enquiry Another obvious precedent for the title of Hume’s first dissertation is John Trenchard’s Natural History of Superstition (1709). Trenchard, it is fair to say, was no great philosopher. His significance was more as a political writer, and he is best known for his weekly periodicals The Independent Whig (1720–1) and Cato’s Letters (1720–3), both written with fellow Commonwealthman Thomas Gordon. The Natural History of Superstition itself had a clear political agenda behind it (see Day 2010). Nevertheless, the work provides some useful context for understanding Hume’s later essay. Like Hume, Trenchard begins his essay with a brief acknowledgement of the force of the design argument, before proceeding to his main topic, an examination of the natural causes of religious error. The aim is announced in a staggering paragraph-long sentence (with which the essay abounds), which begins and ends as follows: But as there is no perfection in this frail State, nor any excellency without some defect accompanying it, so these noble faculties of the Mind have misled and betrayed us into Superstition . . . which could not have thus happened in all Ages, unless something innate in our Constitutions made us easily to be susceptible of wrong Impressions, subject to Panick Fears, and prone to Superstition and Error, and therefore it is incumbent upon us, first of all to examine into the frame and constitution of our own Bodies, and search into the causes of our Passions and Infirmities, for till we know from what Source or Principle we are so apt to be deceived by others, and by our selves, we can never be capable of true Knowledge, much less of true Religion, which is the perfection of it. (1709, p. 9) There are two main aspects of Trenchard’s naturalistic account, both with clear precursors in Hobbes. The first, touched on only briefly near the start, is our ignorance of the causes of events that concern us, joined with a passion of curiosity that demands to be satisfied. Since ‘the Divine Providence has for the most part hid the Causes of Things which chiefly concern us from our View’, and since ‘it is hard to avoid sollicitude till we think we know them’, we must ‘substitute such in their room, as our own Imaginations or Prejudices suggest to us, or take the Words of others whom we think Wiser than our selves’ (1709, p. 10).

104  Part II The second point is the one that was less significant in Hobbes’s story, but which dominates Trenchard’s essay, namely our inability to clearly distinguish imagination from reality. Trenchard endeavours to describe this pathology in detail (and with very few full-stops); suffice it to say here that he diagnosed religious experience (as we would now call it) as essentially the result of an emotional disorder. And for present purposes, the key word here is ‘emotional’: a natural history of superstition is necessarily a psychological history of superstition, locating the origin of such beliefs, experiences, and practices in neither reason nor divine revelation, but the passions. It is not idly that Trenchard states, at the start of his essay, that his subject requires him to ‘search into the causes of our Passions and Infirmities’.7 And so I suggest that it would have come as no surprise to Hume’s readers to find him first publishing his own Natural History of Religion alongside the Dissertation on the Passions. Quite apart from the particular relevance of the passions of hope and fear in his account, his general project in this first dissertation relates directly to the science of emotion as pursued in the second. With this context in mind, it does not seem to me difficult to appreciate the sense in which Hume’s account constitutes a ‘natural history’ of religion: it is a natural history very much in the spirit of Hobbes and Trenchard, where the emphasis is on the first word, with its emotional and irrational (or at least non-rational) overtones, rather than the second. To be clear, there is no sense in which Hume was attempting to reconstruct a genuine history of religion, in the sense of describing what happened where and when. His subject is not human history, so much as human nature: What those principles are, which give rise to the original belief [of invisible, intelligent power], and what those accidents and causes are, which direct its operation, is the subject of our present enquiry. (N 0.1, p. 33) A study of this kind certainly has historical implications, inasmuch as any theory offered will entail constraints on what could have happened (e.g. the constraint that polytheism necessarily preceded monotheism). In the other direction, too, the evidence of history has a bearing on the study, not only because it might directly falsify one’s theory (by pointing to violations of those constraints), but also because it furnishes basic data for theorizing about. But the relevant data are not exclusively historical, and Hume also appeals in the course of his discussion to facts about the present, and to empirical hypotheses about human psychology (drawn, more often than not, from his theory of the passions); see §6.1 and §6.2 for details. Thus I find myself at odds with Michel Malherbe, who has argued that Hume’s account does not conform to any relevant conception of a natural history from the period (Malherbe 1995). In relation to Trenchard’s

The Religious Passion  105 conception (Hobbes is not mentioned), Malherbe complains that, ‘however empirical and historical the required method may appear, it is quite clear that Hume has not written a natural history’ (1995, p. 262), in part because the discussion is ‘[t]oo anecdotal and too abstract’ (1995, p. 263).8 Alan Bailey and Dan O’Brien, in response to this kind of worry (though not explicitly in response to Malherbe), endorse Dugald Stewart’s interpretation of Hume’s account as an example of conjectural history: ‘Hume . . . should not be seen as a shoddy historian; he should, instead, be seen as a different kind of historian  .  .  . interested in what Dugald Stewart (1753–1828) calls a speculative or conjectural history’ (Bailey & O’Brien 2014, p. 168; see Stewart 1793, p. 293). But this does not seem to me quite right (and Malherbe also rejects this interpretation; 1995, pp. 267–70). It is true that the closest Hume comes to any history in this work is conjectural, but that is not because he is engaged in conjectural history. It is because he is not engaged in history at all. In other words, I  suggest that Hume should not be seen as a shoddy historian in the Natural History because he should not be seen as a historian of any kind; he should be seen as a psychologist, precisely the same light in which he appears in Treatise Book 2 and the Dissertation on the Passions.

5.5.  Its Public Entry Was Rather Obscure In his autobiographical My Own Life, Hume had only this to say about his Four Dissertations: In this interval, I published at London my Natural History of Religion, along with some other small pieces: its public entry was rather obscure, except only that Dr. Hurd wrote a pamphlet against it, with all the illiberal petulance, arrogance, and scurrility, which distinguish the Warburtonian school. This pamphlet gave me some consolation for the otherwise indifferent reception of my performance. (MOL 13, p. xxxvii) For convenience, I will follow Hume in attributing this anonymous pamphlet to Richard Hurd, though William Warburton himself almost certainly had a direct hand in it (as Hume himself privately suspected; HL 1, p. 265). A few years previously, Warburton had challenged—with similar petulance—the ideas of Henry St John, Lord Viscount Bolingbroke (Warburton 1754–5). In the pamphlet against Hume, this ‘puny Dialectician from the North’ is directly compared to Bolingbroke, and described as someone who ‘steps forth into his place; and, with the same beggarly troop of routed sophisms, comes again to the attack’ (Hurd 1757, p. 3). Bolingbroke was a leading Tory politician, and editor of the weekly antiWhig journal The Craftsman (1725–36).9 Before returning to E ­ ngland to pursue his opposition to Robert Walpole’s government, however, he spent ten years in exile in France, where he wrote a handful of philosophical

106  Part II essays. These were circulated among his friends, perhaps most notably Alexander Pope, whose deist Essay on Man (1734) was heavily i­ nfluenced by them. They were not published until 1754, however, three years after his death, and so Hume’s own Natural History was almost certainly written—or at least first drafted—in ignorance of them (in fairness to Hurd and Warburton, they themselves could not have known this). In any case, the essays made little impression on Hume, as we learn from this letter to Abbé Le Blanc, dated October 1754: Lord Bolingbroke’s posthumous Productions have at last convinc’d the whole World, that he ow’d his Character chiefly to his being a man of Quality, & to the Prevalence of Faction. Never were seen so many Volumes, containing so little Variety  & Instruction: so much Arrogance & Declamation. The Clergy are all enrag’d against him; but they have no Reason. Were they never attack’d by more forcible Weapons than his, they might for ever keep Possession of their Authority. (HL 1, p. 208) Hume of course recognized philosophical skill (or its absence) when he saw it, and Bolingbroke, like Trenchard, was no great philosopher. But it is nevertheless worth dwelling briefly on how Bolingbroke’s thoughts on the origin of religion relate to Hume’s, if only to set the latter in context (including the context as his contemporary critics saw it). Bolingbroke’s second essay, On the Folly and Presumption of Philosophers, begins with a general attack on metaphysics, of very much the sort that Hume had previously sought to rebut in the introduction to his first Enquiry (E 1, pp. 5–16). Though Bolingbroke’s attack is sweeping and crude, there is a more modest degree of it that accords somewhat with Hume’s own sentiments, particularly regarding the scepticism appropriate when it comes to some of the more profound religious questions, and the preference for keeping our researches within the bounds of common life (see E 11, pp. 132–48; E 12.25, p. 162).10 In section 2 of his essay, Bolingbroke then proceeds to give his account of ‘the rise and progress of philosophy’ (1754, vol. 1, p. 292), an account which begins with the development of superstition or polytheism, the ‘child of ignorance and fear’ (1754, vol. 1, p. 293). As this last quotation indicates, Bolingbroke’s natural history of superstition accords approximately with the Hobbist story. Polytheism is said to arise from mankind’s ‘curiosity to know the causes of the phaenomena’, especially of ‘those, from which they receive much benefit or much hurt’, together with that principle ‘whereby we make ourselves the measure, as well as the final cause, of all things’ (1754, vol. 1, pp. 293–4): It is this that has represented the unknown causes of the ordinary, as well as extraordinary, of the beneficial, as well as hurtful, phaenomena of nature, to the minds of such savages and demi savages as we

The Religious Passion  107 describe, under the images of animal beings, a little different from man, but analogous to him, and endued only with greater power and greater intelligence. (1754, vol. 1, p. 294) In addition to these familiar tropes, Bolingbroke’s essay included two further points worth mentioning here. The first is an appeal to the principle of ‘flattery’, which leads us to deify ‘public benefactors and heroes and kings’ (1754, vol. 1, p. 295). In this he anticipated Hume’s discussion of hero-worship: ‘where men are affected with strong sentiments of veneration or gratitude for any hero or public benefactor, nothing can be more natural than to convert him into a god, and fill the heavens, after this manner, with continual recruits from among mankind’ (N 5.6, p. 50). The second thing of note in Bolingbroke’s essay is his opposition to the opinion of Cudworth and others, ‘that the unity of God was the original belief of mankind, and that polytheism and idolatry were the corruptions of this orthodox faith’ (1754, vol. 1, pp.  298–9). Bolingbroke allows that there are many arguments ‘that demonstrate, beyond a possibility of doubting, the unity of God to have been acknowledged by the most antient of the idolatrous nations’ (1754, vol. 2, p. 2). But he goes on to deny that they demonstrate ‘that this was the primitive faith of mankind: because we see that the things of this world are in a perpetual rotation, and because in several countries, at several periods, men may have gone from idolatry to true religion, and have fallen from this back again into idolatry’ (1754, vol. 2, pp. 2–3). In this, he anticipated Hume’s opening claim in the Natural History that ‘polytheism or idolatry was, and necessarily must have been, the first and most ancient religion of mankind’ (N  1.1, p.  43), as well as his examination in the middle of the dissertation of the ‘flux and reflux in the human mind’ between idolatry and monotheism (N 8.1, p. 58). Bolingbroke had little to offer by way of argument, however, and certainly nothing to compare to the sophistication of Hume’s case, which I will examine in the next chapter (§6.1). Hume’s Warburtonian critics, in placing him alongside Bolingbroke, missed of course the philosophically superior nature of Hume’s discussion, and also the rhetorical nature of his endorsement of true religion and the design argument (Bolingbroke, for his part, was a sincere deist). But they did not miss the relevance of the philosophy of emotion in Hume’s natural history. On the contrary, this seemed to them to be implied by the title of the work itself: You ask, why he chuses to give it this title. Would not the Moral history of Meteors be full as sensible as the Natural history of Religion? Without doubt. . . . But this great Philosopher is never without his Reasons. It is to insinuate, that what the world calls Religion, of

108  Part II which he undertakes to give the history, is not founded in the Judgment, but in the Passions only. (Hurd 1757, p. 9) At the outset of Book 2 of the Treatise, Hume said that his subject there was the ‘nature, origin, causes, and effects’ of the passions (T 2.1.1.43, p. 276). This work contains a great deal about their nature, origin, and causes, but notably less about their effects. Superstitious belief would have been the obvious effect for an Epicurean Sceptic like Hume to focus on; as with the understanding and morals, however, it was only in his later work on the passions that these irreligious implications were brought out explicitly and dealt with at length. It is to this extension of his treatment of the passions that I now turn.

Notes   1 For example: ‘the Natural History of Religion . . . was essentially a sociological enquiry into the origin of religion’ (Jenkins 1992, p. 8); ‘Hume’s purposes in “The natural history” were not fundamentally critical’ (Falkenstein 2003, p. 1).   2 See Martin Bell (1999) for a discussion of this general practice, and of Hume’s appropriation of it for his own irreligious ends.   3 For example, Demea speaks of ‘true Theists’ at D 4.2 (p. 159), only to have Cleanthes claim immediately afterwards that such as Demea has in mind are really atheists without knowing it (D 4.3, p. 159). Philo once refers to ‘the true system of Theism’ in an argument to the effect that it is inconsistent with Cleanthes’s principles (D 5.2, p. 165).   4 At the end of Enquiry section 10, having argued that it is irrational, Hume in fact endorses the superstitious belief in Christian miracles on the grounds of faith (E 10.40, p. 130). But the endorsement is obviously ironic.   5 The two unpublished essays needn’t be placed in this scheme, precisely because Hume never published them. If pushed, I would place Of Suicide alongside the other attacks on superstition, and Of the Immortality of the Soul with the discussions of true religion. This latter essay breaks the stylistic pattern just noted, since in it Hume argues against (the rationality of) a central tenet of most Christians’ conception of true religion, directly and in his own voice, and without describing the belief as ‘superstitious’. (The transparently ironic endorsement of this tenet on the basis of revelation rather than reason hardly makes up for the scandalous nature of the argument; if anything, it adds to the offence.) Being unpublished, however, it strikes me as the exception that proves the rule: Hume wanted his official examinations of such delicate questions to be presented in much less forthright terms.   6 The first of the three comparisons Hume draws in this essay is that ‘superstition is favourable to priestly power, and enthusiasm not less or rather more contrary to it than sound reason and philosophy’ (SE 5, p. 75). But he was himself no friend to priestly power. The second is that ‘religions, which partake of enthusiasm are, on their first rise, more furious and violent than those which partake of superstition; but in a little time become more gentle and moderate’ (SE 7, p. 76). There is a criticism of enthusiasm in this, certainly, but on the whole the ensuing discussion emphasizes how moderate and harmless the enthusiasts are now, in contrast with the superstitious. The third and

The Religious Passion  109 final comparison is that ‘superstition is an enemy to civil liberty, and enthusiasm a friend to it’ (SE 9, p. 78). Here, even the pretence of criticizing enthusiasm has disappeared altogether.   7 As already quoted earlier, what Trenchard says more fully is that it is incumbent upon him ‘to examine into the frame and constitution of our own Bodies, and search into the causes of our Passions and Infirmities’; and it may be noted that he ventures several hypotheses about the physical causes of these emotional disorders as well, something that Hume himself made a point of avoiding.   8 Another part of Malherbe’s worry is that ‘the inductive process which should be employed, going from facts toward the causes or the principles, is superseded by argumentation which is mainly deductive’ (1995, p. 262). But the claim that Hume’s argumentation in the Natural History is deductive rather than inductive simply baffles me, and I can do little more than flatly deny it; see §6.1.  9 It was this journal, together with Joseph Addison’s Spectator, that Hume claimed as the original models for his own Essays, Moral and Political, though at the same time explicitly stating his intention—in contrast to Bolingbroke— to avoid ‘Party-Rage’, and to handle his subject matter with ‘Moderation and Impartiality’ (Ad1741). For a discussion of Hume’s party-neutral stance in these Essays see Harris (2015, pp. 166–74). 10 For further evidence of the more sophisticated nature of Hume’s philosophy, however, see also the Dialogues concerning Natural Religion. Already in part 1 of these Dialogues, Cleanthes pushes Philo hard on precisely this point: ‘These sceptics, therefore, are obliged, in every question, to consider each particular evidence apart, and proportion their assent to the precise degree of evidence, which occurs. This is their practice in all natural, mathematical, moral, and political science. And why not the same, I  ask, in the theological and religious? Why must conclusions of this nature be alone rejected on the general presumption of the insufficiency of human reason, without any particular discussion of the evidence? Is not such an unequal conduct a plain proof of prejudice and passion?’ (D 1.13, p. 137).

Bibliography Bailey, Alan and Dan O’Brien (2014). Hume’s Critique of Religious Belief: Sick Men’s Dreams. Dordrecht: Springer. Bell, Martin (1999). ‘Hume on Superstition’. In Religion and Hume’s Legacy. Eds. by D. Z. Phillips and Timothy Tessin. London: Palgrave Macmillan, pp. 159–70. Bolingbroke, Viscount [Henry St John] (1754). The Philosophical Works of the Late Right Honorable Henry St. John, in Five Volumes. London: David Malley. Day, Matthew (2010). ‘The Sacred Contagion: John Trenchard, Natural History, and the Effluvial Politics of Religion’. History of Religions 50(2), pp. 144–61. Falkenstein, Lorne (2003). ‘Hume’s Project in “The Natural History of Religion” ’. Religious Studies 39(1), pp. 1–21. Gaskin, J. C. A. (1988). Hume’s Philosophy of Religion. Basingstoke: Palgrave Macmillan. ——— (2009). ‘Hume on Religion’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 480–513.

110  Part II Harris, James (2015). Hume: An Intellectual Biography. Oxford: Oxford University Press. Hobbes, Thomas (1651). Leviathan. London: Andrew Crooke. Quotations and page references from Richard Tuck, ed. (1996). Leviathan, by Thomas Hobbes. Cambridge: Cambridge University Press. Hurd, Richard (1757). Remarks on Mr. David Hume’s Essay on the Natural History of Religion. London: M. Cooper. Jenkins, John J. (1992). Understanding Hume. Edinburgh: Edinburgh University Press. Malherbe, Michel (1995). ‘Hume’s Natural History of Religion’. Hume Studies 21(2), pp. 255–74. Millican, Peter (2002). ‘The Context, Aims, and Structure of Hume’s First Enquiry’. In Reading Hume on Human Understanding. Ed. by Peter Millican. Oxford: Oxford University Press, pp. 27–65. ——— (2016). ‘Hume’s Chief Argument’. In The Oxford Handbook of David Hume. Ed. by Paul Russell. Oxford: Oxford University Press, pp. 82–108. Pope, Alexander (1734). An Essay on Man. London: John Wright. Russell, Paul (2008). The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion. Oxford: Oxford University Press. Selby-Bigge, L. A., ed. (1975). Enquiries concerning Human Understanding and concerning the Principles of Morals. 3rd edition, revised by P. H. Nidditch. Oxford: Oxford University Press. Stewart, Dugald (1793). ‘An Account of the Life and Writings of Adam Smith, LL. D.’. From the Transactions of the Royal Society of Edinburgh. Page references are from W. P. D. Wightman, J. C. Bryce and I. S. Ross, eds. (1980). The Glasgow Edition of the Works and Correspondence of Adam Smith, Volume 3. Oxford: Clarendon Press. Trenchard, John (1709). The Natural History of Superstition. London: A. Baldwin. Warburton, William (1754–5). A View of Lord Bolingbroke’s Philosophy; In Four Letters to a Friend. London: John and Paul Knapton. Yandell, Keith E. (1990). Hume’s Inexplicable Mystery. Philadelphia: Temple University Press.

6 The First Religious Principles

Having set the scene in the previous chapter, I turn now to an examination of Hume’s argument in the Natural History itself. I begin with Hume’s unorthodox insistence that polytheism was the first religion of mankind (§6.1), and then—following an interlude on the account of hope and fear offered in section 1 of the Dissertation on the Passions (§6.2)—move on to his explanations of how this polytheism arose and of the subsequent emergence of monotheism on the back of it (§6.3). The second half of the Natural History is mostly taken up with a series of comparisons between polytheism and monotheism, and I look at these in §6.4, before concluding with a general point about whether Hume was an outright atheist or merely an agnostic (§6.5). As in the previous chapter, I will be keen to emphasize throughout the importance of Hume’s philosophy of emotion in understanding the Natural History. But I will also be pointing out various ways in which this dissertation contributes to Hume’s overall critique of (superstitious) religious belief, in addition to discussing the dissertation’s examination of its psychological causes, thereby fulfilling the promise I made in §5.1.

6.1.  That Polytheism Was the Primary Religion of Men Hume’s opening move in the Natural History is to argue that polytheism was temporally prior to monotheism in the history of religion. As noted at the end of the previous chapter (§5.5), this position had been endorsed by deists such as Bolingbroke, but was deeply offensive to orthodox Christians such as Cudworth, who maintained—in accordance with the Biblical narrative of creation—that monotheism came first, and that idolatry was a later corruption of the true faith. Perhaps I am insufficiently critical on this point, being predisposed to agree with Hume, but it seems to me that the argument on this head in the Natural History is quite persuasive, certainly by the standards and evidence of the time, and on the background assumption that the Biblical narrative cannot reasonably be taken as historical fact.1 Twice at the end of the last century, however, his argument has been examined and found wanting. Responding

112  Part II to these criticisms, perhaps of interest in its own right, may also help to clarify something of Hume’s aims and methods in this essay. Hume’s first point in support of his position is simple enough: It is a matter of fact incontestable, that about 1700  years ago all mankind were polytheists. The doubtful and sceptical principles of a few philosophers, or the theism, and that too not entirely pure, of one or two nations, form no objection worth regarding. Behold then the clear testimony of history. The farther we mount up into antiquity, the more do we find mankind plunged into polytheism. . . . The north, the south, the east, the west, give their unanimous testimony to the same fact. What can be opposed to so full an evidence? (N 1.2, p. 34) Malherbe, in the same article I criticized on separate (but related) grounds in §5.4, expresses punctuated consternation that ‘[t]he Jews’ theism stands for nothing; nor do the sceptical principles of some philosophers!’ (1995, p. 262). ‘What could be opposed’, he goes on, ‘to so indistinct and so uncritical an evidence?’ (ibid.). A few years earlier, Mark Webb had voiced similar concerns, albeit less rhetorically: What can be opposed to so full an evidence? Nothing, perhaps, except the ‘doubtful and sceptical principles of a few philosophers, or the theism, and that too not entirely pure, of one or two nations’ (emphasis added) he mentions all too briefly. These he claims ‘form no objection worth regarding’ (NHR 4:310). But for an empirical study of the nature he presumably is undertaking they do form a critically important exception because they militate from the start against not only his account of the origin of polytheism, but especially the subsequent derivation of traditional theism. Unless he can explain, rather than ignore, such anomalies—present as far back as his evidence goes—his causal account is not so much false as fundamentally misleading and incomplete: the empirical evidence available to him offers no reason to believe that polytheism was the original religion of mankind. (1991, p. 147) Both Malherbe and Webb, however, seem to have a faulty conception of Hume’s dialectic, which leads them either to ignore or to misrepresent the full body of evidence that he offers in support of his hypothesis. The first thing to note is that the evidence of history is only the start of Hume’s argument. He goes on to add to it the evidence of ‘our present experience concerning the principles and opinions of barbarous nations’: ‘The savage tribes of America, Africa, and Asia are all idolaters’ (N 1.4, p. 34). More importantly, he then produces an extended argument based

The First Religious Principles  113 on observed principles of human nature, to the effect that the hypothesis that polytheism came first accords much better with those principles. The first principle is that ‘[t]he mind rises gradually, from inferior to superior’, and that ‘[n]othing could disturb this natural progress of our thought, but some obvious and invincible argument, which might immediately lead the mind into the pure principles of theism, and make it overleap, at one bound, the vast interval which is interposed between the human and the divine nature’ (N  1.5, p.  35). He does allow that ‘the order and frame of the universe, when accurately examined, affords such an argument’ (ibid.). But he rejects the possibility that this argument influenced the human animal in its infancy, primarily on the basis of a second principle, concerning the passions of curiosity and wonder: that they are excited much more by what is rare and novel, than by order and regularity (N 1.6, p. 356). This principle is taken, unsurprisingly, from his treatment of the passions (T 2.3.5.2, pp. 422–3; P 6.12, p. 28).2 A third principle provides further confirmation of Hume’s hypothesis, namely the principle that curiosity diminishes the further removed we are from its object: [A]an animal, compleat in all its limbs and organs, is to him [a man on the first origin of society] an ordinary spectacle, and produces no religious opinion or affection. Ask him, whence that animal arose; he will tell you, from the copulation of its parents. And these, whence? From the copulation of theirs. A  few removes satisfy his curiosity, and set the objects at such a distance, that he entirely loses sight of them. (N 1.6, p. 35) This principle features in Hume’s account of curiosity at the end of Treatise Book 2 (T 2.3.7.9, pp. 431–2), an account that does not reappear in the Dissertation on the Passions. It is also a special case of a principle concerning the effect of distance on the passions generally, examined earlier in Book 2 (T 2.3.7.1–3, pp. 427–9).3 Though this discussion is also removed from the Dissertation, I suggested in §1.2 and §1.3 that it may instead have featured in the original fourth dissertation on geometry. Hume concludes his case with a dilemma. The argument required to disturb the ‘natural progress of our thought’ must either be obvious or be abstruse. If it is obvious, then it is implausible to suppose that the belief could be so widely corrupted. But if it is abstruse, then its conclusion will necessarily be limited to a small number of unusually philosophically minded people: ‘Whichever side of this dilemma we take, it must appear impossible, that theism could, from reasoning, have been the primary religion of human race, and have afterwards, by its corruption, given birth to polytheism and to all the various superstitions of the heathen world. Reason, when obvious, prevents these corruptions: When

114  Part II abstruse, it keeps the principles entirely from the knowledge of the vulgar, who are alone liable to corrupt any principle or opinion’ (N 1.8, p. 36). Webb acknowledges Hume’s argument based on principles of human nature, but describes it as ‘non-empirical’ (1991, p. 147), in contrast to the empirical appeal to historical fact at the start of the section. Perhaps a similar attitude is behind Malherbe’s curious diagnoses of much of Hume’s argumentation in the Natural History as ‘deductive’, already complained of in §5.4 (note 8). But historical facts are not the only empirical facts, nor are arguments based on them the only empirical arguments. Hume bases his case as much on present observations of what human beings are like, as on historical reports of our ancient beliefs and practices, and his argument is thoroughly empirical (see Bailey & O’Brien 2014, p. 169). Webb also complains (in the quotation given earlier) that Hume needs to explain, rather than simply ignore, the Jewish exception to his polytheistic rule. The criticism is just in principle, but misplaced in practice, for Hume devotes section 6 of his essay to explaining precisely this. First, he repeats his argument that reasoning from the order of the universe is an implausible explanation (N 6.1, p. 52), adding that religious sentiments are generated much more by ‘[c]onvulsions in nature, disorders, prodigies, miracles, . . . the causes of events seeming then the most unknown and unaccountable’ (N 6.3, p. 53; see also N 3.4, p. 42, quoted in §5.2). He then offers an alternative explanation, based on principles of adulation and fear operating on a polytheistic religion in which one god is thought to be in some way preeminent: Whether this god, therefore, be considered as their peculiar patron, or as the general sovereign of heaven, his votaries will endeavour, by every art, to insinuate themselves into his favour; and supposing him to be pleased, like themselves, with praise and flattery, there is no eulogy or exaggeration, which will be spared in their addresses to him. In proportion as men’s fears or distresses become more urgent, they still invent new strains of adulation; and even he who outdoes his predecessor in swelling up the titles of his divinity, is sure to be outdone by his successor in newer and more pompous epithets of praise. Thus they proceed; till at last they arrive at infinity itself, beyond which there is no farther progress. (N 6.5, p. 53) ‘Thus’, Hume concludes after some additional related observations, ‘the God of Abraham, Isaac, and Jacob, became the supreme deity or Jehovah of the Jews’ (N 6.8, p. 54). The pleasure that we take in praise and flattery was of course a topic of particular interest to Hume in Book 2 of the Treatise, as we have already seen (§3.3). His Mandevillean account of it is briefly restated in the Dissertation on the Passions, apparently contradicting the Butlerian

The First Religious Principles  115 hypothesis of an original desire put forward in the moral Enquiry (§4.3). In any case, however the pleasure itself is to be explained, the fact of it here feeds into another psychological story. When combined with our tendency to conceive of all beings as like ourselves, and our fears concerning future events that we imagine are in the control of some preeminent god, this human vanity affords us a plausible explanation for the development of monotheism from polytheism. On the other hand, Hume’s opponents who maintain the contrary opinion lack any credible story (for the Fall of Adam is scarcely credible) that might explain why an original monotheism should have disappeared from most parts of the world before the birth of Jesus.

6.2.  Agitated by Hopes and Fears Having rejected, in section 1, the idea that we might be led to our first religious beliefs by the design argument, Hume goes on, at the start of section 2, to rule out that possibility even more forcibly as a potential origin of polytheism in particular, since that argument most naturally suggests a single designer (N 2.2, p. 37).4 Instead, he suggests that ‘we trace the footsteps of invisible power in the various and contrary events of human life’ (N  2.3, p.  37), arguing that ‘in all nations, which have embraced polytheism, the first ideas of religion arose not from a contemplation of the works of nature, but from a concern with regard to the events of life, and from the incessant hopes and fears, which actuate the human mind’ (N 2.4, p. 38). I will come to the details of Hume’s account of the origin of polytheism in the next section. First, we should acquaint ourselves with these two crucial components of it, the passions of hope and fear. The importance of these passions in Hume’s explanation is clear in the Natural History itself. As if to ensure that no one would miss the connection, however, Hume also originally published this work in a volume bound up with the Dissertation on the Passions immediately following it, a dissertation that begins with his account of the direct passions, with a particular emphasis on these two: ‘None of these passions seem to contain anything curious or remarkable, except Hope and Fear, which, being derived from the probability of any good or evil, are mixed passions, that merit our attention’ (P 1.7, p. 3). The ingredients of the mixed passions of hope and fear are joy and sorrow (P 1.10, p. 4). On the one hand, joy and sorrow are simply pleasant and painful feelings respectively, regarding some fact, present or past: the joy of a magnificent feast (T 2.1.5.1, p. 285), for example, or regarding the birth of a son (T  2.3.9.14, p.  441; P  1.21, p.  6), or the sorrow at losing a law-suit (ibid.). On the other hand, however, Hume also thinks of joy and sorrow as pleasures and pains regarding imagined situations, possibly future. ‘When good is certain or very probable’, he tells us, ‘it produces joy’, and likewise for evil and sorrow (P 1.4, p. 3;

116  Part II my emphasis; see also T 2.3.9.5, p. 439). Now, the claim that we experience joy in response to possible future pleasures as well as actual present ones strikes me as somewhat odd. There is a precedent for the idea in Locke: ‘Joy is a delight of the Mind, from the consideration of the present or assured approaching possession of a Good’ (1690, p. 231; second emphasis mine).5 This Lockean precedent, however, does not render the notion any less puzzling. To be clear, I  am not troubled by the claim that the anticipation of a future good, as well as the present experience of it, can give rise to a kind of pleasure. What troubles me is the suggestion that there is one single passion—joy—for both of these cases. Instead, I  am inclined to distinguish joy from (as we might say) excitement. The possession of a certain book will give me joy; having ordered it, I am pleasantly excited. But these are two different feelings. Similarly with sorrow and (let us say) apprehension: loss of the book will make me sad, while consideration or anticipation of this loss instead makes me apprehensive. This distinction might make Hume’s claim about the ingredients in the mixtures of hope and fear somewhat easier to understand. For it is the anticipatory sort of joy and sorrow that Hume has in mind here, not the sort of joy and sorrow that arises immediately from some present experience. With this point cleared up, the next most remarkable feature of Hume’s account of these passions is the perhaps surprising extent to which reason is involved in their production. As I pointed out in §3.1, the direct passions are prompted by ideas of good or evil, and hence at the very least require reflection on our present experience. In the case of the anticipatory passions, however, reason itself is also typically implicated. Hope and fear are not unthinking ‘gut reactions’, for Hume, but rather felt responses to the intellectual exercise of weighing up probabilities: Probability arises from an opposition of contrary chances or causes, by which the mind is not allowed to fix on either side; but is incessantly tossed from one to another, and is determined, one moment, to consider an object as existent, and another moment as the contrary. . . . Suppose, then, that the object, concerning which we are doubtful, produces either desire or aversion; it is evident, that, according as the mind turns itself to one side or the other, it must feel a momentary impression of joy or sorrow [excitement or apprehension]. . . . According as the probability inclines to good or evil, the passion of grief or joy [apprehension or excitement] predominates in the composition; and these passions being intermingled by means of the contrary views of the imagination, produce by the union the passions of hope or fear. (P 1.8–10, pp. 3–4; see also T 2.3.9.10–2, pp. 440–1)

The First Religious Principles  117 To be more precise, the immediate cause of these passions is a certain ‘inconstant and wavering survey of an object’ (P 1.14, p. 5), and there are other ways Hume thinks this inconstancy can arise.6 But the paradigmatic and most common underlying cause is the consideration of probabilities. Given the extent to which reason is involved in the production of these passions, and the role that these passions then have in the generation of religious belief, it should, I hope, be immediately doubtful that Hume’s account of the causes of this belief should be straightforwardly separable from the question of the reasons that might be given in support of it. At the very least, we have grounds for approaching the matter with an open mind. When we do approach the matter, as I will now argue, we can see that the two are indeed intimately bound up with each other.

6.3.  Irrational and Superstitious Principles There are two main aspects to Hume’s account of the origin of polytheism. The first is our passionate nature, most of the relevant aspects of which have already been touched on earlier: our curiosity, excited most by things close to us and out of the ordinary; our hopefulness and fearfulness, especially the latter, excited by uncertainty and inconstancy. To these we should also add, simply enough, our ‘concern with regard to the events of life’, which is intimately bound up with ‘the incessant hopes and fears, which actuate the human mind’ (N 2.4, p. 38). Place a creature like this in a world full of chaos and danger, and the ‘unknown causes’ of our good and ill fortune inevitably ‘become the constant object of our hope and fear’, and we are compelled to form ‘ideas of those powers, on which we have so entire a dependence’ (N 3.1, p. 40). This much puts us on the brink of polytheism, but what tips us over the edge is then the second aspect of Hume’s story, namely the ‘universal tendency among mankind to conceive all beings like themselves, and to transfer to every object, those qualities, with which they are familiarly acquainted, and of which they are intimately conscious’ (N 3.2, p. 40). After backing up this general principle with several examples, Hume applies it to the present case: ‘No wonder, then, that mankind, being placed in such an absolute ignorance of causes, and being at the same time so anxious concerning their future fortune, should immediately acknowledge a dependence on invisible powers, possessed of sentiment and intelligence’ (N 3.2, p. 41). As I argued in the previous section, Hume did not consider hope and fear to be immediate unthinking emotional responses. On the contrary, they are typically the result of judgements of probability, i.e. of cause and effect. It is only to be expected, therefore, that the assessment of such judgements is going to be relevant for Hume here, and that his discussion is not likely to be purely descriptive or psychological. Sure enough, Hume is quite clear that this first form of religious belief is contrary to the evidence:

118  Part II Could men anatomize nature, according to the most probable, at least most intelligible philosophy, they would find, that these causes are nothing but the peculiar fabric and structure of the minute parts of their own bodies and of external objects; and that, by a regular and constant machinery, all the events are produced, about which they are so much concerned. But this philosophy exceeds the comprehension of the ignorant multitude, who can only conceive the unknown causes in a general and confused manner; though their imagination, perpetually employed on the same subject, must labour to form some particular and distinct idea of them. (N 3.1, p. 40)7 Polytheism, as far as Hume is concerned, is thus bad science. There is nothing a priori objectionable about it; the idea of invisible, intelligent power is perfectly coherent. Hume makes this point particularly clear in the ambitious claim that ‘the whole mythological system is so natural, that, in the vast variety of planets and worlds, contained in this universe, it seems more than probable, that, somewhere or other, it is really carried into execution’ (N 11.1, p. 65). We needn’t take this suggestion about the existence of gods on other planets too seriously. Toning down the rhetoric, we nevertheless have the important point that the existence of invisible agents is, for Hume, an unproblematic a priori possibility. ‘The chief objection to it with regard to this planet’, he goes on, ‘is, that it is not ascertained by any just reason or authority’ (N 11.2, p. 65). This a posteriori argument that there is no evidence for supernatural intervention will doubtless call to mind Hume’s famous (and much more widely discussed) argument against testimonial evidence for miracles, in section 10 of the first Enquiry. Perhaps it will even be thought that, when Hume claims in the Natural History that there is no evidence for supernatural intervention, he is merely recalling that argument; for of course testimony for miraculous events purports to be just such evidence. The Natural History is not simply referring back to an argument given elsewhere, however, but offering a different argument of its own. In doing so, furthermore, it is filling an important gap that would otherwise be left in Hume’s attack on religion, as I now explain. Perhaps it is easily forgotten that the superstitious see the operation of supernatural agency, not just in miraculous events, but also in natural events that are suitably momentous. Droughts are nothing miraculous, but they are of such devastating consequence that the superstitious are apt to see them as the result of divine wrath. It is no miracle, either, that someone should recover from a potentially (but not necessarily) fatal illness, but when a superstitious friend has been praying for this very thing, they will readily see the hand of a particular providence at work: Even at this day, and in Europe, ask any of the vulgar, why he believes in an omnipotent creator of the world; he will never mention

The First Religious Principles  119 the beauty of final causes, of which he is wholly ignorant. . . . He will tell you of the sudden and unexpected death of such a one: The fall and bruise of such another: The excessive drought of this season: The cold and rains of another. These he ascribes to the immediate operation of providence: And such events, as, with good reasoners, are the chief difficulties in admitting a supreme intelligence, are with him the sole arguments for it. (N 6.1, p. 52) The issue in the case of purported miracles is not whether supernatural agency was responsible, but whether the supposed miracle actually took place (see Millican 2011, p. 168). Thus Hume’s argument in Of Miracles is designed to show that reports of miraculous events are never sufficiently credible, given the antecedent unlikelihood of what is reported, to establish that a miraculous event genuinely occurred. The issue in the present case, however, is not whether the event hoped for or dreaded really took place, but whether anything divine was directly responsible for it. Hume’s argument in the Natural History, accordingly, does not attempt to undermine the reliability of testimony. Rather, Hume points to the observed order and regularity in the world: when we ‘anatomize nature’ more closely, we discover that the causes of these momentous events are in fact ‘nothing but the peculiar fabric and structure of the minute parts of [our] own bodies and of external objects’ (N 3.1, p. 40). The argument here is certainly not as interesting or profound as the argument concerning the credibility of testimony, but it is a piece of the puzzle nonetheless. Potentially rather more interesting is the implication all this has for Hume’s presentation and endorsement of the design argument in the Natural History. The passions of hope and fear, as we have seen, depend on inconstancy and fluctuation, typically arising from chance or uncertainty in the course of nature. But a proper engagement with the design argument necessarily precludes these causes from entering the imagination, requiring us to fix our attention instead on the order and regularity of the universe, which can alone afford any inference to divine design. If the belief in God can arise through this argument, the fear of God never can. Indeed, it is very doubtful that the God of true religion can, for Hume, be the cause or object of any of our passions (except perhaps a calm, disinterested curiosity); as soon as any other passion is implicated, we would seem to be in the territory of superstition or enthusiasm, and in the unwarranted personification of the deity.8 Monotheism differs from polytheism only in the insistence that there is but one true God. As explained in §6.1, Hume charts the emergence of such forms of superstition, not in the persuasive force of the design argument, but in the felt need to flatter one deity above all others. Eventually, after sufficient exaggeration of the praises bestowed on the favoured divinity, all others are seen as vain pretenders. And perhaps this supreme

120  Part II being will even be credited with creating the world, in which case his votaries ‘coincide, by chance, with the principles of reason and true philosophy’ (N 6.5, p. 54). But as long as this deity, however supreme, is still supposed to disturb ‘the settled order of events by particular volitions’ (N 6.2, p. 52), the belief is unjustified, founded on ‘irrational and superstitious principles’ (N 6.4, p. 53; my emphasis). To be clear, it is not just that the religious beliefs in question (both polytheistic and monotheistic) have, as a matter of historical fact, been founded on irrational principles, while there might be other paths to these beliefs that are rational. To repeat, the only rational path that Hume countenances to religious belief ends in true religion, which allows for the existence of only a non-interventionist god, and one who consequently cannot be the object of fear. The superstitious, however, both believe in and fear at least one divinity who interferes in the settled order of things, and these sentiments are not just arrived at by irrational principles, but are themselves irrational, since a closer look at the evidence points to underlying regularity, and to natural causes of the contentious phenomena. Thus we see that Hume’s psychological discussion of the causes of religious belief is intimately bound up with the philosophical issue of the arguments or evidence in support of it. The distinction between reasons and causes, while itself perfectly legitimate and known to Hume, just is not a distinction that can cleanly be applied to his discussions of religion in general, or to the Natural History in particular.

6.4.  Comparison of These Religions It is well known that Hume’s objections to superstition were moral as well as intellectual. In particular, it has not escaped scholarly notice that the Natural History, in addition to offering an account of the causes of religious belief, also spends some time discussing what Hume saw as its pernicious moral effects; see Penelhum (1975, pp.  173–4), Gaskin (1988, ch. 11), Webb (1991), and Ferreira (1999).9 This is the focus of most of sections 9–12, in which Hume compares monotheism and polytheism in various respects, typically pertaining to morality, and with the former almost always coming off the worse of the two. In section 14 he then argues explicitly for the ‘[b]ad influence of popular religions on morality’ (N 14, title, p. 81). Given this moral dimension to the text, it is clear (even by Gaskin’s own admission) that this work is not solely concerned with causes. The distinction between reasons and causes, indeed, leaves no room at all for a discussion of religion’s effects, on morality or anything else. When we see the Natural History instead as part of Hume’s overall attack on superstition, however, everything falls very neatly into place; for this attack is comprehensive, challenging superstition from the point of view of both reason and sentiment.

The First Religious Principles  121 If the introduction to the Natural History is intended to be disarming, by giving the impression that what follows will concern only causes and not reasons, the introductions to Hume’s other attacks on superstition are typically more forthright: I flatter myself, that I have discovered an argument . . . which, if just, will, with the wise and learned, be an everlasting check to all kinds of superstitious delusion, and consequently, will be useful as long as the world endures. (E 10.2, p. 110) One considerable advantage, that arises from philosophy, consists in the sovereign antidote, which it affords to superstition and false religion.  .  .  .  [S]uperstition, being founded on false opinion, must immediately vanish, when true philosophy has inspired juster sentiments of superior powers. (Su 1, pp. 578–9) That the corruption of the best things produces the worst, is grown into a maxim, and is commonly proved, among other instances, by the pernicious effects of superstition and enthusiasm, the corruptions of true religion. (SE 1, p. 73) What Hume says in this last quotation also sets the tone for the comparisons between polytheism and monotheism in the Natural History. Hume dresses up this later, thinly disguised attack on monotheistic religions with the same maxim—‘that the corruption of the best things begets the worst’ (N 11.1, p. 65)—politely allowing that monotheism in the form of true religion is better than pagan idolatry, while enabling him to pull no punches in his criticism of its superstitious corruptions. The importance of the Natural History in Hume’s overall attack on the morality of superstition should not be overlooked, any more than in his overall attack on its rationality. It is of course in the moral Enquiry that Hume develops his own secular account of moral distinctions, and argues that the ‘monkish virtues’ of ‘[c]elibacy, fasting, penance, mortification, self-denial, humility, silence, solitude’ are really vices (M 9.3, p. 270). But the Natural History provides an important addition to this argument, by explaining the motivation behind this behaviour: The duties, which a man performs as a friend or parent, seem merely owing to his benefactor or children; nor can he be wanting to these duties, without breaking through all the ties of nature and morality. A strong inclination may prompt him to the performance: A sentiment of order and moral obligation joins its force to these natural

122  Part II ties: And the whole man, if truly virtuous, is drawn to his duty, without any effort or endeavour. . . . In all this, a superstitious man finds nothing, which he has properly performed for the sake of this deity, or which can peculiarly recommend him to the divine favour and protection. He considers not, that the most genuine method of serving the divinity is by promoting the happiness of his creatures. He still looks out for some more immediate service of the supreme Being, in order to allay those terrors, with which he is haunted. And any practice, recommended to him, which either serves to no purpose in life, or offers the strongest violence to his natural inclinations; that practice he will the more readily embrace, on account of those very circumstances, which should make him absolutely reject it. It seems the more purely religious, because it proceeds from no mixture of any other motive or consideration. (N 14.6, pp. 82–3) Because there are, in human nature, motivations to moral behaviour that do not involve any reference to the divine, such behaviour is only coincidentally motivated (if at all) by the desire of pleasing God. In order to demonstrate their devotion, therefore, and to secure divine favour, the superstitious must pursue actions that are in no way recommended by, or are perhaps even contrary to, our natural benevolent instincts: ‘if he [the superstitious man] fast a day, or give himself a sound whipping; this has a direct reference, in his opinion, to the service of God. No other motive could engage him to such austerities’ (ibid.). Where the moral Enquiry offers a criticism of the monkish virtues, therefore, the Natural History offers an explanation of them, and one that ties them directly to superstition. Though most of Hume’s comparisons between polytheism and monotheism in the Natural History are moral, section 11 compares the two ‘[w]ith regard to reason or absurdity’ (N 11, title, p. 65). What Hume has to say about polytheism in this connection was already noted in passing in §6.3: he takes it to be a straightforward a priori possibility, and therefore not inherently absurd, but one that happens to be contradicted by the evidence. Yet again, monotheism comes off worst from the comparison, for it frequently lands itself in a priori absurdity in addition to involving claims that are contrary to experience. Hume provocatively suggests that an ‘appetite for absurdity and contradiction’ (N 11.3, p. 66) is almost inevitably coincident with popular theology: If that theology went not beyond reason and common sense, her doctrines would appear too easy and familiar. Amazement must of necessity be raised: Mystery affected: Darkness and obscurity sought

The First Religious Principles  123 after: And a foundation of merit afforded to the devout votaries, who desire an opportunity of subduing their rebellious reason, by the belief of the most unintelligible sophisms. (ibid.) Hume doesn’t mention any such ‘sophisms’ explicitly, or spell out the arguments of ‘rebellious reason’ against them; but the insinuation is already clear enough. There are as many a priori objections to monotheistic superstitions, Hume thinks, as there are absurdities associated with them. He is admittedly pessimistic about the persuasive force of these objections. In a memorable passage, he writes that ‘[t]o oppose the torrent of scholastic religion by such feeble maxims as these, that it is impossible for the same thing, to be and not to be, that the whole is greater than a part, that two and three make five; is pretending to stop the ocean with a bull-rush’ (N 11.5, p. 66). But there is no denying that this bullrush, though it contradicts only some forms of superstition, is another part of Hume’s overall attack on the rationality of religious belief, and a further reason to reject the view of the Natural History as an exclusively psychological work.

6.5.  My Abhorrence of Vulgar Superstitions Though mine seems to be a minority view, I am not the only commentator to have argued that the Natural History of Religion contributes to Hume’s discussion of reasons as well as causes. Recently Peter Kail has defended precisely this claim: The prevalent view is incorrect. NHR [the Natural History of Religion] is a philosophically important and powerful component in Hume’s campaign against the rationality of religious belief. (2007, p. 191) Kail’s reasons for thinking this, however, are quite different from my own. And while I welcome the conclusion, I disagree with the argument that gets him there. I  am anxious, therefore, to distance myself from Kail’s position. According to Kail, the Natural History constitutes Hume’s argument against a view that Kail calls ‘rational fideism’: ‘the position that it is rational to maintain religious belief in the absence of evidence or arguments in its favour’ (2007, p. 195). The kernel of Hume’s supposed argument is that some causal explanations of belief are destabilizing, in the sense that they give one a reason to suspend the belief in question in the absence of any evidence in support of it (for example, believing something merely on the basis of the testimony of a habitual liar); and that

124  Part II Hume’s explanation of religious belief is of this kind. Furthermore, Kail thinks that Hume’s attack on religion is incomplete without an argument against the rational fideist. For although Hume argues elsewhere that there is no evidence for religious belief, this does nothing to upset the fideist who thinks it is rational to believe even in the absence of such evidence. It seems to me there are two objections to the suggestion that Hume was offering this argument in the Natural History. The first is that if he was offering it, he certainly had a curiously secretive way of doing so. Hume never mentions or even alludes to the rational fideist as a target, nor does he so much as hint at the crucial premise in Kail’s argument (that his explanation of the causes of religious belief is destabilizing in Kail’s sense). The argument may well have been inspired by Hume, but Kail is unable to produce any textual evidence that shows Hume himself actually rehearsing it. The case is quite different, note, with the a priori bull-rush argument noted in §6.4, and the a posteriori argument that I  looked at in §6.3, both of which are explicit in the text. The second difficulty with Kail’s interpretation is that this argument against rational fideism, were Hume to endorse it, would involve conceding too much to his opponents. For Hume does not in fact believe that there is an absence of evidence in the present case: as we have seen, he is quite explicit in the Natural History that the evidence is against the existence of any supernatural intervention. In Hume’s attack on superstition, therefore, there is quite simply no gap for the rational fideist to move into. Kail is not the first commentator to have read arguments into the Natural History of Religion that are not there. Gaskin himself, though a major promoter of the idea that this work is concerned exclusively with causes, does at least allow that Hume’s psychological story has philosophical implications, in forming the basis of a reply to the argument from common consent: [A]n adequate account of the causes of religious belief, though not itself philosophy, is an essential complement to [Hume’s] philosophical thinking about religion. If, as Hume seems to conclude, the arguments of natural religion are bad or establish only a hesitant and highly attenuated conclusion, and if the authenticity of revelation is suspect, then an appeal will almost inevitably be made by the believer to the argument from general consent: why is it that religious belief is and always has been so very prevalent? (Gaskin 1988, p. 183) My worry with this suggestion is exactly parallel to my worry with Kail’s. While Hume might have used the account of the origins of religious belief given in the Natural History as the basis of a reply to the argument from common consent, there is no line of text in this work where he actually

The First Religious Principles  125 does, or even so much as hints at the possibility. Furthermore, to repeat, Hume doesn’t merely maintain that the arguments in favour of superstition are suspect or unsuccessful; he also insists that there is positive evidence in the other direction, in the form of the observed regularities in nature.10 The situation is analogous in the two essays Of Suicide and Of the Immortality of the Soul, where Hume doesn’t only criticize the arguments in favour of the immorality of suicide and of life after death, but offers positive arguments for the opposite conclusions. In the latter, for example, he writes that ‘[t]he physical arguments from the analogy of nature are strong for the mortality of the soul; and these are really the only philosophical arguments, which ought to be admitted with regard to this question, or indeed any question of fact’ (IS 30, p. 596). His position is not that there is an absence of evidence. His position is that the evidence tells against the religious orthodoxy. There is a general sense—which perhaps underlies the assumption common to Kail and Gaskin that I am here opposing—that Hume tended more towards agnosticism than outright atheism. After all, isn’t that the appropriate stance for a Sceptic to take? There may be some truth to this; I do not have the space or the need to delve into the matter deeply here. But whatever we ought to say about this, we must start with the distinction between superstition and true religion. If there is any sense in which Hume was agnostic, it concerns only true religion. When it comes to superstition, as I have said, he is openly and unequivocally opposed to it all the way down. Perhaps true religion (in some attenuated form) is consistent with the known facts, and with a just sense of morals. But superstition, for Hume, is in accordance with neither.

Notes   1 Hume doesn’t explicitly state this background assumption in the Natural History, still less argue for it; his defence of it, however, is to be found at the end of the first Enquiry, section 10: ‘Here then we are first to consider a book, presented to us by a barbarous and ignorant people, written in an age when they were still more barbarous, and in all probability long after the facts which it relates, corroborated by no concurring testimony, and resembling those fabulous accounts, which every nation gives of its origin. Upon reading this book, we find it full of prodigies and miracles. It gives an account of a state of the world and of human nature entirely different from the present. . . . I desire any one to lay his hand upon his heart, and after a serious consideration declare, whether he thinks that the falsehood of such a book, supported by such a testimony, would be more extraordinary and miraculous than all the miracles it relates; which is, however, necessary to make it be received, according to the measures of probability above established’ (E 10.40, p. 130).  2 In Treatise Book 2, Hume had also endeavoured to explain why ‘surprize is apt to change into fear’ (T  2.3.9.26, p.  446), something that provides a bridge between the present argument and the account Hume goes on to offer concerning the origin of polytheism, dependent as it is on the passion of fear (see §6.3). This additional explanation is absent from the Dissertation on the

126  Part II Passions, however, and the bridge between these two aspects of the Natural History’s account is left implicit in Hume’s overall argument.   3 See McIntyre (1999, pp. 172–3), who also emphasizes the importance of this principle in Hume’s Natural History.  4 In the Dialogues, Philo argues that the design argument does not establish the unity of god, and that the evidence might just as well suggest multiple deities (D 5.8–9, pp. 167–8). This is consistent with Hume’s account in the Natural History, which is about what the design argument naturally ‘leads the mind to acknowledge’ (N 2.2, p. 37), rather than what it in fact establishes. In the Natural History itself, Hume notes that ‘to persons of a certain turn of mind, it may not appear altogether absurd, that several independent beings, endowed with superior wisdom, might conspire in the contrivance and execution of one regular plan’ (N 2.2, p. 37), but dismisses the idea in the present context as serving ‘only to give perplexity to the imagination, without bestowing any satisfaction on the understanding’ (ibid.).   5 Locke’s definition of sorrow, by contrast, is restricted to real evil: ‘Sorrow is uneasiness in the Mind, upon the thought of a Good lost, which might have been enjoy’d longer; or the sense of a present Evil’ (1690, p. 231). But it is easy to see how one might extend the pleasant case to the unpleasant one, as Hume does.   6 For example, an impossible evil may produce fear in virtue of its proximity, ‘as when we tremble on the brink of a precipice, though we know ourselves to be in perfect security’ (P 1.16, p. 5); an inevitable evil may also produce fear if it is so great we cannot bear to think of it, as when ‘[a] man, in a strong prison, without the least means of escape, trembles at the thoughts of the rack’ (P 1.17, p. 5); and even a good may sometimes produce fear, as in the anticipation of losing one’s virginity, when ‘[t]he confusion of wishes and joys, the newness and greatness of the unknown event, so embarrass the mind, that it knows not in what image or passion to fix itself’ (P 1.20, p. 6).   7 This passage is reminiscent of one in the first Enquiry: ‘A peasant can give no better reason for the stopping of any clock or watch than to say that it does not commonly go right: But an artist easily perceives, that the same force in the spring or pendulum has always the same influence on the wheels; but fails of its usual effect, perhaps by reason of a grain of dust, which puts a stop to the whole movement’ (E 8.13, p. 87; see also T 1.3.12.5, p. 132). Superstitious believers are in the same epistemic situation as the peasant; but instead of searching for the grain of dust they allow their imaginations to conjure up an idea of some invisible supernatural agent responsible for keeping watches ticking (or at least with the power to stop them).  8 See McIntyre (1999, pp.  182–4), who argues that Hume rejects God as a possible object of the indirect passions, including most notably love. See also Holden (2010), who argues that Hume was a ‘moral atheist’, denying that God was an appropriate subject of moral assessment.   9 Webb goes further, arguing that the moral critique of religion is Hume’s primary aim in the Natural History, and that its explicitly advertised purpose is subservient to that agenda: ‘while the work purports to be a natural history of religious beliefs and practices, suggesting something of a “scientific” examination of a “natural” phenomenon, that feature of the work clearly is intended by Hume to facilitate his moral critique of religious belief’ (1991, p. 145). I cannot see that Hume had anything so subtle in mind, however, and continue to view his explicit main purpose as his genuine main purpose (albeit one that very naturally leads to a concomitant moral critique). Part of Webb’s case depends on the supposed weakness of Hume’s empirical arguments, and

The First Religious Principles  127 the claim that other arguments are not empirical at all, in opposition to which I have already said something in §6.1. 10 On the face of it, Bailey and O’Brien say things about the Natural History that resonate with my own interpretation, describing Hume’s account as ‘philosophically loaded’ and ‘not merely descriptive’ (2014, p. 167), and insisting that the work ‘plays a key destructive role’ in Hume’s ‘overall case against religious belief’ (2014, p. 178). The basis of these claims, however, is apparently just Gaskin’s idea that the dissertation contains an implicit reply to the argument from common consent: ‘First, given Hume’s natural history, the existence of widespread religious beliefs does not provide evidence that such beliefs are true. Second, his natural history should be taken in conjunction with his other works on religion, and with these, or so we argue, Hume provides a multi-fronted attack on religious belief and on religion’ (2014, p. 179). Thus the attack, for Bailey and O’Brien, still seems to be relegated to Hume’s other writings, with the Natural History itself being largely descriptive after all.

Bibliography Bailey, Alan and Dan O’Brien (2014). Hume’s Critique of Religious Belief: Sick Men’s Dreams. Dordrecht: Springer. Ferreira, M. Jamie (1999). ‘Hume’s “Mitigated Scepticism”: Some Implications for Religious Belief’. In Religion and Hume’s Legacy. Eds. by D. Z. Phillips and Timothy Tessin. Basingstoke: Palgrave Macmillan, pp. 47–67. Gaskin, J. C. A. (1988). Hume’s Philosophy of Religion. Basingstoke: Palgrave Macmillan. Holden, Thomas (2010). Spectres of False Divinity: Hume’s Moral Atheism. Oxford: Oxford University Press. Kail, P. J. E. (2007). ‘Understanding Hume’s Natural History of Religion’. The Philosophical Quarterly 57(227), pp. 190–211. Locke, John (1690). An Essay concerning Humane Understanding. London. Quotations and page references from Peter H. Nidditch, ed. (1975). An Essay concerning Human Understanding, by John Locke. Oxford: Clarendon Press. Malherbe, Michel (1995). ‘Hume’s Natural History of Religion’. Hume Studies 21(2), pp. 255–74. McIntyre, Jane (1999). ‘Passion and Artifice in Hume’s Account of Superstition’. In Religion and Hume’s Legacy. Eds. by D. Z. Phillips and Timothy Tessin. London: Palgrave Macmillan, pp. 171–84. Millican, Peter (2011). ‘Twenty Questions About Hume’s “Of Miracles” ’. In Philosophy and Religion. Ed. by Antony O’Hear. Cambridge: Cambridge University Press, pp. 151–92. Penelhum, Terence (1975). Hume. London and Basingstoke: Palgrave Macmillan. Webb, Mark (1991). ‘The Argument of the Natural History’. Hume Studies 17(2), pp. 141–60.

7 The Object of the Passions

Having concluded my examination of section  1 of the Dissertation on the Passions, alongside the Natural History of Religion, I turn now to the heart of the Dissertation, sections 2–4. (Section 5 will be the subject of the next chapter, and section 6 of Chapter 9.) These central sections contain Hume’s mature treatment of the indirect passions, notably pride, humility, love, and hatred. I will use my discussion of this treatment as the occasion to explore issues related to (what we now call) the intentionality of the passions. I begin by highlighting the importance of cognition in Hume’s account of the causes of the passions (§7.1), and replying to a popular argument to the effect that Hume denied their intentional nature altogether (§7.2). This may help to take the edge off his modern-day reputation as an extreme non-cognitivist. That being said, I then argue that the intentionality of the passions was a derivative or extrinsic matter in the Treatise, the result of our secondary impressions being causally related to appropriate ideas (§7.2 and §7.3). Interestingly, however, all the textual evidence for this interpretation is absent from Hume’s later work, which contains in its place hints of a more cognitivist view, according to which the passions are intrinsically intentional (§7.4 and §7.5). It would be overstating the case to say that Hume clearly embraced this view. My more modest conclusion is that he recognized a difficulty with his earlier position, and attempted to circumvent it—rather than to tackle it head on—by presenting his higher-level claims in such a way that they did not depend on how the lower-level details panned out.1

7.1.  Derived in a Great Measure From Our Ideas The very first distinction that Hume draws in his science of the mind, both in the Treatise and in his later work, is that between ‘feeling and thinking’, or equivalently between ‘impressions and ideas’ (T  1.1.1.1, p.  1; E  2.1–3, pp.  17–8). This distinction, I  trust, is well known, and I have little to add to our existing understanding of it. One thing that is not commonly remarked upon, however, is of particular interest here: whereas in the Treatise (and particularly in Books 1 and 2) Hume almost

The Object of the Passions  129 invariably labelled our felt perceptions ‘impressions’, in his later writing (and to an extent in Book 3 of the Treatise) there is an increasing preference for the term ‘sentiment’. When Hume draws the distinction in section  2 of the first Enquiry, ‘impression’ is still his official choice: Here therefore we may divide all the perceptions of the mind into two classes or species, which are distinguished by their different degrees of force and vivacity. The less forceful and lively are commonly denominated Thoughts or Ideas. The other species want a name in our language. . . . Let us, therefore, use a little freedom, and call them Impressions. (E 2.3, p. 18) But while the term ‘sentiment’ appears not once in the first section of the Treatise (where ‘impression’ appears 46 times), it appears five times in section 2 of the first Enquiry (where ‘impression’ appears ten times), and in the Enquiry as a whole it outnumbers ‘impression’ by almost two to one.2 There can be no doubt that some of the time, at least, Hume intends ‘sentiment’ as just a synonym for ‘impression’. A couple of examples suffice to prove the point (as it happens, any of the five from section 2 of the first Enquiry would do): These faculties [memory and the imagination] may mimic or copy the perceptions of the senses; but they never can entirely reach the force and vivacity of the original sentiment. (E 2.1, p. 17) [W]hen we analyse our thoughts or ideas, however compounded or sublime, we always find, that they resolve themselves into such simple ideas as were copied from a precedent feeling or sentiment. (E 2.6, p. 19) Furthermore, the theory of the ‘double relation of impressions and ideas’ from Book 2 of the Treatise (e.g. T 2.1.5.5, p. 286)—which I will examine in §7.5—has become, in the Dissertation on the Passions, that of the ‘double relation of sentiments and ideas’ (e.g. P 2.22, p. 11; my emphasis). Again, the switch is not complete: the theory is later referred to once by its former name (P 3.1, p. 18). But the new name occurs three times, whereas in the Treatise it was never used. When one considers the increase of the word in Book 3 of the Treatise, a relatively innocuous explanation suggests itself: that Hume adopted it simply in deference to Hutcheson (who used the term a lot). It seems to me doubtful, however, that this can be the whole story, or account for the

130  Part II fact that Hume’s preference became stronger still in his later publications. In any case, I suggest as another likely factor the intellectual connotations of the new term, which I suspect Hume would have been attracted to independently of any conscious attempt to please or outwardly align himself with Hutcheson. Though Hume describes impressions as feelings rather than thoughts, the word ‘sentiment’ is in fact nicely ambiguous or midway between these two things. Johnson’s dictionary defines it as ‘thought, notion, opinion’ (Johnson 1755), without any mention of the feeling sense of the word at all (though that is surely an oversight). Hume himself sometimes uses the word in its opinion sense (e.g. E 4.21, p. 37; E 8.21, p. 92); and when he talks of the ‘sentiments of religion’ in the Natural History (N 6.3, p. 53), he might easily be thinking of belief, passion, or both. The rebirth of the philosophy of emotion in the second half of the last century was, in large part, a defence of ‘cognitivist’ accounts of the passions, in conscious opposition to what were viewed as the implausibly ‘non-cognitivist’ views of Hume and others.3 I would certainly not say that all these worries were unfounded, or that the caricatures of Hume were entirely wide of the mark. But two aspects of Hume’s view are worth emphasizing, in the light of this recent history. The first is that Hume never denied that the passions have objects, or what we nowadays call intentional objects. I will defend this claim in the next section. The second is that some form of ‘cognition’ (not wishing to put too much weight here on that anachronistic term) was for Hume always an essential ingredient in the causes of the passions, and was thereby a central and ineliminable aspect of emotional life, at least broadly construed. Though many contemporary cognitivists will doubtless feel this doesn’t make cognition central enough, the point should still serve to undermine the caricature. Furthermore, as we will see at the end of this chapter, Hume’s later work is consistent with (and indeed suggestive of) a view on which cognition is more central still. It is also worth remembering that belief for Hume is ‘a feeling or sentiment’ (E 5.11, p. 48), in that respect just like the passions; or rather, to make the point in the other direction, the passions are in that respect just like belief. The thesis that the passions are caused by ideas is implied by Hume’s definition of them as impressions of reflection, and is affirmed perhaps most clearly in his brief discussion of Book 2 in the Abstract: ‘it is by means of thought only that any thing operates upon our passions’ (A 35, p. 662). In the Dissertation, Hume says concerning the origin of the direct passions that ‘[a]ll good and evil, whence-ever it arises, produces various passions and affections, according to the light in which it is surveyed’ (P 1.3, p. 3; my emphasis). Thus desire and aversion arise upon considering—i.e. entertaining the idea of—some good or evil, and likewise for joy and sorrow, hope and fear. Hope and fear, moreover, are typically caused in addition by judgements of probability (recall §6.2). Pride, humility, love, and hatred are caused by associations of sentiments and ideas (see

The Object of the Passions  131 §7.5). Malice and envy are caused by comparisons of ourselves with others (§3.4, §4.4). And pity and generosity are caused by the sympathetic communication of passions, a process which crucially involves awareness of oneself and of the other person, and an idea of the passion being communicated (§3.3, §4.4). Kames makes nice use of the double meaning of the word ‘sentiment’ in his definition of it as a ‘thought prompted by passion’ (1762, vol. 1, p. 311; see also vol. 2, p. 741). For Hume, meanwhile, the passions might well be described as feelings prompted by thought. In this way, I suggest that Hume’s later choice of term was very apt, and perhaps less likely than ‘impression’ to invite crude non-cognitivist misreadings.

7.2.  A Passion Is an Original Existence I said in the previous section that Hume never denied that the passions have intentional objects, but there is a notorious argument in the Treatise which many commentators have taken to involve Hume in precisely this denial. Following Phillips (2005, p. 308), I will refer to this as Hume’s representative quality argument. It comes in the context of his wider argument about the combat of passion and reason, and runs as follows: A passion is an original existence, or, if you will, modification of existence, and contains not any representative quality, which renders it a copy of any other existence or modification. When I am angry, I am actually possest with the passion, and in that emotion have no more a reference to any other object, than when I am thirsty, or sick, or more than five foot high. ’Tis impossible, therefore, that this passion can be oppos’d by, or be contradictory to truth and reason; since this contradiction consists in the disagreement of ideas, consider’d as copies, with those objects, which they represent. (T 2.3.3.5, p. 415) The difficulty here arises from Hume’s first premise: that a passion ‘contains not any representative quality’, or that, in having an emotion, I have ‘no more a reference to any other object, than when I am thirsty, or sick, or more than five foot high’. Phillips objects to this premise by reading it as a denial of the intentionality of the passions: ‘Hume is wrong to think that passions cannot represent: anger, his own example, typically has cognitive content (one is angry with a person, institution, etc.)’ (2005, p. 311). Baier is even more forthright, dismissing this whole paragraph as ‘very silly’ on the grounds of this interpretation (Baier 1991, p. 160). Many others have read the premise in the same way, some to complain about it, and some merely to observe that this was Hume’s view; e.g. Kenny (1963, p. 17, n. 1), Millgram (1995, p. 87), Prinz (2004, pp. 11, 52), and Penelhum (2009, pp. 250–1).

132  Part II To clear Hume of the charge of silliness here, we must distinguish between having an (intentional) object, and having a reference to an object; or equivalently between being about something, and being a representation of that thing. ‘The object of pride and humility’, Hume tells us, ‘is self’ (P 2.3, p. 7); ‘the object of envy’, somewhat similarly, is ‘some present enjoyment of another, which by comparison diminishes our idea of our own’ (T 2.2.8.12, p. 377). But pride and humility do not represent self, nor does envy represent the present enjoyment of another. Representation is a quality limited mainly to ideas (which represent the objects or impressions from which they are copied) and to sensory impressions (which represent external objects).4 And the point of denying that the passions represent anything, as should be clear from the dialectical context, is to rule out the possibility that they can be true or false, and hence fall under the jurisdiction of reason. A blanket denial that they have any objects at all might establish this conclusion, but on needlessly strong grounds. All that Hume needs to deny here is that they are copies of objects, that they are representations of objects with which they might disagree. As I read the crucial premise, that is all that he does deny.5 This is the only passage, so far as I am aware, that has led people to believe Hume denied the intentionality of the passions. Elsewhere, he seems to take their intentionality for granted. In section  1 of the Dissertation alone, for example, he talks of trembling ‘at’ the prospect of torture (P  1.17, p.  5), of feeling anxious ‘upon’ the account of a sick friend (P 1.19, p. 5), of being afflicted ‘for’ the loss of a law-suit, or joyful ‘for’ the birth of a son (P 1.21, p. 6). But though it thus seems clear that the passions have intentional objects for Hume, it is less clear how they have intentional objects, and what the precise nature of their relation to those objects is. And here Hume appears to have been drawn in two opposing directions. On the one hand, the atomist in him insisted on a sharp distinction between the passion itself (qua feeling or sensation) and its object, to which it bears—and necessarily can only bear—a contingent causal relationship. On the other hand, the holist in him inclined towards a more integrated view, according to which the passion is to be identified with the combination of its qualitative feel and its intentional object, and from which it follows that the intentionality of the passions is an intrinsic rather than an extrinsic matter. This tension is reflected in discussions of Hume, with some commentators inclined to emphasize one side more than the other, and others exhibiting the same tension in their own reconstructions. For my part, I incline to read Book 2 of the Treatise as very strongly atomistic in this regard, and I believe the supposed holism in this work has been exaggerated. But it is striking that all the best evidence of atomism from the Treatise fails to reappear in Hume’s later works. In the case of the intentionality of the passions, furthermore, these cuts coincide with a small but tantalizing number of additions that—at least on their

The Object of the Passions  133 face-value interpretation—invite a more holistic view. The evidence, it must be admitted, is far from decisive, but my tentative conclusion is that Hume was much less convinced of his earlier atomism in later life, and at the very least wanted to defend his higher-level claims about the passions in ways that didn’t depend on how these more foundational issues panned out. I thus find myself in the odd position of disagreeing with those who have emphasized Hume’s holistic tendencies in the Treatise, only to agree with them in the end about Hume’s considered view. I begin with the negative side of the case, focusing my attention on the Treatise; the contrasts with Hume’s later work will then be drawn out in §7.4 and §7.5. By default, it seems, the object of a passion in the Treatise is simply its cause. For example, I  noted earlier that Hume says the object of envy is some present enjoyment of another. I may now add that he also says, in the previous sentence, that envy is ‘excited’ by that enjoyment (T 2.2.8.12, p. 377); in other words, the object of this passion is its cause. When speaking of the pleasant emotion of beauty, similarly, Hume once refers to ‘[t]he object or cause of this pleasure’ (T 2.1.9.6, p. 306), implying that these things are one and the same. The examples of the loss of a law-suit and the birth of a son, meanwhile, are used to illustrate the principle that ‘[w]hen contrary passions arise from objects entirely different, they take place alternately’ (T 2.3.9.14, p. 411); these two unrelated objects are both what the passions arise from and what they are about. Though this appears to be the default relationship between the passions and their objects, there are a few notable exceptions, namely the passions of pride, humility, love, and hatred. As I will want to refer to these four passions often, it will be convenient to have a label for the set; I propose to call them the double-relation passions, in virtue of the fact that Hume accounts for their origin with his theory of the double relation of sentiments and ideas (I will examine the details of this theory in §7.5). In the case of these double-relation passions, Hume explicitly distinguishes their object (self for pride and humility, another person for love and hatred) from their cause (something agreeable related to the object for pride and love, something disagreeable for humility and hatred). In the present context, however, two things should be noted about this distinction. First, the very fact that Hume is at pains to draw it in these cases provides confirmation of the general rule—that the object of the passion is its cause—in the other cases, where no such distinction is made. Second, although their objects are not the causes of these double-relation passions, the surrounding evidence still clearly indicates that the relation in question is a causal one, just of a different kind. In fact Hume is not consistent in the Treatise about the precise nature of this relation. On the one hand, he maintains that the object of these passions is their effect:

134  Part II The first idea, that is presented to the mind, is that of the cause or productive principle. This excites the passion [of pride], connected with it; and that passion, when excited, turns our view to another idea, which is that of self. Here then is a passion plac’d betwixt two ideas, of which the one produces it, and the other is produc’d by it. (T 2.1.2.4, p. 278) [N]ature has given to the organs of the human mind, a certain disposition fitted to produce a peculiar impression or emotion, which we call pride: To this emotion she has assign’d a certain idea, viz. that of self, which it never fails to produce. (T 2.1.5.6, p. 287) On the other hand, however, he claims that the object is a necessary but insufficient part of the passion’s cause: In order to excite pride, there are always two objects we must contemplate, viz. the cause or that object which produces pleasure; and self, which is the real object of the passion. (T 2.1.6.5, p. 292) Of course, it is perfectly coherent to suppose that the idea of self is both a necessary part of the cause of pride and, as it happens, one of its effects. But this does not appear to be what Hume says. In the first of the three passages just quoted, he says that what produces pride is its cause alone (which is of course the point of calling that its cause). In the third passage, meanwhile, he says that contemplation of self is also necessary in order to excite the passion. It is not at all clear to me which of these two alternatives he should have embraced in the Treatise. Indeed, as I  will explain in §7.5, he had good reason to reject both of them. In any case, it seems inevitable that the connection between the double-relation passions and their objects is, in the Treatise, a contingent rather than an essential matter, just as it is in the more straightforward case where the object and the cause of the passion are identical. Self or another person are causally related to these passions (in one direction or another), but form no part of their definition, any more than the birth of a son is part of the definition of joy. More than this, in fact, Hume is explicit in affirming the simplicity—and hence indefinability altogether— of these passions: THE passions of pride and humility being simple and uniform impressions, ’tis impossible we can ever, by a multitude of words, give a just definition of them, or indeed of any of the passions. (T 2.1.2.1, p. 277)

The Object of the Passions  135 ’TIS altogether impossible to give any definition of the passions of love and hatred; and that because they produce merely a simple impression, without any mixture or composition. (T 2.2.1.1, p. 329) Thus we are still firmly in the territory of the extrinsic interpretation of the intentionality of the passions, whichever side of the present dilemma we choose.

7.3.  From the Primary Constitution of the Mind Hume’s equivocation over whether the object of the double-relation passions is their effect or a part of their cause does not appear to have been clearly noted before, though Penelhum comes close (1975, p. 99; 2015, p. 215). This complication aside, the extrinsic interpretation of Hume on the intentionality of the passions is widely held, and is probably fairly described as the mainstream view.6 There has been some resistance to it, however, with a handful of commentators maintaining (not always, perhaps, with complete clarity or consistency) that Hume thought the passions were complex perceptions, whose causes and objects formed a part of their essence or definition, or relatedly that the connection with their causes and objects was a priori and not contingent.7 I will not dwell on these views here, as I have criticized them in detail elsewhere (Merivale 2009, pp. 192–5, 198–200). Instead, I will examine a more recent interpretation in a similar vein, in order to illustrate the point. My position, as hinted at in the introduction, is that these interpretations are wrong as applied to the Treatise, but might well be closer to the mark regarding Hume’s later view of the passions. The more recent interpretation I have in mind was put forward by Hsueh Qu (2012). Qu argues that, although the passions are simple impressions, they nevertheless have an intrinsic, essential, and n ­ on-­contingent ­intentionality. Qu acknowledges an immediate difficulty with this proposal, in that it would seem to contradict Hume’s so-called separability principle: ‘that whatever objects are different are distinguishable, and that whatever objects are distinguishable are separable by the thought and imagination’ (T  1.1.7.3, p.  18). On Qu’s account, a Humean passion comprises both a qualitative character and an intentional object; but as these two components appear to be distinguishable and separable, the consequence is that the passions must be complex rather than simple impressions. Qu avoids this conclusion by denying that the two components are separable, claiming instead that the difference between a passion’s character and its object is a ‘distinction of reason’, like that between ‘figure and the body figur’d; motion and the body mov’d’ (T  1.1.7.17, p.  24). The sensation and the intentionality both reside in one and the

136  Part II same simple impression, but this impression can serve as the model for different abstract ideas, one for the quality and another for the object. In the first instance, it should be noted that the most this interpretation can claim to be is an account of what Hume might have maintained, consistently with his other principles; as a suggestion about what he in fact believed, it suffers from the obvious problem that Hume never even came close to stating it. Abstract ideas and distinctions of reason are not so much as mentioned anywhere in Book 2. This does not mean that the interpretation is necessarily unmotivated; it could be motivated by pointing to a problem that it solves, in which case it would have a good claim to being, if not what Hume actually thought, then at least what he could and should have thought. This, I take it, is Qu’s position, since he argues that alternatives such as the atomistic interpretation I have been defending (in the case of the Treatise) ‘ultimately fail to capture something crucial in Hume’s treatment of the intentionality of the passions’ (2012, p. 102). Qu offers three objections to extrinsic interpretations of the intentionality of the passions: ‘first, they cannot account for pride and humility’s being essentially directed towards the self; second, they contradict Hume’s claim that the intentionality of pride and humility is an original quality; and, third, Hume seems to see the intentionality of pride and humility as a quality in the passions (rather than an extraneous quality of them)’ (2012, pp. 102–3). I will return to the first objection later. The second and third are closely related, both drawing on the same passage from Book 2 for their principal evidence. The passage, slightly abridged, runs as follows: Having thus in a manner suppos’d two properties of the causes of these affections . . . I proceed to examine the passions themselves, in order to find something in them, correspondent to the suppos’d properties of their causes. First, I find, that the peculiar object of pride and humility is determin’d by an original and natural instinct, and that ’tis absolutely impossible, from the primary constitution of the mind, that these passions shou’d ever look beyond self, or that individual person, of whose actions and sentiments each of us is intimately conscious. . . . For this I pretend not to give any reason; but consider such a peculiar direction of the thought as an original quality. The second quality, which I discover in these passions, and which I likewise consider as an original quality, is their sensations, or the peculiar emotions they excite in the soul, and which constitute their very being and essence. Thus pride is a pleasant sensation, and humility a painful; and upon the removal of the pleasure and pain, there is in reality no pride nor humility. (T 2.1.5.3–4, pp. 285–6) In this passage, Hume describes the relationship to self as a quality ‘in’ the passions of pride and humility, on a par with their sensations or qualitative character, and this is the core of Qu’s third objection. But there is

The Object of the Passions  137 a leap from saying that something is a quality ‘in’ the passions to saying that it is an intrinsic or essential quality, and the text here seems to rule out precisely this leap, for two reasons. First, the second quality—the sensation of these passions—is explicitly said to ‘constitute their very being and essence’; by implication, then, the first quality is not a part of their essence, but is an extrinsic, relational property. Second, and more importantly, Hume says that what makes it impossible for pride and humility ever to look beyond self is not the nature of these passions themselves, as it would be on Qu’s interpretation, but rather ‘the primary constitution of the mind’. Thus, while pride is pleasant simply because it is a pleasant sensation, it is directed at self because of the relational properties of the mind. In response to my earlier paper, in which I defended this view, Qu remarks that my interpretation ‘is difficult to reconcile with Hume’s phrasing’ (2012, p.  114, n. 22). But he is apparently placing a lot of weight on the little word ‘in’, at the expense of all the others. On the contrary, it seems to me it is Qu’s view that is difficult to reconcile with the text. Qu’s second and ‘most serious worry’ (2012, p.  105) stems from Hume’s claim, as also exhibited in the present passage, that the relationship of pride or humility to self is the result of an original quality or principle. Qu argues, quite rightly I believe, that original principles are explanatorily basic for Hume, in the strong sense that they rule out all further kinds of explanation; not only those appealing to intermediate causal connections, but also those resolving the principle into another that is more general (2012, p. 106). The upshot of this, supposedly, is that the relationship between pride or humility and self cannot be extrinsic and causal, because in that case it ‘would be resolvable into more general principles, namely the three principles of association, which are the only general principles that can unite our perceptions’ (2012, p. 107). I confess the reasoning here escapes me, and it seems to me on the contrary that because the principle in question is original it must be causal. Original principles are, by their very nature, causal principles; they are the endpoint of a posteriori scientific investigation. Just because the connection in question is causal, meanwhile, that does not mean it must be resolvable into the association of ideas between cause and effect. Pride is not connected to self merely through the association of ideas, which prompts us to think of its effect (or a necessary but insufficient part of its cause); it is connected to the idea of self, first and foremost, because that is its effect (or a necessary but insufficient part of its cause).

7.4. It Is Essential to Pride to Turn Our View on Ourselves While Qu’s second two objections to extrinsic interpretations of the intentionality of the passions thus seem to me unconvincing, his first objection is rather more interesting. The worry, recall, is that ‘pride and humility

138  Part II seem to be essentially directed at their intentional objects’ (2012, p. 103). If true, this would indeed be in direct contradiction to the extrinsic interpretation. The evidence Qu offers from the Treatise, however, does not strike me as particularly strong, and certainly far from strong enough to counterbalance the evidence on the other side. It consists of the following three passages (the third of which we already met in the previous section), in which Hume emphasizes the closeness of the connection between these passions and their object: When self enters not into the consideration, there is no room either for pride or humility. (T 2.1.2.2, p. 277) ’Tis always self, which is the object of pride and humility. . . . That this proceeds from an original quality or primary impulse, will likewise appear evident, if we consider that ’tis the distinguishing characteristic of these passions. (T 2.1.3.2–3, p. 280) I find, that the peculiar object of pride and humility is determin’d by an original and natural instinct, and that ’tis absolutely impossible, from the primary constitution of the mind, that these passions shou’d ever look beyond self. (T 2.1.5.3, p. 286) Qu draws attention to the absolute impossibility that these passions should ever look beyond self, and to Hume’s description of this as their distinguishing characteristic. The first of these points, however, can easily be read as a claim about causal impossibility, as Qu himself acknowledges. The second is admittedly harder to read in this way, and in isolation it is tempting to treat it as indicative of a more holistic view. But in explicating the claim further in the same paragraph, Hume writes that the connection in question is based on one of those original qualities that is ‘most inseparable from the soul’ (T 2.1.3.3, p. 280), and not—as the holistic view would have it—most inseparable from the passions themselves. However, Qu also offers as evidence this intriguing claim from the Dissertation on the Passions, which has no ancestor in Treatise Book 2: With regard to all these passions [the double-relation passions], the causes are what excite the emotion; the object is what the mind directs its view to when the emotion is excited. Our merit, for instance, raises pride; and it is essential to pride to turn our view on ourselves with complacency and satisfaction. (P 2.4, p. 7) This claim, it seems to me, admits of no ambiguity or alternative interpretation: Hume explicitly affirms that the relationship to self is an essential

The Object of the Passions  139 feature of pride. With regard to this passage, Qu writes that the essentiality of this connection ‘is even more clearly emphasized in the Dissertation’ (2012, p. 103). But this misdescribes the relationship between Book 2 and the Dissertation on this point, forgetting that in the earlier work Hume had said the essences of pride and humility were simply ‘their sensations, or the peculiar emotions they excite in the soul’ (T 2.1.5.4, p. 286). On the face of it, at least, what we have in the Dissertation is not Hume emphasizing more clearly something that was previously mentioned only in passing; it is Hume contradicting one of his earlier claims. By itself, this one passage might be thought a slim basis for inferring any substantial change in Hume’s position. But the passage is backed up by another very striking addition, and several equally striking deletions. Recall, first, that in the Treatise Hume explicitly affirmed the simplicity and hence indefinability of the passions in general, and of the doublerelation passions in particular (T 2.1.2.1, p. 277; T 2.2.1.1, p. 329). In the Dissertation, in contrast, not only are these affirmations removed, but they are replaced with what, to all appearances, look like precisely the kinds of definitions he had earlier insisted were not possible: Pride is a certain satisfaction in ourselves, on account of some accomplishment or possession, which we enjoy: Humility, on the other hand, is a dissatisfaction with ourselves, on account of some defect or infirmity. Love or Friendship is a complacency in another, on account of his accomplishments or services: Hatred, the contrary. (P 2.1–2, p. 7) In the Treatise, pride was a certain simple and indefinable satisfaction, which happened as a matter of contingent fact to be causally related to self. In the Dissertation, however, it is defined as a complex perception, part satisfaction and part self. And similarly for the other three. Meanwhile, the two conflicting claims from Book 2—that the objects of these passions are at once their effects and (a part of) their causes—are both suppressed from Hume’s subsequent presentation of his view. What Hume’s definitions of the double-relation passions in the Dissertation most naturally suggest is an intrinsic account of their relation to their objects, but not along the lines that Qu proposes. For Qu, recall, the passions are simple impressions, whose intentionality comes apart from their qualitative character only through a ‘distinction of reason’. In this way, their dual nature and simplicity do not contradict Hume’s separability principle. But the present definitions suggest instead a more straightforward view: that these passions are now complex perceptions, part impression and part idea. However, this suggestion—though I think it is closer to the mark—should be tempered by consideration of a general difference between the Treatise and Hume’s later works. In the Dissertation and the first Enquiry, Hume’s taxonomy of impressions—which had

140  Part II previously appeared at the start of both Books 1 and 2—is entirely absent. The most that remains is the vaguest hint of the distinction between direct and indirect passions at P 2.1 (p. 7). Similarly, the distinction between simple and complex perceptions is almost entirely gone, leaving only an implicit trace of itself in Hume’s restatement of his copy principle, that all (simple) ideas are copied from a prior (simple) impression (E 2.5–6, pp.  19–20). Relatedly, the separability principle itself, which causes so much difficulty for Qu’s account, makes no reappearance.8 It would be overstating the case to say that Hume definitively abandoned any of these earlier principles and unambiguously embraced a different view of the nature of the passions. The available evidence does not seem to support so strong a conclusion. What the evidence does suggest, it seems to me, is a general weakening of Hume’s commitment to these earlier claims, and an attempt to extricate the conclusions that subsequently mattered to him most from these possible hostages to fortune. And in the particular case of the passions, the door is seemingly left open for the possibility that these are now somehow complex—and yet indivisible—­perceptions, which are essentially and not merely contingently connected to their objects.

7.5.  The Double Relation of Sentiments and Ideas Hume’s definitions of the double-relation passions in the Dissertation, as apparently comprising a feeling together with an idea of their object, give us some reason for walking through this door, at least with regard to these four passions. Having entered this far, meanwhile, it is tempting to go in all the way, and indeed we shall see that consideration of Hume’s theory of the double relation of impressions and ideas gives us a reason for extending this analysis to all the passions. This theory, with which Hume accounts for the origin of the double-relation passions, comprises three general principles: The first of these is the association of ideas, or that principle, by which we make an easy transition from one idea to another. However uncertain and changeable our thoughts may be, they are not entirely without rule and method in their changes. They usually pass with regularity, from one object, to what resembles it, is contiguous to it, or produced by it. . . The second property, which I shall observe in the human mind, is a like association of impressions or emotions. All resembling impressions are connected together; and no sooner one arises, than the rest naturally follow. Grief and disappointment give rise to anger, anger to envy, envy to malice, and malice to grief again. . . In the third place, it is observable of these two kinds of association, that they very much assist and forward each other, and that the

The Object of the Passions  141 transition is more easily made, where they both concur in the same object. Thus, a man, who, by an injury received from another, is very much discomposed and ruffled in his temper, is apt to find a hundred subjects of hatred, discontent, impatience, fear, and other uneasy passions; especially, if he can discover these subjects in or near the person, who was the object of his first emotion. (P 2.6–8, pp. 7–8; see also T 2.1.4.2–4, pp. 283–4) In establishing that these three principles together account for the origin of the double-relation passions, Hume spends relatively little time (in both Book 2 and the Dissertation) remarking that the causes of these passions are related either to ourselves (for pride and humility) or to another person (for love and hatred). He takes the relevance of the association of ideas to be obvious, and is principally concerned with proving the relevance of the association of impressions (and consequently the principle of mutual assistance), by arguing that the causes of these passions always first produce a separate pleasure (for pride and love) or pain (for humility and hatred). Hume’s emphasis on this point distracts from a more fundamental difficulty in applying this theory—and in particular the mutual assistance principle—to the origin of the double-relation passions. If the mutual assistance principle explains why one passion is more likely to give rise to another that resembles it—why anger, say, is more likely to give rise to envy—when the causes of these two passions are closely related, then it cannot also explain the origin of hatred (and equivalently for the other three double-relation passions). This is because the cause of hatred is not related to its passion in the same way that the causes of anger and envy are related to their passions. What the mutual assistance principle might explain, consistently with the general case, is the origin of the separate pain associated with hatred, but at this point in the train of emotions the pattern Hume is appealing to is broken. Anger as a result of someone mocking my singing might lead to envy of their own beautiful voice, which might lead to a displeasure at their tendency to laugh at others. In this we may recognize the operation of a single general principle. But going from here to a hatred of the person himself would then require a different principle, because the relation of this person to the passion of hatred is not, as in the other cases, that of cause to effect. And this difficulty arises, note, whichever side of the earlier dilemma Hume takes. Whether the idea of another is the effect of hatred, or a necessary but insufficient part of its cause, the pattern with the other passions such as anger and envy is still broken. This problem can be avoided (if not exactly solved) by stating the mutual assistance principle in terms of the objects of the passions rather than their causes, and assuming that the relationship between a passion and its object is independent of the relationship between a passion and

142  Part II its cause. For in this way the relevant pattern—the pattern involving the ­object—can be preserved in the special case of the double-relation passions, where the pattern involving the cause is different. I  say this avoids the problem more than solves it, because it requires us to take a lot of things for granted, and we are legitimately left wondering what this common passion-object relationship is, and whether we really have any grounds for supposing it exists at all. But what we have, at least, is the form of a theory that could work. If there is some common relationship that holds between anger and its object, envy and its object, and hatred and its object (even though in this last case, unlike with the others, the object is not also the cause), then that relationship could justify Hume’s application of the mutual assistance principle to the case of hatred (and likewise for pride, humility, and love). Hume’s presentations of the double-relation theory in the Treatise and the Dissertation are almost exactly identical, differing mainly in superficial points of style. There is only one substantial change, and it is to be found in Hume’s statement of the mutual assistance principle. In the Treatise, Hume writes that when a man is angry with someone, he ‘is apt to find a hundred subjects of discontent, impatience, fear, and other uneasy passions; especially if he can discover these subjects in or near the person, who was the cause of his first passion’ (T 2.1.4.4, p. 284; my emphasis). In the Dissertation, he says that the transition occurs when these subjects are discovered ‘in or near the person, who was the object of his first emotion’ (P 2.8, p. 8; my emphasis). Now, it could easily be that Hume had meant to say ‘object’ in the Treatise, and that ‘cause’ was simply a slip of the pen. But in the light of the argument just given, it is tempting to suppose that this was in fact a conscious change of mind. It is tempting, that is to say, to suppose that Hume recognized his principle of mutual assistance needed to be stated in terms of the objects of the passions rather than their causes. This would fit neatly with the introduction of his definitions of the double-relation passions as essentially involving the ideas of their objects. Only with this essential and non-causal intentionality, it would seem, can the causal mechanism of the double-relation theory do the work that Hume wanted it to do.

Notes 1 I presented an earlier version of this argument in Merivale (2009). My view is now rather different (albeit similar in spirit). Most significantly, my understanding of the double-relation theory, and of the mutual assistance principle in particular, has now changed (see §7.5). I was previously carried away by Hume’s claim that the two principles of association must ‘concur in the same object’, a phrase I believe I interpreted too literally. As a result, the problem I identify with this theory and the solution to it that seems to me to be suggested by Hume’s later work both look very different to me now. 2 The exact figure is 63 to 33. The trend here is quite general. In the Treatise as a whole, ‘impression’ wins by 616 to 236 (a ratio of more than five to two). Most

The Object of the Passions  143 occurrences of ‘impression’ are in Books 1 and 2, while most occurrences of ‘sentiment’ are in Book 3; in Books 1 and 2 by themselves the ratio is more than seven to one. In the Four Dissertations and the two Enquiries combined, meanwhile, ‘sentiment’ wins by an enormous 362 to 50 (a ratio of more than seven to one, the exact inverse of Books 1 and 2). In the two Enquiries alone the ratio is only fractionally smaller (261 to 37). These statistics do not measure exactly what we want: sometimes ‘impression’ is used in its ordinary sense (e.g. ‘[t]he inference is by no means just, that, because a system of religion has made no deep impression on the minds of a people, it must therefore have been positively rejected by all men of common sense’; N 12.19, p. 73), and sometimes ‘sentiment’ is used to mean opinion (see later in the chapter). Ideally, the data would be sifted and these instances discounted. Nevertheless, the figures are telling. 3 See e.g. Kenny (1963), Pitcher (1965), Solomon (1977), and Deigh (1994). 4 The first half of this conjunction is relatively clear: ‘Ideas always represent the objects and impressions, from which they are deriv’d’ (T 1.2.3.11, p. 37). The second half is subject to some controversy; see Garrett (2006) for a discussion and defence. Garrett also suggests that ‘moral and aesthetic sentiments may represent moral and aesthetic qualities’ (2006, p. 303), though this suggestion seems to me problematic, precisely in light of Hume’s anti-rationalism; see §8.5, note 5. However that may be, ideas represent and passions do not, and these two contrasting cases should suffice to give a sense of what is at stake in the present context. 5 Similar views on this matter can be found in Weller (2002), Schmitter (2008), and Qu (2012). I  will discuss Qu’s article further later. Cohon also agrees (1994, pp. 188–9), although Weller, curiously, cites her as an opponent. For additional confirmation of this interpretation, see §11.1, note 1. 6 See e.g. Árdal (1966, chs. 1, 2), Gardiner (1966, p.  37), Davidson (1976, p. 749), Garrett (2006, p. 303), McIntyre (2006, p. 210), and Schmitter (2008). 7 Deitl (1968), Sutherland (1976), Baier (1978), Inoue (2003), and Alanen (2006). 8 For a discussion of these cuts and their potential significance, see Millican (2002, pp. 48–52). Needless to say for those who know me, or have read the acknowledgements at the start of this book, my thinking on these matters owes a great deal to discussions with Millican.

Bibliography Alanen, Lilli (2006). ‘The Powers and Mechanisms of the Passions’. In The Blackwell Guide to Hume’s Treatise. Ed. by Saul Traiger. Oxford: Blackwell, pp. 179–98. Árdal, Páll S. (1966). Passion and Value in Hume’s Treatise. Edinburgh: Edinburgh University Press. Baier, Annette C. (1978). ‘Hume’s Analysis of Pride’. The Journal of Philosophy 75(1), pp. 27–40. ——— (1991). A Progress of Sentiments: Reflections on Hume’s Treatise. Cambridge, MA: Harvard University Press. Cohon, Rachel (1994). ‘On an Unorthodox Account of Hume’s Moral Psychology’. Hume Studies 20(2), pp. 179–94. Davidson, Donald (1976). ‘Hume’s Cognitive Theory of Pride’. Journal of Philosophy 73(19), pp. 744–57. Deigh, John (1994). ‘Cognitivism in the Theory of Emotions’. Ethics 104(4), pp. 824–54.

144  Part II Dietl, Paul (1968). ‘Hume on the Passions’. Philosophy and Phenomenological Research 28(4), pp. 554–66. Gardiner, P. L. (1966). ‘Hume’s Theory of the Passions’. In David Hume: A Symposium. Ed. By D. F. Pears. London: Palgrave Macmillan, pp. 31–42. Garrett, Don (2006). ‘Hume’s Naturalistic Theory of Representation’. Synthese 152(3), pp. 301–19. Inoue, Haruko (2003). ‘The Origin of the Indirect Passions in the Treatise: An Analogy Between Books 1 and 2’. Hume Studies 29(2), pp. 205–22. Johnson, Samuel (1755). A Dictionary of the English Language, in two volumes. London. Kames, Lord [Henry Home] (1762). Elements of Criticism, in three volumes. Edinburgh. Quotations and page references from Peter Jones, ed. (2005). Elements of Criticism, in two volumes. Indianapolis: Liberty Fund. Kenny, Anthony (1963). Action, Emotion and Will. London: Routledge & Kegan Paul. McIntyre, Jane (2006). ‘Hume’s “New and Extraordinary” Account of the Passions’. In The Blackwell Guide to Hume’s Treatise. Ed. by Saul Traiger. Oxford: Blackwell, pp. 199–215. Merivale, Amyas (2009). ‘Hume’s Mature Account of the Indirect Passions’. Hume Studies 35(1&2), pp. 185–210. Millgram, Elijah (1995). ‘Was Hume a Humean?’ Hume Studies 21(1), pp. 75–94. Millican, Peter (2002). ‘The Context, Aims, and Structure of Hume’s First Enquiry’. In Reading Hume on Human Understanding. Ed. by Peter Millican. Oxford: Oxford University Press, pp. 27–65. Penelhum, Terence (1975). Hume. London: Palgrave Macmillan. ——— (2009). ‘Hume’s Moral Psychology’. In The Cambridge Companion to Hume. Eds. by David Fate Norton and Jacqueline Taylor. Cambridge: Cambridge University Press, pp. 238–69. ——— (2015). ‘The Indirect Passions, Myself, and Others’. In The Cambridge Companion to Hume’s Treatise. Eds. by Donald C. Ainslie and Annemarie Butler. Cambridge: Cambridge University Press, pp. 206–29. Phillips, David (2005). ‘Hume on Practical Reason: Normativity and Psychology in Treatise 2.3.3’. Hume Studies 31(2), pp. 299–316. Pitcher, George (1965). ‘Emotion’. Mind, New Series 74(295), pp. 326–46. Prinz, Jesse J. (2004). Gut Reactions: A Perceptual Theory of Emotion. Oxford: Oxford University Press. Qu, Hsueh (2012). ‘The Simple Duality: Humean Passions’. Canadian Journal of Philosophy, 42(supp 1), pp. 98–116. Schmitter, Amy M. (2008). ‘Making an Object of Yourself: On the Intentionality of the Passions in Hume’. In Topics in Early Modern Philosophy of Mind. Ed. by J. Miller. Dordrecht: Springer, pp. 223–240. Solomon, Robert C. (1977). ‘The Logic of Emotion’. Noûs 11(1), pp. 41–9. Sutherland, Stewart R. (1976). ‘Hume on Morality and the Emotions’. The Philosophical Quarterly 26(102), pp. 14–23. Weller, Cass (2002). ‘The Myth of Original Existence’. Hume Studies 28(2), pp. 195–230. Zemach, Eddy M. (2001). ‘What is Emotion?’ American Philosophical Quarterly 38(2), pp. 197–207.

8 The Combat of Passion and ­Reason

Undoubtedly the most famous thing that Hume wrote about the passions is that ‘[r]eason is, and ought only to be [their] slave’ (T 2.3.3.4, p. 415). This remark is a victim of its own success. Moral philosophers since have been so taken with the project of either confirming or refuting it, that the discussion in which it appears has become increasingly divorced from its original context. For example, Charles Pigden’s recent collection devoted to this topic contains only two passing references to Samuel Clarke, and no mention at all of Nicolas Malebranche, though these were Hume’s principal targets (Pigden 2009; the references to Clarke are on pages 7 and 213, where his name is mentioned, but nothing more). Its contributors seem content to study Hume’s argument without so much as a glance in the direction of the views that it was intended to refute. In this ­chapter, I  mean to inject some much-needed historical context back into this discussion, examining first the views of Hume’s opponents Malebranche and Clarke (§8.1 and §8.2), and then of his ally Hutcheson (§8.3). Having done so, making sense of Hume himself will be a comparatively straightforward matter (§8.4 and §8.5).

8.1.  To Give the Preference to Reason In §2.4, I mentioned Hutcheson’s criticism of Clarke’s moral rationalism, but set it to one side, the better to focus on the egoism debate. Now is the time to pick it up again and look at it more closely. To understand the view that Hume followed Hutcheson in opposing, however, the best place to start—according to Hume’s own account of the matter—is not with Clarke, but with Malebranche: ‘Malebranche, as far as I can learn, was the first that started this abstract theory of morals, which was afterwards adopted by Cudworth, Clarke, and others’ (M 3.n12.1, p. 197). Malebranche’s two major influences were St. Augustine and René Descartes (see Nadler 2006, p. 1). From the latter he inherited—among many other things—the doctrine of the real distinction between mind and body, and its attendant conception of the mind as a non-extended, thinking thing. From the former he inherited—again among other things—a

146  Part II neo-Platonic conception of the doctrine of original sin, according to which God is our only true good and is to be loved above all things, the corporeal world is to be shunned as leading us away from this sole good, and our close connection with our bodies (which draws us so strongly in the wrong direction) is God’s punishment to all mankind for Adam’s eating of the apple. It is the Augustinian strand of Malebranche’s thought that is of interest here, for it is this that dominates his treatment of the passions. Furthermore, it was precisely this sort of view that Hume was reacting against in his famous discussion of the combat of passion and reason. Malebranche’s most extended examination of the passions is to be found in book 5 of his De la Recherche de la Vérité (1675), but for a slightly fuller picture we must begin with book 4. This preceding book discusses what Malebranche calls the ‘natural inclinations’ of the mind, which are the purely mental aspects of the will, untainted by bodily interference. They stand to the passions as thought stands to sensation or imagination (1675, p. 337). Because Malebranche’s God loves himself above all else, and because he creates our wills in the image of his own, Malebranche concludes that our primary natural inclination must be towards God or his glory: It is an unquestionable truth that God can have no other principal end for His operations than Himself.  .  .  . Since the mind’s natural inclinations are undoubtedly the constant impressions of the Will of Him who has created and preserves them, it seems to me that these inclinations must be exactly like those of their Creator and Preserver. By their very nature, then, they can have no other principal end than His glory. (1675, p. 226) Since God’s creation participates in his goodness, however, God may have as a secondary end the preservation of his creatures: ‘God wills His glory, then, as His principal end and the preservation of His creatures only for His glory’ (1675, p. 226). Again, our natural inclinations follow suit: ‘since God unceasingly imprints in us a love like His own . . . He also provides us with all those natural inclinations that . . . necessarily dispose us toward preserving our own being and the being of those with whom we live’ (1675, p. 267). For reasons that are somewhat obscure (or which at any rate have so far escaped my understanding), God’s self-love manifests itself in us as ‘the love of the good in general’ (1675, p. 267), rather than as the love of God (though for Malebranche, I take it, God is the only truly good thing, which perhaps goes some way towards explaining the switch). Whatever its basis, this love that we have of the good in general does not force us to love only what is in fact good; we cannot love what we do not think is good, but God has left us free to make mistakes in this regard: ‘the power

The Combat of Passion and Reason  147 of loving badly, or rather of loving well what we should not love at all, depends on us, because as free beings we can and do in effect determine toward particular, and consequently false, goods the good love that God unceasingly imprints in us’ (1675, p. 267). The passions are for Malebranche the chief causes of precisely these sorts of errors. He does allow that they have a useful purpose, namely the preservation of the body, by directing us towards things that are good for it: The senses and the passions were given to us only for the good of the body. Sensible pleasure is the mark that nature has attached to the use of certain things in order that without having to bother with a rational examination we might use them for the preservation of the body. (1675, p. 359) On the whole, however, it would seem that for Malebranche the passions do us more harm than good. The inclination that they give us towards the good of the body might have been fine by itself, if only we could remember that ‘we are not our body; it is a thing belonging to us’ and that ‘[t]he good of our body is therefore not our good’ (1675, p. 359). But since the Fall of Adam, ‘God has withdrawn from us’ (1675, p. 360), weakening our connection with him and what is truly good, and strengthening our tie to our bodies. As a result of God’s departure, we are now in constant peril of such sinful things as the love of strawberries (my example, not Malebranche’s, but this really does seem to be his position). For we mistakenly think that strawberries (or whatever it might be) are the causes of our pleasure, not realizing that they are merely the occasions on which God chooses to give us pleasure. Moreover, we mistakenly think that they are good for us and not just good for our bodies. And so we direct our love where it should not be directed (e.g. at strawberries), thereby exciting God’s righteous anger: We can and must love what is capable of making us sense pleasure, granted. But it is for this reason that we must love only God, because only God can act in our soul and because sensible objects can do no more than move our sense organs. But what difference does it make, you will ask, where these pleasant sensations come from? I only want to enjoy them. Ingrate, take a look at the hand heaping goods upon you! You exact unjust rewards from a just God; you would have Him reward you for the crimes you commit against Him, and at the same time you commit them. . . . But death will corrupt this body, and God, whom you have made serve your unjust desires, will mock you when His turn comes and then will make you serve His own just wrath. (1675, pp. 359–60)

148  Part II The mind, for Malebranche, lies ‘between God and body, between good and evil, between that which enlightens it and that which blinds it, that which sets it in order and that which disrupts it, between that which can make it perfect and happy and that which can make it unhappy and imperfect’ (1675, p. 363). The passions, as we have seen, are the emissaries of the body. They drag us away from God and true felicity; they ‘involve us in error with regard to the good’, and ‘they must be resisted continuously’ (1675, p.  357). Reason, by contrast, is the emissary of God. The free consent of our wills must therefore shun the passions and follow instead our God-given reason: This consent must be regulated and kept free in spite of all the efforts of the passions. Only to God should it subjugate its freedom; it should surrender only to the voice of the Author of nature, to inner certainty, to the secret reproaches of reason. (1675, p. 357) Hume’s famous discussion of reason and passion in the Treatise (Of the influencing motives of the will, T 2.3.3, pp. 413–8) begins with an outline of the view that he is opposing: Nothing is more usual in philosophy, and even in common life, than to talk of the combat of passion and reason, to give the preference to reason, and assert that men are only so far virtuous as they conform themselves to its dictates. . . . The eternity, invariableness, and divine origin of the former have been display’d to the best advantage: The blindness, unconstancy, and deceitfulness of the latter have been as strongly insisted on. (T 2.3.3.1, p. 413) As I trust the previous summary has shown, Malebranche was firmly in Hume’s sights here. In his endeavouring ‘to show the fallacy of all this philosophy’ (ibid.), we should remember that Hume’s interests were not purely metaethical, as many recent commentaries might suggest. His target was a stern religious ethics, bound up with a heavily theistic conception of human reason, and—at least in the case of Malebranche—with the odious doctrine of original sin.

8.2.  Virtue Is Nothing but a Conformity to Reason Hume’s other major target was Samuel Clarke, whose views on this matter are expounded at the start of his Discourse on the Unchangeable Obligations of Natural Religion (the second of his two sets of Boyle Lectures, delivered in 1705 and first published in 1706). The essence of Clarke’s position, as stated in the contents, is this:

The Combat of Passion and Reason  149 That from the Eternal and Necessary Differences of Things, there naturally and necessarily arise certain Moral Obligations, which are of themselves incumbent on all Rational Creatures, antecedent to all positive Institution, and to all Expectation of Reward and Punishment. (1706, p. 588) That these obligations are incumbent on all rational creatures, antecedent to any expectation of reward and punishment, is of course an antiegoist view; and Clarke devotes much of his discussion to a criticism of Hobbes. These arguments of his are not of present interest, however, first because it is his rationalism rather than his anti-egoism that we are concerned with here, and second because Hume was also an anti-egoist (at least after the Treatise), as we have already seen at length. The summary proposition just quoted refers to the necessary differences of things as the source of moral obligations, but when we get down to the details Clarke also speaks more simply of ‘the nature of things’. One might safely assume that the differences between things simply follow from their individual natures, so that on one level this distinction does not really matter. But the distinction does seem to be playing a role in Clarke’s account, since natures and differences appear to ground distinct moral obligations. Thus we are told first that ‘there is a Fitness or Suitableness of certain Circumstances to certain Persons, and an Unsuitableness of others; founded in the nature of Things’; and then that ‘from the different relations of different Persons one to another, there necessarily arises a fitness or unfitness of certain manners of Behaviour of some Persons towards others’ (1706, p. 608). The immediate example given to illustrate this latter kind of obligation is theological: That God is infinitely superior to Men; is as clear, as that Infinity is larger than a Point, or Eternity longer than a Moment: And ’tis as certainly Fit, that Men should honour and worship, obey and imitate God, rather than on the contrary in all their Actions indeavour to dishonour and disobey him. (1706, p. 608) To illustrate the former kind of obligation, meanwhile, Clarke writes: In like manner; in Mens dealing and conversing one with another; ’tis undeniably more Fit, absolutely and in the Nature of the thing itself, that all Men should indeavour to promote the universal good and welfare of All; than that all Men should be continually contriving the ruin and destruction of All. (1706, p. 609)

150  Part II However they are founded—whether on the differences or just the natures of things—these ‘fitness’ facts are at the core of Clarke’s moral philosophy. To begin with, Clarke offers no argument for this highly dubious deduction of the fitness of certain sorts of behaviour from the differences or natures of things. He simply asserts that ‘[t]hese things are so notoriously plain and self-evident, that nothing but the extremest stupidity of Mind, corruption of Manners, or perverseness of Spirit, can possibly make any Man entertain the least doubt concerning them’ (1706, p. 609). He goes on to accuse Hobbes of some such failing. Later, however, when he turns to examine some particular moral obligations, we do get more by way of argument. For example, he offers the following justification for the principle of equity, that every man must ‘deal always with another, as he would reasonably expect that Others should in like Circumstances deal with Him’ (1706, p. 619): Whatever relation or proportion one Man in any Case bears to another; the same That Other, when put in like Circumstances, bears to Him. Whatever I judge reasonable or unreasonable for another to do for Me; That, by the same Judgment, I declare reasonable or unreasonable, that I in the like Case should do for Him. And to deny this either in Word or Action, is as if a Man should contend, that, though two and three are equal to five, yet five are not equal to two and three. (1706, p. 619) There are further arguments for some additional example obligations, but an examination of these would involve us in more detail than is presently required. For what it’s worth, this argument—which has a familiar rationalist appeal—strikes me as unsuccessful. The most that it can show is that one judgement concerning what is reasonable or unreasonable follows from another such judgement. But this is beside the point: the question is where the first judgement came from, and whether, in particular, it can be deduced from the eternal differences or natures of things. Humeans of course deny that any such deduction is possible. The final piece of Clarke’s view is that the fitness of certain sorts of behaviour gives rise to an obligation on all rational agents to behave in that way. The argument at this point again strikes me as far from persuasive: Wherefore all rational Creatures, whose Wills are not constantly and regularly determined, and their Actions governed, by right Reason . . . but suffer themselves to be swayed by unaccountable arbitrary Humours, and rash Passions; by Lusts, Vanity and Pride; by private Interest, or present sensual Pleasures; These, setting up their

The Combat of Passion and Reason  151 own unreasonable Self-will in opposition to the Nature and Reason of Things, endeavour (as much as in them lies) to make things be what they are not, and cannot be. Which is the highest Presumption and greatest Insolence, as well as the greatest Absurdity, imaginable. ’Tis acting contrary to that Understanding, Reason and Judgment, which God has implanted in their Natures on purpose to enable them to discern the difference between good and evil. (Clarke, 1706, pp. 613–4) First, it is unclear why agents who indulge their rash and lusty passions, however unreasonable this behaviour may be, are thereby endeavouring ‘to make things be what they are not, and cannot be’. That might be their intention, I suppose, but surely most of the time nothing is further from their (our) minds. Second, there is the merest assertion, without any supporting argument, that such an endeavour is immoral. This assertion is hard to credit, however, not least because the supposed endeavour is necessarily—and obviously—doomed to failure. The contrast in the previous quotation between ‘right Reason’ and ‘rash Passions’, and the identification of the former as the source of all good behaviour and the latter of all evil, is a general theme throughout this section. God, in Clarke’s design, cannot but act reasonably; lesser intelligent beings, however, may ‘wilfully and perversely allow themselves to be over-ruled by absurd Passions, and corrupt or partial Affections’ (1706, p. 612). Such ‘wilful Passions or Lusts’, together with negligent misunderstanding, ‘are . . . the only Causes which can make a reasonable Creature act contrary to Reason’ (1706, p. 613).

8.3.  Reason Alone Can Never Be a Motive As noted in §2.4, Hutcheson’s first work, An Inquiry into the Original of Our Ideas of Beauty and Virtue, prompted an objection from the rationalist Gilbert Burnet, who appealed to the views of Cudworth, Clarke, and Wollaston. This objection, in turn, prompted Hutcheson’s subsequent Illustrations on the Moral Sense, with its sentimentalist attack on moral rationalism. This attack, as we will see in the next section, gave Hume much material for his own views on the matter. Hutcheson’s first move is to distinguish between moral motivation and moral approbation, and consequently between what he calls exciting and justifying reasons: When we ask the Reason of an Action we sometimes mean, ‘What Truth shews a Quality in the Action, exciting the Agent to do it?’ Thus, why does a Luxurious Man pursue Wealth? The Reason is given by this Truth, ‘Wealth is useful to purchase Pleasures.’ Sometimes for a Reason of Actions we shew the Truth expressing a

152  Part II Quality, engaging our Approbation. Thus the Reason of hazarding Life in just War, is, that ‘it tends to preserve our honest Countrymen, or evidences publick Spirit:’ The Reason for Temperance, and against Luxury is given thus, ‘Luxury evidences a selfish base Temper.’ The former sort of Reasons we will call exciting, and the latter justifying. (1728, p. 138) This would seem to be a version of the distinction between reasons that explain why someone does something, and reasons that show why someone should do something. Notable modern variants include the popular distinction between motivating and normative reasons (Smith 1994, pp. 94–8; Parfit 1997, p. 99; Dancy 2000, pp. 1–5), and Bernard Williams’s distinction between internal and external reasons (Williams 1981).1 Hutcheson’s criticism of moral rationalism proceeds on these two separate fronts. I will restrict my attention here to the first front, the case of exciting reasons. Although there is of course a close connection between the two, exciting reasons and motivation belong within the philosophy of emotion, while justifying reasons and approbation are more a topic for moral philosophy. This same division is also reflected very clearly in Hume’s own texts, where the issue of exciting reasons is taken up in the Dissertation on the Passions (P 5, pp. 24–5), and that of justifying reasons in the moral Enquiry (M App1, pp. 285–94). As space here is limited, and the latter would take me too far from my present concern, I will focus on the Dissertation topic. (When turning to Hume himself, I will be forced to glance in the direction of morals and approbation; but still my primary focus is on motivation.) Regarding exciting reasons or motivation, then, Hutcheson’s core argument is very simple: As to exciting Reasons, in every calm rational Action some end is desired or intended; no end can be intended or desired previously to some one of these Classes of Affections, Self-Love, Self-Hatred, or desire of private Misery, (if this be possible) Benevolence toward others, or Malice: All Affections are included under these; no end can be previous to them all; there can therefore be no exciting Reason previous to Affection. (1728, p. 139) By a ‘calm rational Action’, Hutcheson means one in which reason and reflection are involved, as opposed to one in which passion alone moves us suddenly and thoughtlessly (see below). By urging that the passions or affections are necessary even in these cases (to fix the ends of action), therefore, he is urging that they are necessary in all cases.

The Combat of Passion and Reason  153 What Hutcheson is offering here is the first clear presentation of what is now widely known (rather unfairly) as the ‘Humean’ theory of motivation. According to this account, there is no motivation without some end which the action is believed to further somehow (as a probable means to that end, a constitutive part, or whatever it might be); and there is no end without some affection or desire towards it. As the formula goes, motive equals belief plus desire.2 On this view, reason and sentiment both play a necessary role, and accordingly Clarke’s talk of the combat between passion and reason is dismissed as incoherent: We have indeed many confused Harangues on this Subject, telling us, ‘We have two Principles of Action, Reason, and Affection, or Passion (i.e. strong Affection): the former in common with Angels, the latter with Brutes: No Action is wise, or good, or reasonable, to which we are not excited by Reason, as distinct from all Affections; or, if any such Actions as flow from Affections be good, ’tis only by chance, or materially and not formally.’ As if indeed Reason, or the Knowledge of the Relations of things, could excite to Action when we proposed no End, or as if Ends could be intended without Desire or Affection. (1728, p. 139) Hutcheson accounts for this confusion by acknowledging that particular violent affections can sometimes prevent us from thinking clearly: Perhaps what has brought the Epithet Reasonable, or flowing from Reason, in opposition to what flows from Instinct, Affection, or Passion, so much into use, is this, ‘That it is often observed, that the very best of our particular Affections or Desires, when they are grown violent and passionate, thro’ the confused Sensations and Propensities which attend them, do make us incapable of considering calmly the whole Tendency of our Actions, and lead us often into what is absolutely pernicious, under some Appearance of relative or particular Good.’ This indeed may give some ground for distinguishing between passionate Actions, and those from calm Desire or Affection which employs our Reason freely: But can never set rational Actions in Opposition to those from Instinct, Desire or Affection. (1728, p. 175) Here we see why Hutcheson restricted his attention earlier to cases of ‘calm rational Action’. Even in these cases, where reason is operating at its full, the agent must still have some end that is the object of their desire. Affection or sentiment must always play a role. The debate between Hutcheson and the rationalists was about how to divide up the mind, and how to conceptualize human agency at the most basic level. Malebranche and Clarke were operating with a hierarchical

154  Part II division of the mind, with reason (mind, spirit) firmly on the top, and passion (body, matter) firmly on the bottom. Each of these two components, moreover, formed for them a complete and fully functioning unit, capable of acting on its own account. For these thinkers, there is thus a constant battle in the soul between our higher self and our lower self, our angel and our devil. Hutcheson, however, preferred what we might call a specialization view. On this understanding, there is no hierarchy: the rational and passionate components are on the same level. But more than this, the two components are distinguished, not by their different position on the scale, but by their different psychological roles, with each one having a necessary and indispensable part to play in motivation: roughly, reason to gather the information, and sentiment to set the goals. Hutcheson (and Hume after him) also thought, of course, that sentiment was responsible for evaluating the goals that it set. This is not to say that reason has no role to play here; but the final verdict is down to feeling. As Hume neatly puts it in the moral Enquiry: The final sentence . . . which pronounces characters and actions amiable or odious, praise-worthy or blameable . . . depends on some internal sense or feeling, which nature has made universal in the whole species. . . . But in order to pave the way for such a sentiment, and give a proper discernment of its object, it is often necessary, we find, that much reasoning should precede, that nice distinctions be made, just conclusions drawn, distant comparisons formed, complicated relations examined, and general facts fixed and ascertained. (M 1.9, pp. 172–3) But though reason, when fully assisted and improved, be sufficient to instruct us in the pernicious or useful tendency of qualities and actions; it is not alone sufficient to produce any moral blame or approbation. Utility is only a tendency to a certain end; and were the end totally indifferent to us, we should feel the same indifference towards the means. It is requisite a sentiment should here display itself, in order to give a preference to the useful above the pernicious tendencies.  .  .  . Here, therefore, reason instructs us in the several tendencies of actions, and humanity makes a distinction in favour of those which are useful and beneficial. (M App1.3, p. 286) It seems to me that the question of approbation is distinct from that of motivation—although the two are of course very closely related—and as I have said my focus here is on the latter more than the former. A proper examination of Hume’s views on moral approbation is a large project in its own right, not attempted here.

The Combat of Passion and Reason  155 To emphasize the distinctness of these two issues, notwithstanding their close relationship, note that it is possible to agree with Hutcheson and Hume in the motivational case, while disagreeing in the case of approbation. It is possible, that is to say, to hold that reason and sentiment must work together to generate motivation, in the way that Hutcheson proposed, but to think that reason alone is sufficient to produce moral approbation and disapprobation. The thought here, roughly, would be that there is information available about what is morally right and wrong, and that reason can deliver this information. This, very approximately, is the view adopted in recent times by Michael Smith (1994). Hutcheson and Hume, though they both very clearly rejected this view, did so for rather different reasons, and Hume’s rejection had considerably more sceptical implications. Hume rejected it, or so it seems to me, because he believed that there was simply no such information to be had. Hutcheson, on the other hand, rejected this view because he thought that there was such information, but that we arrived at it through the direct perception of our moral sense, rather than through reason. I am not at all sure that Hutcheson’s view on this matter is coherent; but that is a large question for another day.

8.4.  Where They Neither Excite Desire Nor Aversion The question of approbation aside, Hume’s anti-rationalism with regard to motivation was not significantly different from Hutcheson’s. It also remained the same throughout his life. Given what we have seen of Hutcheson in the previous section, therefore, making sense of Hume’s remarks on this matter—both in section 2.3.3 of the Treatise and in section 5 of the Dissertation—is now a fairly straightforward matter. Section  2.3.3 of the Treatise begins with an introductory paragraph calling attention to the talk of the combat of passion and reason, and the tendency of many philosophers and moralists—Hume evidently had Malebranche and Clarke in mind, as we have seen—to give the preference to reason. Hume then announces his intention ‘to shew the fallacy of all this philosophy’ (T 2.3.3.1, p. 413). Section 5 of the Dissertation omits this introduction, diving straight into the main point: It seems evident, that reason, in a strict sense, as meaning the judgment of truth and falsehood, can never, of itself, be any motive to the will, and can have no influence but so far as it touches some passion or affection. Abstract relations of ideas are the object of curiosity, not of volition. And matters of fact, where they are neither good nor evil, where they neither excite desire nor aversion, are totally indifferent; and whether known or unknown, whether mistaken or rightly apprehended, cannot be regarded as any motive to action. (P 5.1, p. 24; see also T 2.3.3.2–3, pp. 413–4)

156  Part II Hume doesn’t explicitly say here—as Hutcheson did—that passions or desires are necessary to set the ends or goals of action, but this is implicit in his account of motivation (already examined in Chapters 3 and 4). Incidentally, the distinction appealed to here—between relations of ideas and matters of fact, which Hume had earlier drawn in the first Enquiry (E 4.1–2, pp. 25–6)—was also anticipated by Hutcheson in the context of the reason/passion debate: Reason denotes either our Power of finding out Truth, or a collection of Propositions already known to be True. Truths are either Speculative, as ‘When we discover, by comparing our Ideas, the Relations of Quantities, or of any other Object, among themselves;’ or Practical, as ‘When we discover what Objects are naturally apt to give any Person the highest Gratifications or what means are most effectual to obtain such Objects.’ (Burnet & Hutcheson 1735, p. 18) The truths on Hutcheson’s practical side in this quotation are perhaps surprisingly narrow (as Hume’s matters of fact are not), but that was presumably just because of the motivational context of the argument, in which truths about means to given ends are particularly salient.3 At this point in the Treatise, Hume says that, ‘[a]s this opinion may appear somewhat extraordinary, it may not be improper to confirm it by some other considerations’ (T 2.3.3.4, p. 415). He then produces his famous representative quality argument (familiar from §7.2), and two follow-on paragraphs in the same vein (T  2.3.3.5–7, pp.  415–7). This material is not repeated in section 5 of the Dissertation, and I consider this fact to be significant. I will have more to say about it in the next section, but for now let us stay with the common aspects of the earlier and later discussions. Hume’s next move, in both the Treatise and the Dissertation, is to offer an explanation or reinterpretation of the talk of the combat of passion and reason, one that is consistent with his specialization view of the relationship between these two faculties. Here again, Hume apparently drew on Hutcheson for inspiration, for his account similarly turns on a distinction between calm and violent passions. The specifics of Hume’s suggestion, however, are different. Whereas Hutcheson had claimed that the violent passions sometimes prevent us from thinking clearly, and so to that extent do genuinely interfere with reason, Hume says instead that his opponents have confounded reason with calm desire, because these two things feel the same (P  5.2, p.  24; T  2.3.3.8, p.  417). This is the springboard for his consequent discussion of the causes of the violent passions, which I will turn to in the next chapter. Section  5 of the Dissertation and section  2.3.3 of the Treatise end with essentially the same paragraph, save for a few insignificant stylistic changes. In this paragraph Hume draws attention to one further point

The Combat of Passion and Reason  157 that Hutcheson didn’t make: ‘The common error of metaphysicians has lain in ascribing the direction of the will entirely to one of these principles [i.e. calm desires or violent passions], and supposing the other to have no influence’; but, says Hume, people in general are moved both by ‘the view of the greatest possible good’ and by ‘present uneasiness’ (P 5.4, p. 24; T 2.3.3.10, p. 418). This point is perhaps a little puzzling, first because it just seems so obviously true, but also because neither Malebranche nor Clarke had ever denied it. Their position was not that we are moved by only one of the competing principles (a position that would after all make a nonsense of the idea that there is a constant combat between them), but that we should be moved by only one of them. So far as I am aware, no one has yet remarked on the curious nature of this concluding paragraph; attention, no doubt, has been drawn away by the other provocative things that Hume says in this context, especially in the Treatise presentation. Perhaps I thus have the honour of venturing the first explanation. In any case, the only way that I have been able to make sense of the comment is to suppose that it is aimed, not at Clarke or Malebranche, but at Locke. For Locke appears to have oscillated on precisely this point: in the first edition of his Essay, he wrote that ‘the greater Good is that alone which determines the Will’ (1690, p. 251n); but in the second edition he said that ‘upon second thoughts I am apt to imagine [that what determines the Will] is not, as is generally supposed, the greater good in view: But some (and for the most part the most pressing) uneasiness a Man is at present under’ (1694, pp.  250–1). Hume’s midway position on this matter seems obvious; but I suppose if Locke had swung from one extreme to the other, then the obvious point needed to be made. This solution to the puzzle is not perfect, for Hume describes the view that he is rejecting as a common error of metaphysicians, and presents his rejection in the context of an argument that is obviously directed at Malebranche and Clarke. Thus we should expect it to be a point of agreement among all or most rationalists, rather than an error peculiar to Locke. Nevertheless, it is the most satisfying explanation that I have been able to come up with; I would gladly be persuaded by a better one.

8.5.  ’Tis Not Contrary to Reason Section 5 of the Dissertation does not contain anything that was not already present in section 2.3.3 of the Treatise. However, when examining the former in the light of the latter, one cannot help but be struck by the very glaring omissions: all the best-known parts of the Treatise discussion have disappeared. This includes the representative quality argument already looked at in §7.2, the notorious claim that ‘[r]eason is, and ought only to be the slave of the passions’ (T 2.3.3.4, p. 415), and the several provocative illustrations of the anti-rationalist position:

158  Part II ’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger. ’Tis not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me. ’Tis as little contrary to reason to prefer even my own acknowledg’d lesser good to my greater, and have a more ardent affection for the former than the latter. (T 2.3.3.6, p. 416) I suppose that the standard thought, insofar as this curious fact has been noticed at all, is that Hume made these cuts simply in order to keep things short. But a little reflection shows this explanation to be most unlikely. Section 2.3.3 of the Treatise was not long to begin with. If section 5 of the Dissertation had simply repeated it in its entirety, it would still have been the shortest section in this later work. As it is, section 5 is radically shorter than all the other sections, and less than half the length of its Treatise ancestor. Considerations of space are thus not a plausible explanation, and we must look for some other reason for Hume’s drastic castration of his earlier discussion. The conclusion of Hume’s first argument in this section of the Treatise, the argument that is repeated in the Dissertation, is that ‘reason alone can never produce any action, or give rise to volition’ (T 2.3.3.4, p. 414); or that ‘reason  .  .  . can never, of itself, be any motive to the will, and can have no influence but so far as it touches some passion or affection’ (P 5.1, p. 24). This is very clearly a point about motivation, as I was of course assuming in the previous section. It is not, to be explicit, a point about approbation; and although Hume held an exactly parallel view in the case of approbation, and these two anti-rationalist theses are of course very closely related, they are nevertheless distinct claims (as I have already insisted). In Hume’s later work, the distinction between motivation and approbation is firmly in his grasp, and he tackles the two topics separately. The question of approbation first comes up, after the Treatise, in the essay The Sceptic, in which Hume makes his anti-rationalist stance very clear (at least if we take him to be in agreement with his Sceptic on this point, which I do). It then comes up again, of course, in the moral Enquiry. The question of motivation, meanwhile, is addressed in the Dissertation on the Passions, which is entirely silent on the matter of approbation. This is all relatively clear and straightforward. In section 2.3.3 of the Treatise, however, Hume rather confusingly runs these two things together, attempting to take on his rationalist opponents on both fronts simultaneously. The representative quality argument is his case for denying that reason alone gives rise to approbation, its conclusion being that passions cannot be ‘contradictory to truth and reason’ (T 2.3.3.5, p. 415), ‘contrary to reason’ (T 2.3.3.6, p. 416), or ‘unreasonable’ (T 2.3.3.7, p. 416), except in a trivial and uninteresting sense (i.e.

The Combat of Passion and Reason  159 when they are founded on some false belief). I suppose one might be able to interpret this as a conclusion about motivation rather than approbation, but that would be a fairly serious strain on the words. The obvious interpretation of Hume’s thinking in these passages is that reason does play an evaluative role in human psychology—it can evaluate beliefs as either true or false—but that this evaluative role does not extend to classifying passions and desires as morally right or wrong. One very good reason why Hume should have removed this argument from the Dissertation on the Passions, then, is that it doesn’t belong there: it belongs, if anywhere, in the moral Enquiry. But this cannot be the whole explanation, because, most curiously, it isn’t in the moral Enquiry either. Hume didn’t just move the representative quality argument in his later work; he abandoned it. And he abandoned it, I believe, for a very good reason, namely that it isn’t a very good argument. To see this, let us take a look at the argument again, in as clear a form as possible: For something to be contrary to reason is for it to be a misrepresentation of what it represents. Passions do not represent anything. Therefore, passions cannot be contrary to reason. The argument is obviously valid. It may even be sound. Nevertheless, it is a terrible argument for anti-rationalism, because the premises simply assume—without attempting to explain or justify—far too much of what is to be proved. As arguments go, it amounts to little more than a bare assertion of the claim to be established. Given the obvious validity of the argument, rationalists wishing to escape its conclusion must deny at least one of the premises. But why shouldn’t they do this? Hume has given us little or nothing in support of either. It is agreed on both sides of this debate that passions and desires can be morally right or wrong. The question concerns not the reality of this distinction, but its basis.4 Clarke says that reason alone determines which side of the line things fall on, while Hume insists that sentiment has the final say. Suppose that you agree with Clarke, and that you agree with the second premise of Hume’s argument (that the passions are not representations). In this case you will of course deny that the only way for something to be contrary to reason is for it to be a misrepresentation. You will say that things can be contrary to reason either when they are misrepresentations or when they are morally wrong. Or suppose instead that you agree with Hume’s first premise; now you will obviously deny the second. You will say that passions represent moral reality. This may sound implausible for passions like anger or benevolence, but it is not obviously wrong to say that love and hatred, for example, represent their

160  Part II objects as virtuous and vicious respectively. And it is a small step from here to the moral sentiments themselves. Hume and his fellow Humeans will not accept either of these rationalist responses to the argument, of course, and there is nothing in these considerations to persuade them that they should.5 But that is not the point. The point is that there is nothing in Hume’s argument to persuade anyone that they shouldn’t. The argument simply doesn’t touch on the controversy at all. Hume helps himself to his preferred views of reason and representation in the premises, from which his anti-rationalist conclusion trivially follows. But the debate concerns whether these are indeed the right views of reason and representation in the first place. And so I suggest that it was good philosophical judgement, on Hume’s part, that prompted him to supress this argument, both in the Dissertation on the Passions and in the moral Enquiry.

Notes 1 It seems to me, incidentally, that Williams set up the question in the wrong way, by asking whether there are any external reasons (and notoriously answering in the negative). At least in the eighteenth century, however, the question dividing sentimentalists and rationalists concerned not the existence but the basis of external (normative, justifying) reasons: are they grounded in reason, in perceptions of the moral facts, or in sentiment, in non-rational feelings of approval and disapproval? See Blackburn (2010). 2 See e.g. Smith (1994, pp. 92–125), who offers in essence the same argument as Hutcheson did for the ‘Humean’ theory. 3 The distinction between relations of ideas and matters of fact is commonly referred to as ‘Hume’s Fork’. For a recent and very thorough discussion of it, see Millican (2017). 4 Following the uncharitable interpretation of the Treatise as ‘sapping the Foundations of Morality, by denying the natural and essential Difference betwixt Right and Wrong, Good and Evil’ (L 19, p. 425), Hume was at pains to make this point very clear in the introduction to the moral Enquiry (M 1.2–3, pp. 169–70). 5 It is for this reason that I hesitate to follow Garrett’s suggestion that the moral sentiments may represent the properties of virtue and vice (Garrett 2006, p. 303; recall §7.2, note 4). If Hume were to endorse this claim, while retaining his views on reason and representation, that would seem to undermine his sentimentalism.

Bibliography Blackburn, Simon (2010). ‘The Majesty of Reason’. Philosophy 85(1), pp. 5–27. Burnet, Gilbert and Francis Hutcheson (1735). Letters concerning the True Foundation of Virtue or Moral Goodness. London. Clarke, Samuel (1706). A Discourse Concerning the Unchangeable Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation. London. Dancy, Jonathan (2000). Practical Reality. Oxford: Oxford University Press.

The Combat of Passion and Reason  161 Garrett, Don (2006). ‘Hume’s Naturalistic Theory of Representation’. Synthese 152(3), pp. 301–19. Hutcheson, Francis (1728, 1742). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense. London. Quotations and page references from Garrett, Aaron, ed. (2002). An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense, by Francis Hutcheson. Indianapolis: Liberty Fund. Locke, John (1690, 1694). An Essay concerning Humane Understanding. London. Quotations and page references from Nidditch, Peter H., ed. (1975). An Essay concerning Human Understanding, by John Locke. Oxford: Clarendon Press. Malebranche, Nicolas (1675). De la Recherche de la Vérité. Paris. Quotations and page references from Thomas M. Lennon and Paul J. Olscamp, trans. (1997). The Search After Truth, by Nicolas Malebranche. Cambridge: Cambridge University Press. Millican, Peter (2017). ‘Hume’s Fork and His Theory of Relations’. Philosophy and Phenomenological Research 95, pp. 3–65. Nadler, Steven, ed. (2006). The Cambridge Companion to Malebranche. Cambridge: Cambridge University Press. Parfit, Derek (1997). ‘Reasons and Motivation’. Proceedings of the Aristotelian Society, Supplementary Volumes 71, pp. 99–130. Pigden, Charles R., ed. (2009). Hume on Motivation and Virtue. Basingstoke: Palgrave Macmillan. Smith, Michael (1994). The Moral Problem. Oxford: Blackwell. Williams, Bernard (1981). ‘Internal and External Reasons’. In Moral Luck: Philosophical Papers 1973–1980. Cambridge: Cambridge University Press, pp. 101–13.

9 The Causes of the Violent Passions

In this chapter and the next, I look at section 6 of the Dissertation on the Passions together with the third of the four dissertations, Of Tragedy. As we will see, these two parts of the Four Dissertations very clearly belong together: in the former Hume introduces a general principle in his account of the causes of the violent passions, which in the latter he then applies to a particularly intriguing case, namely the pleasure that we take in tragic art. In §9.1, I  place Hume’s discussion in its theoretical context. In the next two sections, I add some historical context by examining the previous solutions of Du Bos (§9.2) and Fontenelle (§9.3). In §9.4, I turn to Hume’s objections to these accounts, arguing that in the case of Du Bos they are disappointingly weak, and suggesting there may have been another reason why Hume rejected the Frenchman’s account. Finally, in §9.5 I will highlight what is most distinctive about Hume’s new solution, while offering a conjecture as to what this other reason might have been. The precise details of Hume’s solution will then be the topic of the next chapter.

9.1.  A Mixture of Passions Hume’s argument concerning the combat of passion and reason, discussed in the previous chapter, is very well known. Having argued that there can be no such combat, strictly speaking, he goes on to claim that the real combat people have been thinking of in this connection is between calm and violent desires. Less well known is Hume’s ensuing discussion, in both the Treatise and the Dissertation on the Passions, of the causes of the violent emotions. In Book 2 of the Treatise, this discussion constitutes the core of part 3, taking up five of this final part’s ten sections (T 2.3.4–8, pp. 418–38). Section 5 discusses the effects of custom, section 6 looks at the influence of the imagination, while sections 7 and 8 examine the effects of distance in time and space. As I  explained in §1.3, very little from these sections remains in the Dissertation on the Passions, but this later work does reproduce section 4 more or less in its entirety. In this section (which becomes section  6 of the Dissertation), Hume introduces and defends his conversion principle: ‘It is a property

The Causes of the Violent Passions  163 in human nature, that any emotion, which attends a passion, is easily converted into it. . . . The predominant passion swallows up the inferior, and converts it into itself. The spirits, when once excited, easily receive a change in their direction’ (P 6.2, p. 26; see also T 2.3.4.2, pp. 419–20).1 Section  1 of the Dissertation on the Passions, meanwhile, contains Hume’s account (examined in §6.2) of the generation of hope and fear from the mixture, in different proportions, of joy and sorrow. Having given this account, Hume then briefly expands it into a general theory concerning the mixing of ‘contrary’ passions (i.e. one positive and one negative): In contrary passions, if the objects be totally different, the passions are like two opposite liquors in different bottles, which have no influence on each other. If the objects be intimately connected, the passions are like an alcali and an acid, which, being mingled, destroy each other. If the relation be more imperfect, and consist in the contradictory views of the same object, the passions are like oil and vinegar, which, however mingled, never perfectly unite and incorporate. (P 1.24, p. 6; see also T 2.3.9.17, p. 443) The third of these principles is the one responsible for the production of hope and fear. Hope, on this chemical analogy, is like oil with a dash of vinegar, while fear is like vinegar with a dash of oil. Hume does not say very much about the first two principles, in either the Treatise or the Dissertation. In the case of contrary passions concerning the same object cancelling each other out, he offers nothing at all beyond a bare statement of the principle (P  1.22, p.  6; T  2.3.9.15, p. 442). In the case of contrary passions concerning different objects having no effect on each other, he merely offers one illustration: ‘when a man is afflicted for the loss of a law-suit, and joyful for the birth of a son, the mind, running from the agreeable to the calamitous object; with whatever celerity it may perform this motion, can scarcely temper the one affection with the other, and remain between them in a state of indifference’ (P 1.21, p. 6; see also T 2.3.9.14, pp. 441–2). Section 1 then ends with a hint of a fourth principle, concerning the mixture of any two passions, not just a positive and a negative: ‘The effect of a mixture of passions, when one of them is predominant, and swallows up the other, shall be explained afterwards’ (P 1.25, p. 6). This is of course a reference ahead to section 6, and to the conversion principle that I have already introduced. Thus this principle belongs in the overlap between two of Hume’s general theories: concerning the mixture of passions, and concerning the causes of the violent passions. The two topics overlap, straightforwardly enough, because the effect of mixing two passions sometimes is precisely an increase in the force or violence of one of them.

164  Part II I have already looked at Hume’s account of hope and fear, and, as I  said previously, Hume says very little about his other two principles concerning the mixing of contrary emotions. That leaves the conversion principle. As for the causes of the violent passions, I will hereafter restrict my attention to this same principle. It will become clear presently (if it isn’t already) why this principle is of particular importance to me here. Furthermore, it is the only aspect of the general theory that Hume continued to discuss in detail after the Treatise, and so it makes sense that a study of Hume’s later philosophy of emotion should exhibit the same bias. The conversion principle is a considerable piece of evidence in support of my viewing the Four Dissertations as Hume’s Enquiry concerning the Passions. The principle itself was already present in Book 2 of the Treatise (T 2.3.4.2, pp. 419–20), and Hume’s discussion of it in the Dissertation differs only in superficial points of style. It was only after writing the Treatise, however, that Hume came to realize that his principle might be applied to the curious phenomenon of the pleasure that we take in tragic art (see §9.4 and §9.5). He develops this application, furthermore, not in the Dissertation on the Passions, but in the following dissertation Of Tragedy. One could hardly wish for a clearer case of unity among the Four Dissertations, nor of difference between these works and the Treatise.

9.2.  The Languid, Listless State of Indolence Hume’s dissertation Of Tragedy is very familiar to philosophers of art for its presentation of one of the canonical solutions to the paradox of art and negative emotions. Put simply, the puzzle is this: why do we so much enjoy and appreciate things in art (tragedy, horror, blood and gore) that we find so unpleasant in real life? Hume motivates his own famous answer to this question with a brief look at the earlier solutions offered by Jean-Baptiste Du Bos and Bernard Le Bovier de Fontenelle. Probably because of Hume’s criticisms, these solutions have since disappeared from the debate (though Hume’s account continues to receive much attention). In the case of Fontenelle, I am not persuaded that this is any great loss. The case of Du Bos, however, is different. Du Bos’s principal work for our purposes (he also wrote some historical books) is the two-volume Réflexions Critiques sur la Poésie et sur la Peinture (1719). This work contains a wealth of interesting psychological and art-historical observations, together with intriguing suggestions in the philosophy of art and art criticism. Here, I will restrict my attention to just the introduction and the first three sections of volume 1. The initial aim of this first volume is ‘to explain what the beauty of a picture or poem chiefly consists in’ (1719, vol. 1, p. v). The nature of the pleasure that we receive from poetry and painting, Du Bos thinks, is on the face

The Causes of the Violent Passions  165 of it rather puzzling, because a great deal of the time it bears a striking resemblance to sorrow and other unhappy feelings: ‘The arts of poetry and painting are never more applauded, than when they are most successful in moving us to pity’ (1719, vol. 1, p.  1). This is of course the familiar paradox of art and negative emotions. Du Bos might even have been the first person to formulate this paradox in modern terms (see Livingston 2013, pp. 401–4). Du Bos’s own solution, put forward in section 3, depends on an interesting view of human psychology developed in sections  1 and 2. The core of this view is that many of the emotions that we typically view as negative or painful in fact have a good deal to recommend them. The first point made in support of this is the idea that ‘one of the greatest wants of man is to have his mind incessantly occupied’ (1719, vol. 1, p. 5). So strong is this desire, according to Du Bos, that we will often seek out great pain in order to satisfy it: ‘The heaviness which quickly attends the inactivity of the mind, is a situation so very disagreeable to man, that he frequently chuses to expose himself to the most painful exercises, rather than be troubled with it’ (1719, vol. 1, p. 5). In section 2, Du Bos then turns more particularly to the emotions that we experience ‘upon seeing our fellow creatures in any great misfortune or danger’ (1719, vol. 1, p. 10). Although he begins by saying that these emotions have nothing to recommend them beyond the general relief from ennui introduced in section  1, his subsequent elucidation of the sorts of relief that they provide, and of what exactly it is that we enjoy about this emotional stimulation, could (to my mind) just as well be viewed as additions to the basic general point. I shall not dwell here on the details of Du Bos’s examples. Some depend on more morally dubious sides of human nature, such as the pleasure that we take in public executions (1719, vol. p. 10), in gladiatorial battles (1719, vol. p. 12), or in bull-fights (1719, vol. p. 18). Others are more innocent, however, at least by our standards, such as the thrill of tight-rope walking (1719, vol. 1, pp. 11–2) or gambling (1719, vol. 1, pp. 19–20). Having argued that many negative experiences, for all their undoubted negativity, also have much to recommend them—emotional stimulation, the thrill of danger and high stakes, satisfaction of our bloodlust (whether we approve of this or not)—Du Bos then introduces his account of the pleasure of poetry and painting. What these art forms offer us, he says, are artificial experiences with all these pleasant effects, but none of the usual drawbacks: Since the most pleasing sensations that our real passions can afford us, are balanced by so many unhappy hours that succeed our enjoyments, would it not be a noble attempt of art to endeavour to separate the dismal consequences of our passions from the bewitching pleasure we receive in indulging them? Is it not in the power of art

166  Part II to create, as it were, beings of a new nature? Might not art contrive to produce objects that would excite artificial passions, sufficient to occupy us while we are actually affected by them, and incapable of giving us afterwards any real pain or affliction? (1719, vol. 1, p. 21) The crucial idea here, which was hinted at in the previous sections, but now appears more clearly, is that in real life our emotional reactions to tragic, dangerous, and horrific events are mixed, containing both positive and negative aspects; but the negative aspects are so strong that they typically drown out the positive. What poetry and painting do, thinks Du Bos, is present us with ‘imitations’ of these events. And an imitation of this sort, because it is only fictional, ‘does not affect our reason, which is superior to the illusory attack of those sensations, . . . [but] affects only the sensitive soul’ (1719, vol. 1, p. 23). As such, it is able to induce the pleasurable feelings in isolation, untainted by the pain that attends them in real life: The pleasure we feel in contemplating the imitations made by painters and poets, of objects which would have raised in us passions attended with real pain, is a pleasure free from all impurity of mixture. It is never attended with those disagreeable consequences, which arise from the serious emotions caused by the object itself. (1719, vol. 1, p. 24) Du Bos is not entirely explicit in listing these ‘disagreeable consequences’, and a more careful and thorough interpretation would disentangle what he takes to be the positive and the negative aspects of our experience. But that is more than I have the space (or need) for here, and I trust that the above has given a sufficient sense of the overall structure, at least, of Du Bos’s solution.

9.3.  A Pleasure From the Bosom of Uneasiness Du Bos’s Réflexions Critiques sur la Poésie et sur la Peinture is a long book, containing several astute psychological observations, and an intriguing and well thought-out solution to the paradox of art and negative emotions. By contrast, Fontenelle’s similarly titled Réflexions sur la Poétique (1742) is a medium-sized essay, containing just a single paragraph discussing this particular puzzle. This is the paragraph that Hume quotes, in his own (perfectly good) translation, in Of Tragedy: ‘Pleasure and pain,’ says he [Fontenelle], ‘which are two sentiments so different in themselves, differ not so much in their cause. From the instance of tickling, it appears, that the movement of pleasure, pushed a little too far, becomes pain; and that the movement of pain,

The Causes of the Violent Passions  167 a little moderated, becomes pleasure. Hence it proceeds, that there is such a thing as a sorrow, soft and agreeable: It is a pain weakened and diminished. The heart likes naturally to be moved and affected. Melancholy objects suit it, and even disastrous and sorrowful, provided they are softened by some circumstance. It is certain, that, on the theatre, the representation has almost the effect of reality; yet it has not altogether that effect. However we may be hurried away by the spectacle; whatever dominion the senses and imagination may usurp over the reason, there still lurks at the bottom a certain idea of falsehood in the whole of what we see. This idea, though weak and disguised, suffices to diminish the pain which we suffer from the misfortunes of those whom we love, and to reduce that affliction to such a pitch as converts it into a pleasure. We weep for the misfortune of a hero, to whom we are attached. In the same instant we comfort ourselves, by reflecting, that it is nothing but a fiction: And it is precisely that mixture of sentiments, which composes an agreeable sorrow, and tears that delight us. But as that affliction, which is caused by exterior and sensible objects, is stronger than the consolation which arises from an internal reflection, they are the effects and symptoms of sorrow, that ought to predominate in the composition.’ (Tr 7, pp. 218–9) In this quite long paragraph, Fontenelle does not seem to me to offer a coherent explanation of the phenomenon, but rather presents a few different ideas, which are not obviously consistent with each other. The first idea is an echo of Du Bos’s claims that we have a strong desire for emotional stimulation of all kinds, and that ostensibly negative emotions may in truth be more desirable than is often thought: ‘The heart likes naturally to be moved and affected. Melancholy objects suit it’. A second idea in this paragraph also has some similarities with Du Bos’s account. This is the idea that the fictional nature of the events depicted weakens or softens the emotions, and that a passion that is typically painful in fact becomes pleasurable when it is thus softened. Du Bos too, as we have seen, thought that the fictional nature of the events neutralized the unpleasant aspects of the experience. For Du Bos, however, our reallife experience combines pleasant with unpleasant emotional responses, and what the fictional case does is simply remove the unpleasant; for Fontenelle, at least according to this second idea, there is no pleasure until the falsehood of what we are seeing weakens the otherwise purely painful response. For Du Bos, the real-life reaction is mixed, and the artificial reaction is pure; for Fontenelle, both reactions are pure, but one is purely painful and the other purely pleasant. There also seems to be yet a third idea towards the end of Fontenelle’s paragraph, where he suggests that our experience at the theatre is not after all purely pleasant, but is rather a mixture of pleasure and pain,

168  Part II an ‘agreeable sorrow’ that combines the sadness with the comforting thought that what we are seeing is not real. I leave to others the task of rendering this idea consistent with the rest.

9.4.  Neither Is the Sorrow Here Softened by Fiction Though he does not name them explicitly there, Hume had apparently read Du Bos and Fontenelle when he wrote the Treatise. For in a couple of passages in Book 1 (the second a later addition from the appendix) he shows an awareness of the paradox of art and negative emotions, and gestures at a (somewhat vague) solution that contains elements of both these earlier accounts: We may add to this a remark; that in matters of religion men take a pleasure in being terrify’d, and that no preachers are so popular, as those who excite the most dismal and gloomy passions. In the common affairs of life, where we feel and are penetrated with the solidity of the subject, nothing can be more disagreeable than fear and terror; and ’tis only in dramatic performances and in religious discourses, that they ever give pleasure. In these latter cases the imagination reposes itself indolently on the idea; and the passion, being soften’d by the want of belief in the subject, has no more than the agreeable effect of enlivening the mind, and fixing the attention. (T 1.3.9.15, p. 115) There is no passion of the human mind but what may arise from poetry; tho’ at the same time the feelings of the passions are very different when excited by poetical fictions, from what they are when they arise from belief and reality. A passion, which is disagreeable in real life, may afford the highest entertainment in a tragedy, or epic poem. In the latter case it lies not with that weight upon us: It feels less firm and solid: And has no other than the agreeable effect of exciting the spirits, and rouzing the attention. (T 1.3.10.10, pp. 630–1; appendix addition) At some point after writing the Treatise, however, Hume evidently gave the matter more thought, and came to be dissatisfied with this account. The result was the third of the four dissertations, Of Tragedy, in which he abandons his earlier tacit allegiance to Du Bos and Fontenelle, and offers a very different account of his own. Hume’s later objection to Du Bos is disappointing. To begin with, his summary of Du Bos’s solution mentions only the point about our strong desire for emotional stimulation of all kinds (Tr 3, p. 217). His endorsement of this point (Tr 4, p. 217) should be taken entirely at face value, for he had

The Causes of the Violent Passions  169 already appealed to it in his account of the love of family relations in Book 2 of the Treatise (T 2.2.4.4, p. 352), which may be further evidence that he had already read Du Bos’s work at that time. But he makes no mention whatsoever of Du Bos’s additional discussions of the inherent pleasures of gruesome spectacles, or of the thrill of risk and danger. More importantly, he completely ignores the core of Du Bos’s actual solution, namely that the artificial passions aroused by art are pleasures purified from their usual negative aspects, by virtue of the fictional nature of the events depicted. The extent of Hume’s objection is simply this: There is, however, a difficulty in applying to the present subject, in its full extent, this solution, however ingenious and satisfactory it may appear. It is certain, that the same object of distress, which pleases in a tragedy, were it really set before us, would give the most unfeigned uneasiness; though it be then the most effectual cure to languor and indolence. (Tr 6, p. 218) As a criticism of Du Bos, this is simply beside the point (see Livingston 2013, p. 405). Du Bos after all explicitly agrees, as, for example, here: The massacre of the innocents must have left most gloomy impressions in the imaginations of those, who were real spectators of the barbarity of the soldiers slaughtering the poor infants in the bosom of their mothers, all imbrued with blood. (1719, vol. 1, p. 24) The point of Hume’s objection seems to be that the desire for emotional stimulation by itself cannot explain the phenomenon; we also need to be told, at the very least, what the effective difference is between art and real life. Why are we so entertained in the former case, when we are so distressed in the latter? Whatever one thinks of the ultimate success of Du Bos’s account, it must be admitted that he does at least have an answer to this question. Artworks are imitations, he maintains, and as such they affect only our sensation and imagination, not our reason; they excite only artificial emotions that have none of the negative consequences that go along with their real-life counterparts. When Hume then introduces Fontenelle’s account as offering an ‘addition’ to Du Bos’s that speaks to this objection (Tr 6, p. 218), by locating the key difference between art and real life in the fictional nature of the former, his unfairness to Du Bos is even more striking. For Du Bos, as we have seen, the key difference between imitations and real life is precisely that the former are fictional. Fontenelle’s supposed ‘addition’ was already in Du Bos’s account to begin with.

170  Part II Be that as it may, Hume goes on to make short work of Fontenelle’s novel idea that a pain weakened becomes pleasant: ‘You may by degrees weaken a real sorrow, till it totally disappears; yet in none of its gradations will it ever give pleasure’ (Tr 10, p. 221). He is similarly dismissive of the suggestion that the fictional nature of artworks is the key feature that distinguishes them from reality: The pathetic description of the butchery, made by Verres of the Sicilian captains, is a masterpiece of this kind: But I believe none will affirm, that the being present at a melancholy scene of that nature would afford any entertainment. Neither is the sorrow here softened by fiction: For the audience were convinced of the reality of every circumstance. (Tr 8, p. 219) With this remark Hume does after all give us (though apparently unintentionally) a genuine criticism of Du Bos, as well as of Fontenelle, since the two agreed on this point. The comment, something of an aside in Hume’s dissertation, raises to my mind a substantial issue, which I will return to in the next section. Given the general ineffectiveness of Hume’s criticisms—this one point aside—it seems to me unlikely that he changed his mind on this topic because of these worries. They are not, on inspection, very compelling reasons for abandoning the ideas hinted at by Fontenelle, or the theory carefully developed by Du Bos. It is possible, therefore, that Hume simply hit upon his own new solution, and then hunted around for objections to these earlier writers so as to motivate it. But there is also another and potentially more interesting possibility, namely that Hume had independent reasons for wanting a different explanation, reasons that are not brought out in his explicit and rather weak objections. I will suggest such a reason in the next section.

9.5.  A New Direction From the Sentiments of Beauty What is most distinctive about Hume’s later explanation of the pleasure that we take in tragic art, when compared to Du Bos’s or Fontenelle’s, is the introduction of a completely new set of emotional responses: responses to the artwork itself, rather than to the events that it depicts, such as our delight in its beauty and in the skill of the artist. With these (positive) emotions now in the picture, Hume is able to apply his conversion principle. The negative passions arising from the events themselves, Hume says, are converted into or swallowed up by the predominant feelings of delight evoked by the artistry: The genius required to paint objects in a lively manner, the art employed in collecting all the pathetic circumstances, the judgment

The Causes of the Violent Passions  171 displayed in disposing them: the exercise, I say, of these noble talents, together with the force of expression, and beauty of oratorial numbers, diffuse the highest satisfaction on the audience, and excite the most delightful movements. By this means, the uneasiness of the melancholy passions is not only overpowered and effaced by something stronger of an opposite kind; but the whole impulse of those passions is converted into pleasure, and swells the delight which the eloquence raises in us. . . . The impulse or vehemence, arising from sorrow, compassion, indignation, receives a new direction from the sentiments of beauty. The latter, being the predominant emotion, seize the whole mind, and convert the former into themselves, at least tincture them so strongly as totally to alter their nature. (Tr 9, pp. 219–20) If Hume had only engaged properly with Du Bos here, there would have been a nice contrast to be drawn between their solutions. As we have seen, Du Bos’s position is that in the real-life case there is a mixture of pleasure and pain, in which the pain predominates; while in the artistic case, the painful aspects are removed. Hume however holds that the reallife case is purely painful, while in the artistic case there is a mixture of pleasure and pain in which the pleasure predominates. In the Treatise, when Hume introduces Du Bos’s idea of the strong desire for emotional stimulation of all kinds, he attributes it (without mentioning Du Bos by name) to those ‘who take a pleasure in declaiming against human nature’ (T 2.2.4.4, p. 352). He is in this way implicitly placing Du Bos in the wider group (which includes the egoists) of those who ‘give us a mean opinion of our nature’ (DM 2, p. 81; recall §4.2). This points to the important Epicurean context of Du Bos’s solution. As Hume was probably aware, Hobbes had earlier expressed some closely related ideas: [F]rom what Passion proceedeth it, that men take pleasure to behold from the shore the danger of them that are at Sea in a tempest; or in Fight, from a safe Castle to behold two armies charge one to another in the field? It is certainly, in the whole sum, Joy; else men would never flock to such a spectacle. Nevertheless there is in it both Joy and Grief: for as there is novelty and remembrance of our own Security present, which is delight: so there is also Pity, which is Grief: But the Delight is so far predominant, that men usually are content in such a case to be spectators of the misery of their friends. (Hobbes 1651, p. 114) Whether Du Bos was directly inspired by Hobbes, I do not know, but there is a similarity in that both claim our experience in such cases is a mixture of pleasure and pain. Du Bos also discusses the same two examples

172  Part II of watching a shipwreck from the shore, and a battle from a castle (Du Bos 1719, p. 11). There is nothing in that to tie Hobbes directly to Du Bos—the examples are from Lucretius’s poem De Rerum Natura (recall §3.4)—but the common source is significant in itself. Lucretius’s poem is an explication of Epicurean philosophy (indeed, it is one of our major sources of the ideas of this school), and both Hobbes and Du Bos are recognizably influenced by the Epicurean tradition in their psychological theories. This context makes Hume’s rejection of Du Bos’s solution—indeed, his complete failure to engage with it properly—potentially more intelligible. As I argued extensively in Part I of this book, Hume’s philosophy of motivation underwent a major change after the Treatise away from egoism and the Epicurean tradition, and towards sentimentalism and the rival Stoical tradition. I suggested at the end of the previous section that Hume may have abandoned his earlier solution to the paradox of art and negative emotions simply because he came up with a novel one that seemed better. But I  also suggested a more interesting possibility: that he later had other reasons for disliking Du Bos’s account that he did not make explicit. One such potential reason is his general attempt to disassociate himself from Hobbes and Epicurus, and to offer more morally palatable psychological explanations of the phenomena that he was investigating. Kames’s essay on the paradox of tragedy, in his Essays on the Principles of Morality and Natural Religion (1751), also took as its starting point a disagreement with Du Bos. Unlike Hume, Kames was explicit in placing Du Bos in the Epicurean tradition, and his own rejection of Du Bos’s account depended crucially on an appeal to Butler’s motivational pluralism: Self-love is a strong motive to search about for every thing that may contribute to happiness. Self-love operates by means of reflection and experience; and every object, as soon as discovered to contribute to our happiness, raises in us of course a desire of possessing. Hence it is, that pleasure and pain are the only motives to action, as far as selflove is concerned. But our appetites and passions are not all of them of this kind. They frequently operate by direct impulse, without the intervention of reason, in the same manner as instinct does in brute creatures. As they are not influenced by any sort of reasoning, the view of shunning misery or acquiring happiness, makes no part of the impulsive motive. It is true, that the gratification of our passions and appetites, is agreeable; and it is also true, that, in giving way to a particular appetite, the view of pleasure may, by a reflex act, become an additional motive to the action. But these things must not be confounded with the direct impulse arising from the appetite or passion; which, as I have said, operates blindly, and in the way of instinct, without any view to consequences. (Kames 1751, p. 15)

The Causes of the Violent Passions  173 Though Butler is not mentioned explicitly, I trust it is clear that it was his view Kames was appropriating here; even the terminology of ‘particular’ appetites and passions is Butler’s. And Butler is approvingly cited and appealed to in the following essay on the principles of morality (1751, p. 33). Given the close relationship between Hume and Kames, this adds weight to the suggestion of the previous paragraph. For Hume was certainly aware of Kames’s views.2 Returning to the other unanswered question from the previous ­section— of what the crucial distinguishing feature of artworks is that makes our response different from what it would be in real life—the obvious first thought, given what we have now seen of his solution, is that Hume’s answer ought to be simply that it is an artwork. The difference in reactions, thinks Hume, is down to the addition of some positive emotions, but these are not in any way a response to the fictional nature of the events depicted (the events need not even be fictional). Rather, they arise because the events are depicted so beautifully, with such great skill and artistry. This is certainly right as far as it goes. On closer inspection, however, it cannot be the whole of Hume’s answer, for the following reason. Supposing that some work of art evokes pleasant emotions because of its beauty, and negative emotions because of its content, it is still an open question which of these two sets of emotions will be predominant in the mixture. In some cases, Hume says, the negative feelings dominate, and all the eloquence in the world will only serve to make us feel worse: Who could ever think of it as a good expedient for comforting an afflicted parent, to exaggerate, with all the force of elocution, the irreparable loss, which he has met with by the death of a favourite child? The more power of imagination and expression you here employ, the more you encrease his despair and affliction. (Tr 21, p. 223) More examples are given in the following paragraphs: Cicero’s rhetorical skill in describing Verres’s massacre delights the audience, but only distresses Verres himself all the more (Tr 22, p. 223); Clarendon overlooks the death of King Charles I in his history of the civil war, because ‘[h]e himself, as well as the readers of that age, were too deeply concerned in the events’ (Tr 23, p. 223); the English theatre depicts too much blood and gore, creating a sense of horror that no amount of artistry can overcome (Tr 24, p. 224); even our sorrow at seeing virtuous characters suffer at the hands of the vicious predominates over any feelings of beauty, unless softened by having the virtue turn into a courageous despair, or the villains receive their proper punishment (Tr 25, p. 224). Though this shows that a positive reaction to the beauty and brilliance of the artwork cannot be the sole difference between art and reality, however, it does not show us what the crucial missing factor is. What

174  Part II is different about the cases in which this positive reaction predominates, as opposed to those in which it only serves to increase our pain? Having thus excited the reader’s curiosity, I  will follow Hume’s advice (Tr  13, p. 221) and refrain from revealing the answer to this question until §10.4.

Notes 1 This name for the principle seems to have been coined by Margaret Paton (1973, p.  121). It or something like it has since been used by most people writing on the topic: Hill (1982), Feagin (1983), Packer (1989), Schier (1989), Budd (1991), Yanal (1991, 1992), Neill (1992, 1998), Leibowitz (1993), and Galgut (2001). I have previously referred to it as the ‘principle of the predominant passion’ (Merivale 2011), taking my cue from Immerwahr (1994). I had a reason for this idiosyncrasy: I felt that the emphasis on conversion, at the expense of Hume’s other ways of describing the process (one passion swallowing up another, the spirits changing direction), was unfortunate. But I now think that the conversion idea probably is the most important after all, and in any case I regret the cumbersome nature of the longer name. 2 Kames’s own solution to the paradox is a precursor to so-called compensation theories (e.g. Feagin 1983). Kames insists that, while the objects of grief excite aversion in us, grief itself does not. Reflecting on this feeling, especially in the case of compassion or sympathy with others, is on the contrary attended with desire: ‘We submit willingly to such painful passions, and reckon it no hardship to suffer under them. In being thus constituted, we have the consciousness of regularity and order, and that it is right and meet we should suffer after this manner’ (1751, pp. 19–20). Hume never engages with this proposal, but I expect it would have seemed too Stoical, even after the Treatise, to his still somewhat Epicurean tastes.

Bibliography Budd, Malcolm (1991). ‘Hume’s Tragic Emotions’. Hume Studies 17(2), pp. 93–106. Du Bos, Jean-Baptiste (1719). Réflexions Critiques sur la Poesie et sur la Peinture, in 2 volumes. Paris. Quotations and page references from volume 1 of Thomas Nugent, trans. (1748). Critical Reflections on Poetry, Painting, and Music, by Jean-Baptiste Du Bos, in 3 volumes. London. Feagin, Susan L. (1983). ‘The Pleasures of Tragedy’. American Philosophical Quarterly 20(1), pp. 95–104. Fontenelle, Bernard (1742). Oeuvres de Fontenelle. Paris. Galgut, Eliza (2001). ‘The Poetry and the Pity: Hume’s Account of Tragic Pleasure’. The British Journal of Aesthetics 41(4), pp. 411–24. Hill, Eric (1982). ‘Hume and the Delightful Tragedy Problem’. Philosophy 57(221), pp. 319–26. Hobbes, Thomas (1651). Leviathan. London: Andrew Crooke. Quotations and page references from Richard Tuck, ed. (1996). Leviathan, by Thomas Hobbes. Cambridge: Cambridge University Press. Immerwahr, John (1994). ‘Hume’s Dissertation on the Passions’. The Journal of the History of Philosophy 32(2), pp. 225–40. Kames, Lord [Henry Home] (1751). Essays on the Principles of Morality and Natural Religion. Edinburgh: R. Fleming. Quotations and page references from

The Causes of the Violent Passions  175 Mary Catherine Moran, ed. (2005). Essays on the Principles of Morality and Natural Religion, by Henry Home, Lord Kames. Indianapolis: Liberty Fund. Leibowitz, Flo (1993). ‘The Logic of Hume’s Conversion Theory’. The Journal of Aesthetics and Art Criticism 51(4), pp. 625–6. Livingston, Paisley (2013). ‘Du Bos’ Paradox’. The British Journal of Aesthetics 53(4), pp. 393–406. Merivale, Amyas (2011). ‘Mixed Feelings, Mixed Metaphors: Hume on Tragic Pleasure’. The British Journal of Aesthetics 51(3), pp. 259–69. Neill, Alex (1992). ‘Yanal and Others on Hume on Tragedy’. The Journal of Aesthetics and Art Criticism 50(2), pp. 151–4. ——— (1998). ‘ “An Unaccountable Pleasure”: Hume on Tragedy and the Passions’. Hume Studies 24(2), pp. 335–54. Packer, Mark (1989). ‘Dissolving the Paradox of Tragedy’. The Journal of Aesthetics and Art Criticism 47(3), pp. 211–19. Paton, Margaret (1973). ‘Hume on Tragedy’. The British Journal of Aesthetics 13(2), pp. 121–32. Schier, Flint (1989). ‘The Claims of Tragedy: An Essay in Moral Psychology and Aesthetic Theory’. Philosophical Papers 18(1), pp. 7–26. Yanal, Robert J. (1991). ‘Hume and Others on the Paradox of Tragedy’. The Journal of Aesthetics and Art Criticism 49(1), pp. 75–6. ——— (1992). ‘Still Unconverted: A Reply to Neill’. The Journal of Aesthetics and Art Criticism 50(4), pp. 324–6.

10 The Predominant Passion

In the previous chapter, I placed Hume’s conversion principle in its context, and looked at how Hume came to apply it to the paradox of art and negative emotions. I ended with an unanswered question: what is it that makes one passion predominant over another? Why is it, for example, that for most of us the pleasure of Cicero’s rhetoric is heightened by the horrific subject matter of Verres’s massacre, while for Verres himself the shame and terror are heightened by the rhetoric? This is one of three main interpretative questions regarding Hume’s principle. The other two concern the nature of the process itself (how the predominant and subordinate passions interact), and its end result (what becomes of these two passions after their interaction). In this chapter, I offer answers to each of these questions. It will be convenient to tackle the middle question first, concerning the nature of the process itself, since this question is easier to answer, and has an impact on the answers to the other two questions (§10.1, §10.2, and §10.3). From the outset I should say that we cannot assume, simply because these other two questions naturally arise, that Hume himself had precise answers in his own mind. I will initially proceed on the basis of this assumption, and see which answers are best supported by the text (§10.4). In §10.5, however, I will argue that Hume was not in fact quite so clear on these matters as we might have wished.1

10.1.  Any Emotion, Which Attends a Passion Hume’s statement of his conversion principle includes three images or metaphors: It is a property in human nature, that any emotion, which attends a passion, is easily converted into it; though in their natures they be originally different from, and even contrary to each other. . . . The predominant passion swallows up the inferior, and converts it into itself. The spirits, when once excited, easily receive a change in their direction; and it is natural to imagine, that this change will come from the prevailing affection. (P 6.2, p. 26; see also T 2.3.4.2, pp. 419–20)

The Predominant Passion  177 The first image invoked is that of the conversion of one passion into another. This is perhaps the most important idea; certainly it is the one that has dominated discussions of Hume’s view, and indeed given the principle its familiar name. I will say more about this idea in §12.3. The second image is a meal metaphor: Hume talks of the predominant passion ‘swallowing up’ the subordinate. The third is a gesture at the underlying physiology: the spirits are said to receive a change in their direction. I will say more about this last in §12.2. The meal metaphor does not seem especially puzzling or problematic, and I have nothing to say about it directly. It is well to be aware of it, at least, for it will prove helpful when turning to consider what happens to the passions following their interaction (see §12.4). But in itself it does not seem to raise any interpretative questions. This metaphor does however provide the occasion for settling an important preliminary issue. The issue in question arises from an influential suggestion by Alex Neill (1998). Neill’s idea is that it is not the subordinate passion itself that is swallowed up by the predominant, in Hume’s process, but rather a distinct item called an ‘emotion’ that ‘attends’ the subordinate passion. The textual springboard for this idea is Hume’s very first statement of the principle: ‘It is a property in human nature, that any emotion, which attends a passion, is easily converted into it’ (P 6.2, p. 26). Interesting though this suggestion is, it does not seem to be supported by the text. As far as I can see, the only difference discernible in Hume’s use of the terms ‘passions’ and ‘emotions’ is the connotation—entirely standard at the time—that the former are more violent (and perhaps less ‘rational’) than the latter. Thus at the start of Book 2, in distinguishing between the calm and the violent secondary impressions, Hume writes: The raptures of poetry and music frequently rise to the greatest height; while those other impressions, properly call’d passions, may decay into so soft an emotion, as to become, in a manner, imperceptible. But as in general the passions are more violent than the emotions arising from beauty and deformity, these impressions have been commonly distinguish’d from each other. (T 2.1.1.3, p. 276) Strictly speaking, then, the ‘passions’ are a subset of the ‘emotions’. This itself undermines Neill’s suggestion. But as this distinction between the emotions in general and the passions in particular is seldom Hume’s focus, for the most part the two words are effectively just stylistic variants. Many passages are suggestive of this straightforward interpretation, but perhaps the clearest piece of evidence comes from a revision that Hume made to a paragraph from Treatise Book 2: Thus a man, who, by any injury from another, is very much discompos’d and ruffled in his temper, is apt to find a hundred subjects of discontent,

178  Part II impatience, fear, and other uneasy passions; especially if he can discover these subjects in or near the person, who was the cause of his first passion. (T 2.1.4.4, p. 284) The repetition of ‘passion’ at the end here is somewhat inelegant, as Hume himself no doubt realized later. For in the corresponding passage in the Dissertation on the Passions it is avoided by using ‘emotion’ instead: Thus, a man, who, by an injury received from another, is very much discomposed and ruffled in his temper, is apt to find a hundred subjects of hatred, discontent, impatience, fear, and other uneasy passions; especially, if he can discover these subjects in or near the person, who was the object of his first emotion. (P 2.8, p. 8) This provides clear evidence that these terms differ in style more than substance.2 Since Hume never explicitly draws a distinction between ‘passions’ and their attendant ‘emotions’, and since there was no such standard distinction that would have been familiar to Hume’s readers, particularly strong evidence is required to suggest that Hume really was operating with some such contrast in mind. Neill’s evidence, however, is all perfectly consistent with the more straightforward synonymy interpretation. For example, it is unsurprising that Hume should say ‘any emotion, which attends a passion, is easily converted into it’ (P 6.2, p. 26), rather than use the more clumsy expression ‘any passion, which attends a passion’. Moreover, Neill seems in any case to have misread this passage: the emotion in question is said to attend the predominant passion, but Neill’s interpretation requires that it attend the subordinate one. More directly, the effective synonymy of these two words in the present context is established by the fact that Hume refers to the subordinate item, the thing that is converted or swallowed up, as both a ‘passion’ and an ‘emotion’. Neill acknowledges one place in which Hume refers to the subordinate item as a passion (T 2.3.4.2, p. 420; see also P 6.2, p. 26), but goes on to say that this ‘is in fact very much an exception to the rest of his talk about affective conversion’, in which ‘emotion’ is the more common term (Neill 1998, p. 345). But Neill’s selective quotations in support of this claim are not representative. In the Dissertation on the Passions and Of Tragedy combined, the tally is in fact exactly equal, with the subordinate item being referred to as a ‘passion’ five times and an ‘emotion’ five times.3 The relevant section of the Treatise (T 2.3.4, pp. 418–22), which includes most of the Dissertation’s references, does not alter this picture.4 Thus Neill’s proposal is not borne out by the text.

The Predominant Passion  179

10.2.  The Spirits Receive a Change in Their Direction Hume’s reference to ‘spirits’ in connection with his conversion principle is to be understood in the context of the physiological theory of the time. In the second century, Galen had hypothesized that food was converted into natural spirit by digestion, which latter was then carried (by the veins) to nourish the body’s tissues. It was also sent to the heart, where— mixed with air breathed into the lungs—it was converted into vital spirit, which was transported throughout the body by the arteries. Some of this vital spirit, when it reached the brain, was transformed into animal spirits, which were sent through the nervous system. These animal spirits served a two-way communicative purpose: to send information from the sense organs to the brain, and to send instructions from the brain to the muscles.5 Aside from differences at the level of detail that needn’t concern us here (chiefly concerning how the animal spirits effected these two communicative tasks), Galen’s view of the nervous system was the standard one until Luigi Galvani produced compelling evidence (in 1780, shortly after Hume’s death) that electricity rather than the hypothesized animal spirits performed the communicative function.6 It is thus no surprise to find Descartes and Malebranche positing movements in the animal spirits as the immediate causes of the passions (or the occasions for God to bring about the passions, for Malebranche).7 The two Frenchmen had their own particular take on how the spirits managed to do this, but otherwise this claim was simply received wisdom: all the perceptions of the mind were presumed to be brought about in this way. Hume, like any learned man of his age, was familiar with the gist of the physiological theory of the day, as we see particularly clearly in his essay The Epicurean: The stomach digests the aliments: The heart circulates the blood: The brain separates and refines the spirits. (Ep 5, p. 140) It should go without saying, however, that Hume was no physiologist. Though his work on the passions is peppered—very lightly—with mentions of the spirits and of movements in the spirits, in every case he is simply taking the received wisdom for granted. With just one exception, these claims never play any indispensable role in his psychological theories, and in the exceptional case Hume explicitly apologizes for the anomaly: ’Twou’d have been easy to have made an imaginary dissection of the brain, and have shewn, why upon our conception of any idea, the animal spirits run into all the contiguous traces, and rouze up the

180  Part II other ideas, that are related to it. But tho’ I have neglected any advantage, which I might have drawn from this topic in explaining the relations of ideas, I am afraid I must here have recourse to it, in order to account for the mistakes that arise from these relations. (T 1.2.5.20, p. 60) It may also be noted that this is in the context of Hume’s discussion of the association of ideas; in the case of the passions, remarks about the spirits add some nice visual imagery, but never serve any significant theoretical purpose. In the context of the conversion principle, Hume’s talk of the spirits receiving a change in their direction is obviously to be understood as a gesture at the underlying physiology. But the principle itself is a psychological one, as far as Hume is concerned, and he supports it exclusively with psychological examples; the physiological descriptions of it do not seem to play any essential role in Hume’s argument, as far as I can see, and might well be deleted at no significant philosophical cost. John Wright, as part of his insistence that Malebranche was an important positive influence on Hume, is keen to emphasize Hume’s occasional references to animal spirits. He claims that Hume ‘seems to have thought that the search for something like the physical causes described in Cartesian psychophysiology was directly relevant to his own activities as a moral philosopher’ (Wright 1983, p.  15). But apart from the isolated and apologetic example regarding the association of ideas that we have already seen, I am aware of no textual evidence to support this surprising claim. With regard to his Malebranchean interpretation of Hume’s associationism, Wright does acknowledge that ‘[t]his part of his account may not be entirely explicit’ (1983, p.  68). And again with regard to Hume’s account of personal identity: ‘there is no explicit reference back to the neurophysiological account when Hume explains how we develop our idea of the continuous identical self’ (1983, p. 73). A less ingenious reader might rather contend that these things are simply non-existent (recall §1.5, note 17). Hume’s allusions to the physiology underlying his conversion principle may provide us with some clues pertaining to the interpretation of that principle (see §10.4). But it would be quite inappropriate to read them as serious attempts to uncover the physical processes of the brain. Hume seems to have been broadly familiar with the theory of the day, and that familiarity shows through in some of his writing, but we should not read anything more into this unsurprising fact.

10.3.  Converted Into the Predominant Passion This brings us to the aspect of Hume’s general description of the conversion principle that has dominated recent commentaries, and indeed given

The Predominant Passion  181 the principle its name: the talk of one passion ‘converting’ another into itself (P 6.2, p. 26; P 6.6, p. 27; Tr 9–12, pp. 219–21; Tr 19–20, pp. 222– 3). There has been much speculation as to what Hume might have meant by this conversion in his discussion of tragedy. The most likely candidate, I  suggest, is not quite the natural reading of the term. It is difficult to make sense of a way in which, for Hume, a particular perception of pride (for example) might have previously been a different passion. It seems that all Hume really can mean by conversion in this context is causation: when one passion is converted into another, this amounts to the former passion causing the latter to appear (and presumably vanishing itself in the process; see §10.4). This suggestion is supported by Hume’s use of the notion in other contexts. In section 2 of the Dissertation on the Passions, when discussing the causal genesis of pride and humility, Hume writes: Beauty of all kinds gives us a peculiar delight and satisfaction; as deformity produces pain, upon whatever subject it may be placed, and whether surveyed in an animate or inanimate object. If the beauty or deformity belong to our own face, shape, or person, this pleasure or uneasiness is converted into pride or humility; as having in this case all the circumstances requisite to produce a perfect transition, according to the present theory. (P 2.17, p. 10; my emphasis; see also T 2.1.8.1, p. 298) The ‘present theory’ here referred to is that of the double relation of sentiments and ideas, according to which the pleasure or uneasiness that Hume says is converted into pride or humility is a key step in the causal chain that produces these passions (recall §7.5). Although this is the only one to remain in the Dissertation, there are more examples of this talk of conversion to be found in Book 2 of the Treatise (e.g. T 2.1.5.5, pp. 286–7; T 2.1.6.2, p. 290; T 2.1.12.8, p. 327). It doesn’t immediately follow, of course, that Hume meant the same thing by this word in his discussion of the conversion principle; but there is a default assumption in favour of this simple hypothesis. Moreover, there is direct evidence of synonymy when this principle is first described, in two sentences that I skipped over when quoting this passage at the start of this chapter: It is a property in human nature, that any emotion, which attends a passion, is easily converted into it; though in their natures they be originally different from, and even contrary to each other. It is true, in order to cause a perfect union amongst passions, and make one produce the other, there is always required a double relation, according to the theory above delivered. But when two passions are already produced by their separate causes, and are both present in the mind, they readily mingle and unite; though they have but one relation, and

182  Part II sometimes without any. The predominant passion swallows up the inferior, and converts it into itself. (P 6.2, p. 26; see also T 2.3.4.2, pp. 419–20) Notice how Hume relates the conversion mechanism here to his theory of the double relation of sentiments and ideas. In order to make one passion ‘produce’ another (cause and effect), a double relation is required between both the sentiment components and the idea components. But when both passions are already present in the mind, one will be converted into the other with just a single relation (meaning a relation of ideas).8 I suggest that this is not just an artful juxtaposition: production and conversion boil down to one and the same thing, namely the causing of one passion by another.9 In the case of the conversion principle, the predominant passion in question is already present in the mind. And what good, one might reasonably ask, is a cause that only ever operates when its effect has already been produced by independent means? To make sense of this, we need to appreciate that Hume has two ways of visualizing the difference between calm and violent passions. Most familiar is his talk of the ‘force’ or ‘violence’ of a passion. But Hume also often speaks in terms of an increase of the passion itself, rather than an increase in its violence, about the ‘heightening’ or ‘raising’ of an emotion, and about variations in ‘quantity’ or ‘degree’.10 While it would be a mistake to suppose that Hume had different properties in mind for these different metaphors,11 it is important to be aware that he conceived of this singular property in (at least) two different ways: as akin to ‘violence’ or ‘turbulence’ on the one hand, and as akin to ‘degree’ or ‘quantity’ on the other. The talk of conversion does not seem to make sense on the model of violence or turbulence, and this is the reason for the present worry. But when we think of the difference between calm and violent passions as a difference in quantity, the idea of conversion as causation makes perfect sense. When two passions are both present to the mind, and one is converted into the other, this amounts to the one causing more of the other to appear.

10.4.  Too Deeply Concerned in the Events Now that we have a sense of how Hume pictured the process itself, we may tackle the other two questions that I posed at the start of this chapter: what determines which of the two passions will predominate, and what happens to the passions at the end of the process? What happens to the predominant passion at the end of the process, at least, is obvious: the main result of the interaction is ‘to give additional force to the prevailing passion’ (P 6.3, p. 26). Hume states this unambiguously, and we could have deduced it easily enough anyway from the context that we saw in the previous chapter. This principle is one of those that Hume offers

The Predominant Passion  183 concerning the causes of the violent emotions, and it wouldn’t belong in this context if its effect was not an increase in force or violence. This is also presumably the point of Hume’s talk of one passion swallowing up another: the predominant passion is strengthened by its metaphorical meal. The question of what happens to the subordinate passion at the end of the process, however, is not so easily settled. Furthermore, this question has often been thought to be particularly important for the evaluation of Hume’s account, since there is some evidence to suggest that he supposes the mechanism utterly annihilates these melancholy passions. We might call this the destruction interpretation of the principle. This destruction interpretation, furthermore, is sometimes thought to entail that our experience of tragic drama is entirely pain-free. This, in turn, is believed to be incorrect as a description of the experience, and the inevitable conclusion is that Hume’s account fails; see Hill (1982, pp. 323–4), Packer (1989, p.  212), and Budd (1991, pp.  103–4). Those who don’t accuse Hume of this error still typically accept that it would be an error, and develop interpretations explicitly designed to overcome it; see Yanal (1991, 1992) and Neill (1992, 1998). Taken at face value, the talk of conversion plainly implies the destruction of the passion that gets converted. This is certainly so if we give the word ‘conversion’ its usual interpretation, and understand Hume as saying that the subordinate passion itself somehow turns into the predominant. But even if we understand conversion as essentially the causing of one passion by another, as I  suggest we should, the natural reading is still that of a causal process involving the destruction of the cause. Nor can we dismiss this as a possibly unintended connotation of the word, for Hume could perfectly well have chosen a causal term with no such connotation (e.g. ‘produces’, ‘creates’). It is chiefly with a view to sidestepping this evidence in favour of the destruction interpretation that Neill introduces his distinction between ‘passions’ and ‘emotions’ (recall §10.1), urging that the emotion is converted rather than the passion, and therefore that the subordinate passion itself does not vanish at the end of the process. We have seen, however, that this distinction has no convincing basis in the text. Hume’s gesture at the underlying physiology points in the same direction. If the stream of the subordinate passion is diverted, the spirits will no longer ‘rummage’ the appropriate cell (T 1.2.5.20, p. 61) and actuate the felt emotion, and so it seems from this picture too that the subordinate passion is fated to disappear. As for the meal metaphor, once again the evidence points in the same direction; for it is difficult to suppose that the subordinate passion might live on in the belly of the predominant, so to speak. Neill himself acknowledges that the meal metaphor implies the destruction of the subordinate passion (1998, p. 344). This evidence is all in principle defeasible. It is open to Neill, for example, to maintain that Hume chose his general descriptions carelessly or

184  Part II badly, and that he never intended these implications. This hypothesis might then be backed up by reference to what Hume says about particular instances of his principle. And indeed it is Neill’s position that the particular instances—with one notable exception—point to the survival of the subordinate emotion (1998, p.  348f). The exception is at Tr  27 (p.  225), where Hume says that ‘too much jealousy extinguishes love’ (my emphasis). Neill acknowledges the exception (1998, p.  350), but thinks that the other examples outweigh the evidence of this one. On inspection, however, his arguments at this stage all rest on the supposed implausibility of applying the destruction interpretation to the particular cases (an example will be given in §10.5), and not on any direct textual evidence. At no point does Hume in fact say or even imply that the subordinate passion does survive the process. It seems to me that many of Neill’s criticisms of the destruction idea are not in fact decisive. I have argued this in detail elsewhere (Merivale 2011, pp. 267–8), and will not repeat the arguments here.12 The present point is simply that they are philosophical objections, not textual. Ultimately, however, I do think that Neill came very close to putting his finger on a difficulty here that is both philosophical and textual, and in the next section I will suggest a subtler refinement of the destruction interpretation. For now, however, I submit that interpretation as provisionally the best of the available options. So much for the fate of the passions after the process. I turn now to the question of what makes one passion predominant over the other in the first place. To the best of my knowledge, only one answer has been given to this question in the secondary literature, namely that the predominant passion is the more forceful of the two.13 I agree that this is the best available answer, but it is worth pointing to the reasons in support of this interpretative claim, since those who have made it before seem to have done so without offering any such support (and the quality of this evidence will also be important in §10.5). The general argument runs as follows: there is only one way in which a passion can vary (other than in its causes or objects), and that is in its force or violence; this is therefore the only salient way in which the predominant passion might differ from the subordinate one; and since it is absurd to suppose that the predominant passion is the less forceful of the two, it must be the more forceful. To this general argument we can add two specific passages from Of Tragedy. Towards the end of this dissertation, Hume offers a handful of examples in support of the ‘inversion’ of the conversion principle (Tr 27, p. 224), which are in fact just examples of the principle itself, but with the predominant passion being a passion that in other (more common) situations is subordinate. The shame of Verres on hearing Cicero’s speech in prosecution of him, we are told, would have been ‘too strong for the pleasure arising from the beauties of elocution’ (Tr 22, p. 223; my emphasis). This is what makes that painful passion predominate

The Predominant Passion  185 in Verres’s mind; and ‘strength’ here is presumably to be equated with ‘force’ and ‘violence’.14 Furthermore, Hume later adds that ‘[t]oo much jealousy extinguishes love’ (Tr 27, p. 225; my emphasis again). Here the metaphor is that of volume rather than violence, but, as we have seen, the underlying property is the same. There is a difficulty with this answer to the question of which passion predominates, however, one that has to some extent been noted already by Malcolm Budd, who criticizes this aspect of Hume’s proposed solution to the problem of tragic pleasure (Budd 1991, p. 96f). His objection, in a nutshell, is that it just isn’t plausible to suppose that the conversion principle, thus understood, applies in the case of tragedy, because one would have to suppose that the pleasurable passions roused by the eloquence with which the melancholy scene is depicted are always more forceful, to begin with, than the painful passions caused by the scene itself. The example of Verres gives us a hint of why Hume might have thought that the pleasurable passions are (in the appropriate cases) always more forceful than the painful. All Cicero’s rhetorical skill supposedly could not induce feelings strong enough to efface Verres’s shame and guilt, which painful passions instead predominated and thereby gained greater force as a result of the orator’s eloquence (Tr 22, p. 223). Why is it, then, that for us Cicero’s rhetoric wins out over the horror? The salient difference for Hume would seem to be that we are not so intimately connected with the tragic facts. Similarly with the case of Lord Clarendon overlooking King Charles I’s death in his history of the civil war: ‘He himself, as well as the readers of that age, were too deeply concerned in the events, and felt a pain from subjects, which an historian and a reader of another age would regard as the most pathetic and most interesting, and, by consequence, the most agreeable’ (Tr 23, pp. 223–4; my emphasis). The difference, for Hume, is in the distance. This at least is how Hume might attempt to meet Budd’s objection, though how successful such an attempt could be I  leave for others to judge. For even if the objection can be met, a broader worry would still remain regarding this aspect of the conversion principle. This broader worry, meanwhile, is closely analogous to a similar worry with the other aspect of this principle discussed here (the fate of the subordinate passion), and I will therefore examine both matters in the next section. For now, suffice it to conclude that the best reconstruction of Hume’s principle, on the basis of the text, is as I have been arguing: the predominant passion (i.e. the antecedently more forceful of the two) destroys the subordinate passion, and acquires yet greater force as a result.

10.5.  Every Thing That Is New, Is Most Affecting The textual evidence that I offered in the previous section is far from decisive. Hume’s metaphors certainly indicate that the fate of the subordinate

186  Part II passion is to disappear, but there is only one comment that backs this up in a particular case. It seems that the only plausible answer to the question of which passion predominates, meanwhile, is that the predominant passion is the more forceful of the two; but again only two particular examples back this up explicitly. I  have argued that these answers are more defensible than any alternative, but the weakness of the evidence in any direction here suggests a subtler interpretation: that although this might be (or should be) Hume’s considered position, he never actually considered either of these questions very closely, so as to frame clear and definite answers in his own mind. I shall now give some reasons for thinking that this subtler interpretation is in fact closer to the mark. The main textual difficulty with the cruder interpretation that I have been developing so far is also, it seems to me, the main philosophical difficulty with Hume’s account. Hume wants his conversion principle to be a general psychological law, explaining a range of distinct phenomena. The particular examples that he appeals to must therefore all have a sufficient amount in common, with these common elements composing the essence of the principle. It is not enough, clearly, that the various cases all involve two passions interacting somehow. Rather, the two passions must be related in the same way every time, and the end results for both must be the same every time. The difficulty here is that Hume himself seems to have been somewhat lax in arguing that these requirements are always met. Nowhere does he take the time to argue, for any of his examples, that the subordinate passion is destroyed at the end of the process. Nor does he ever take the time to argue, for any of his examples, that the passion that predominates is, at the start of the process, the more forceful of the two. In the absence of this detail, Hume’s justification of his principle falls far short. His several examples, we can allow, are all similar at some broad level of description. But are they similar enough? Are they really instances of the very same phenomenon? We have seen hints of this problem before that can now be brought out more clearly. Neill, we may recall, thinks that in several of Hume’s particular examples of the conversion principle it is highly implausible to suppose that the subordinate passion is destroyed at the end of the process. (This is in the context of arguing that Hume didn’t in fact suppose this; if I am right in thinking this argument unsuccessful, the point now becomes an objection to Hume himself.) Take the case of grief at a friend’s death increasing one’s love for him, for example (Tr 16, p. 222). Neill thinks that to suppose one’s love destroys the grief in this case is ‘grotesquely implausible’ (1998, p. 351), and I am inclined to agree. But that Hume (if he was committed to the destruction of the subordinate passion) might be wrong about some of his examples is not so much the problem; more immediately problematic is the fact that he systematically fails to give any argument for this commitment. I have said that the

The Predominant Passion  187 destruction interpretation is the most textually plausible of the available options here. But if we take a step back, ultimately the most plausible interpretation would seem to be that Hume never quite realized the need to give a definite answer to the question of what happens to the subordinate passion at the end of the process. Much the same is true concerning the question of what makes one passion predominant over the other. Budd, we have seen, comes close to appreciating the difficulty here when he challenges the claim that, in the case of tragic pleasure, the joys of the eloquence are always stronger than the pains of the drama. I went some way, in the previous section, towards defending Hume against this charge, but really this is just the tip of the iceberg. For it is not enough for this to be true in the case of tragedy; it must also be true in every case that Hume produces as support for his principle. Hume makes no attempt, however, to argue for this in any of the cases, and it is far from obvious that it is indeed true for them all. Consider, for example, Hume’s claim that ‘when good or evil is placed in such a situation as to cause any particular emotion, besides its direct passion of desire or aversion, this latter passion must acquire new force and violence’ (P 6.5, p. 27; see also T 2.3.4.4, p. 421). For this to be true, it must also be true that no passion is ever more violent than the passions of desire and aversion; for if there were such a passion, it would convert the desire or aversion into itself, rather than the other way around. Or again, consider Hume’s account of why ‘every thing that is new, is most affecting’ (P 6.12, p. 28; see also T 2.3.5.2, p. 423), namely that the passion of surprise or wonder nourishes any emotion attending the novel object. For this to be a genuine instance of the conversion principle, it must be the case that surprise or wonder is the weakest of all the passions; for if something new were the object of a weaker passion, this weaker passion would nourish our surprise, rather than be nourished by it. None of these commitments seems particularly plausible. But again, the immediate problem is not so much that they might be false, but rather that Hume makes no attempt to argue for them. I have suggested that the best answer to the question of what makes one passion predominant over another, for Hume, is that it is the antecedently more forceful of the two. He does more or less say as much in two places in Of Tragedy, and in any case this seems like the only answer Hume has available to him. But if we take a step back, ultimately the most plausible interpretation seems to be that Hume never really thought through this commitment fully. Regarding both of these questions, then, concerning the status of the passions before and after the conversion process, I submit that no definite answer can be given. If we must give answers, then the view best supported by the text seems to be that the predominant passion is the more forceful of the two, and that it destroys the subordinate passion. But it would be overstating the case to say that this is definitely what Hume

188  Part II thought. I suspect that he had no definite thoughts on the matter one way or the other. And after all, this is perhaps only to be expected: Hume was pioneering a new science of experimental psychology. It is unreasonable to suppose that his principles must be as clearly and precisely formulated as we would expect today. This point will re-emerge, in more detail, towards the end of Chapter 12.

Notes   1 I have presented much the same view before in answer to the third question (what becomes of these two passions after their interaction); see Merivale (2011). Here I extend my answer to include the first question as well (what makes one passion predominant over the other).   2 Hume also changes the word ‘cause’ to ‘object’ in this last sentence, something that may have been prompted by more than just stylistic considerations; recall §7.5.   3 The ‘subordinate’ or ‘inferior’ item is termed a ‘passion’ three times (P 6.2, p. 26; Tr 13, 27, pp. 221, 224). The item that stands in the relevant relation to the predominant passion, meanwhile, when it is not explicitly labelled as ‘subordinate’, is termed a ‘passion’ twice (P 1.25, 6.4; 6, 27) and an ‘emotion’ five times (P 6.2, 6.3, 6.6, 6.8, pp. 26, 27; Tr 14, p. 221), bringing the total to five apiece.   4 The item said to be ‘converted’ or ‘transfused’ into the predominant passion is called an ‘emotion’ twice (T 2.3.4.2, p. 419; T 2.3.4.5, p. 421) and a ‘passion’ twice (T 2.3.4.2, p. 420; T 2.3.4.7, p. 421). There is a fifth reference to the ‘inferior emotion’ (T 2.3.4.3, p. 420).  5 Though Galen made some notable progress in our understanding of these matters, not all these ideas were original, and the three sorts of spirits were not first hypothesized by him. For a thorough, book-length treatment of the history of this subject—from a physiologist, who therefore may be presumed to understand it all much better than I do—see Ochs (2004).   6 Other aspects of Galen’s view were most notably challenged by William Harvey in 1628, who argued for the circulation of the blood, with the heart as its pump.   7 Descartes, moreover, was an influential early critic of Harvey, agreeing with the principle of circulation, but rejecting the idea of the heart as a pump. See Gorham (1994).   8 To be precise, Hume says with ‘but one relation, and sometimes without any’ (my emphasis). It is unclear what this further qualification is doing, however. Hume never gives us any examples of one passion being converted into another even when their objects are unrelated. In fact, he told us earlier that ‘[i]n contrary passions, if the objects be totally different, the passions are like two opposite liquors in different bottles, which have no influence on each other’ (P 1.24, p. 6; see also T 2.3.9.17, p. 443). It seems, therefore, that this later qualification was a slip.   9 The notion of conversion is also present in Hume’s account of the sympathetic communication of passions (T 2.1.11.3, p. 317; T 2.1.11.7–8, pp. 318–20). Here, however, it is not one sentiment being converted into another, but the idea of a sentiment being converted into that very sentiment. In this case, it seems possible we have a literal conversion, i.e. one and the same perception changing from an idea into an impression. Since impressions and ideas differ

The Predominant Passion  189 only in their felt intensity, all that is required for this conversion is an injection of force and vivacity. 10 For examples of the first metaphor, see T 2.1.4.4 (p. 284), T 2.2.11.1 (p. 394), T  2.3.4.1 (pp.  418–9), P  5.3 (p.  24), P  6.3 (p.  26), and P  6.6 (p.  27). For examples of the second, see P 1.12 (p. 4), P 2.9 (p. 8), P 2.21 (p. 11), P 2.25 (p. 12), P 2.28 (p. 12), P 2.41 (p. 15), P 2.42 (p. 15), P 2.47 (p. 16), P 6.1 (p. 26), P 6.3 (p. 26), P 6.6 (p. 27), P 6.11 (p. 28), and Tr 27 (pp. 224–5). 11 Aside from the fact that Hume deploys the different metaphors with considerable freedom and variety, with no hint of any underlying pattern, ‘violence’ and ‘degree’ appear to be fairly explicitly equated at T 2.2.2.22 (pp. 343–4), in a passage that does not reappear in the Dissertation on the Passions. 12 Very briefly, Neill’s major worry is that, on the destruction interpretation, Hume’s account has the implausible consequence—already advertised at the start of this section—that our experience of tragedy is entirely pain-free. But this follows only on the very questionable assumptions that the conversion process is very short and that it takes place only once in the course of our engagement with the story. More plausible, one might think, is that several slow conversions could take place. I owe this suggestion to illuminating discussions with Filippo Contessi. 13 This answer is first given clearly and explicitly by Hill (1982, p. 322). Hill attributes it to Paton, who does seem to hint at it (Paton 1973, pp. 127–8), but I do not myself find it obvious that this is what she is saying. The same answer is later given by Budd (1991, p. 95). 14 In one place in the Treatise Hume does distinguish between ‘violent’ and ‘strong’ passions, the latter being those that have a greater influence on the will, regardless of their felt intensity (T  2.3.4.1, pp.  418–9). But elsewhere ‘strength’ is clearly just another term for ‘violence’ (e.g. T 2.2.2.23, p. 344), and the present passage makes much more sense read in this way than in the other. It is not likely that Hume means to say here that Verres’s shame has a greater influence on his actions or volitions than the sentiment of beauty. That may be true, but hardly seems relevant.

Bibliography Budd, Malcolm (1991). ‘Hume’s Tragic Emotions’. Hume Studies 17(2), pp. 93–106. Gorham, Geoffrey (1994). ‘Mind-Body Dualism and the Harvey-Descartes Controversy’. The Journal of the History of Ideas 55(2), pp. 211–34. Hill, Eric (1982). ‘Hume and the Delightful Tragedy Problem’. Philosophy 57(221), pp. 319–26. Merivale, Amyas (2011). ‘Mixed Feelings, Mixed Metaphors: Hume on Tragic Pleasure’. The British Journal of Aesthetics 51(3), pp. 259–69. Neill, Alex (1992). ‘Yanal and Others on Hume on Tragedy’. The Journal of Aesthetics and Art Criticism 50(2), pp. 151–4. ——— (1998). ‘ “An Unaccountable Pleasure”: Hume on Tragedy and the Passions’. Hume Studies 24(2), pp. 335–54. Packer, Mark (1989). ‘Dissolving the Paradox of Tragedy’. The Journal of Aesthetics and Art Criticism 47(3), pp. 211–19. Paton, Margaret (1973). ‘Hume on Tragedy’. The British Journal of Aesthetics 13(2), pp. 121–32.

190  Part II Ochs, Sidney (2004). A History of Nerve Functions: From Animal Spirits to Molecular Mechanisms. Cambridge: Cambridge University Press. Wright, John P. (1983). The Sceptical Realism of David Hume. Manchester: Manchester University Press. Yanal, Robert J. (1991). ‘Hume and Others on the Paradox of Tragedy’. The Journal of Aesthetics and Art Criticism 49(1), pp. 75–6. ——— (1992). ‘Still Unconverted: A Reply to Neill’. The Journal of Aesthetics and Art Criticism 50(4), pp. 324–6.

11 The Sentiments of Beauty

In Chapter 8, we saw that, as a consequence of the specialization view of reason and passion, the passions for Hume cannot be contrary to reason (except in a trivial sense, when they are founded on some false belief). This naturally prompts the question of how the passions can be evaluated beyond the minimal constraint that they be based only on true beliefs. In this final pair of chapters, I turn to the fourth dissertation, Of the Standard of Taste, and to a part of Hume’s reply to this question, namely the part concerning our emotional responses to works of art. In this chapter, I introduce the problem of providing a standard of taste (§11.1), and describe the outline of Hume’s solution, namely an appeal to the sentiments of true judges (§11.2). I then describe the official characteristics of these true judges (§11.3), before arguing that we need to posit an additional implicit characteristic (§11.4). The chapter concludes with a defence of Hume against the charge of vicious circularity in his account (§11.5). The next chapter will turn to the assessment of Hume’s account more generally.

11.1.  There Is Nothing, in Itself, Beautiful or Deformed At the heart of Hume’s anti-rationalism, as we saw in Chapter 8, is what I have called the specialization view of reason and passion (as opposed to Malebranche’s and Clarke’s hierarchical view). On this picture, both reason and sentiment have a role to play in motivation; roughly, the latter to set the goals, and the former to tell us how to achieve those goals. My focus then was on motivation, but as I noted at the time Hume held a similar view with regard to approbation, in the case of morals as well as art. Reason can assess how well an action or artwork realizes its end (which is a question of fact), but it belongs to sentiment to evaluate the ends themselves: Thus the distinct boundaries and offices of reason and of taste are easily ascertained. The former conveys the knowledge of truth and falsehood: The latter gives the sentiment of beauty and deformity, vice and virtue. (M App1.21, p. 294)

192  Part II In his essay The Sceptic, Hume has the title character endorse something I will refer to as Hume’s sceptical principle: If we can depend upon any principle, which we learn from philosophy, this, I think, may be considered as certain and undoubted, that there is nothing, in itself, valuable or despicable, desirable or hateful, beautiful or deformed; but that these attributes arise from the particular constitution and fabric of human sentiment and affection. What seems the most delicious food to one animal, appears loathsome to another: What affects the feeling of one with delight, produces uneasiness in another. This is confessedly the case with regard to all the bodily senses: But if we examine the matter more accurately, we shall find, that the same observation holds even where the mind concurs with the body, and mingles its sentiment with the exterior appetite. (Sc 8, p. 162; my emphasis) There is no guarantee, from this passage alone, that Hume himself endorsed this sceptical principle (for he is not speaking here in his own voice). But the principle is closely related to the specialization view of reason and passion, being arguably just another way of saying the same thing. And in any case, Hume explicitly endorses the principle himself, at least in the case of art (my present focus), in Of the Standard of Taste: ‘it [is] certain, that beauty and deformity, more than sweet and bitter, are not qualities in objects, but belong entirely to the sentiment, internal or external’ (ST 16, p. 235). The obvious worry, for those who are disinclined to agree with this sceptical principle or the specialization view of reason and passion, is that it leads to relativism. When different people find different things desirable, valuable, or beautiful, on account of their different constitutions (and not because of any differences of opinion concerning the facts), reason—on this view—is not in a position to adjudicate between the rival sets of sentiments. But if not reason, then what? On what basis can we pronounce one constitution or resulting sentiment better than another? Hume himself was keenly aware of this difficulty, and the dissertation Of the Standard of Taste constitutes his attempt to respond to it in the case of beauty in works of art. (His parallel attempt in the case of morals and motivation is of course to be found in the moral Enquiry.) Putting the case forcefully on the other side, he begins this fourth dissertation as follows: The difference, it is said, is very wide between judgment and sentiment. All sentiment is right; because sentiment has a reference to nothing beyond itself, and is always real, wherever a man is conscious of it.1 But all determinations of the understanding are not right; because they have a reference to something beyond themselves, to wit,

The Sentiments of Beauty  193 real matter of fact; and are not always conformable to that standard. Among a thousand different opinions which different men may entertain of the same subject, there is one, and but one, that is just and true. . . . On the contrary, a thousand different sentiments, excited by the same object, are all right: Because no sentiment represents what is really in the object. . . . Beauty is no quality in things themselves: It exists merely in the mind which contemplates them; and each mind perceives a different beauty. . . . [E]very individual ought to acquiesce in his own sentiment, without pretending to regulate those of others. To seek the real beauty, or real deformity, is as fruitless an enquiry, as to pretend to ascertain the real sweet or real bitter. (ST 7, pp. 229–30) The starting point for this relativistic argument is precisely the ­sceptical principle, which as I have said Hume himself accepts (at least in the context of art). But he is anxious to resist the relativist’s inference from this principle to ‘the natural equality of tastes’ (ST 8, p. 231). He is ­anxious, that is to say, to erect a standard of taste, according to which ­sentiments— though not true or false—can nevertheless be assessed as in some other sense right or wrong. This then is the problem Hume sets himself in his fourth and final dissertation. The general shape of his solution is the topic of the present chapter. In the next chapter I will go on to place Hume’s account in the broader context of his philosophy as a whole (specifically his philosophy of emotion), and offer at least a partial defence. The best things to be said in defence of Hume, I will argue, arise in connection with his wider views on the passions. This constitutes the core of my case for thinking that Of the Standard of Taste was a fitting choice for filling the gap left by the abandoned dissertation on geometry.

11.2.  The True Standard of Taste and Beauty The essence of Hume’s response to the relativist, very simply put, is that the right sentiments are those felt by the right sort of people. The ‘standard of taste’ of the dissertation’s title is precisely the sentiments of these people: In each creature, there is a sound and a defective state [of the organ]; and the former alone can be supposed to afford us a true standard of taste and sentiment. (ST 12, pp. 233–4) Thus, though the principles of taste be universal, and nearly, if not entirely the same in all men; yet few are qualified to give judgment on any work of art, or establish their own sentiment as the standard of

194  Part II beauty. . . . Strong sense, united to delicate sentiment, improved by practice, perfected by comparison, and cleared of all prejudice, can alone entitle critics to this valuable character; and the joint verdict of such, wherever they are to be found, is the true standard of taste and beauty. (ST 23, p. 241) Judgements, recall, ‘have a reference to something beyond themselves, to wit, real matter of fact’ (ST 7, p. 230), and real matter of fact affords us the standard of judgement. Sentiments, by contrast, do not have a reference to anything beyond themselves (in that same sense), and so the only things that can afford us a standard of taste are other sentiments. Which sentiments, then? Hume answers: the natural ones, the ones felt in the sound state of the organ, or equivalently those felt by the true judges. This is of course just the barest outline of a sceptical response to the relativist, and many expository and philosophical questions remain. The most immediate question is who these true judges are, or equivalently what the characteristics are that constitute the ‘sound’ state of the organ. Hume lists them in the paragraph just quoted (‘strong sense, united to delicate sentiment’, etc.), and I  will turn to them in more detail in the next section. First, however, some argument is required to confirm that what I have said so far is indeed structurally correct. The text seems to me quite clear, but the matter has been obscured by a pair of influential articles by Jeffrey Wieand (1984) and James Shelley (1994), both of whom see a tension between two apparently conflicting definitions of the standard of taste in Hume’s dissertation. The starting point for their worry is this one sentence, which Shelley describes as ‘perhaps . . . the most important and least understood’ of the whole work: It is natural for us to seek a Standard of Taste; a rule, by which the various sentiments of men may be reconciled; at least, a decision afforded, confirming one sentiment, and condemning another. (ST 6, p. 229) Wieand and Shelley both read this introductory remark as constituting Hume’s definition of the standard of taste; whereupon they are puzzled, because on that reading it appears to contain not one but two definitions (Wieand 1984, p.  130; Shelley 1994, p.  437). Which one is it, then: a ‘rule’ or a ‘decision’? As I read it, however, this sentence is not offering a definition at all, but simply a trivial observation, at the outset of the enquiry, about why we might want a standard of taste. The standard tells us which sentiments are right and which are wrong. If we find ourselves disagreeing, therefore, we can reconcile our sentiments by submitting them to that standard. Failing that—i.e. if any party to the dispute refuses to budge—those of

The Sentiments of Beauty  195 us in possession of the standard can at least use it to settle the matter to our own satisfaction. This, it seems to me, is all that Hume is saying here. I submit that mine is the more natural reading. Indeed, it would never have occurred to me to read it in any other way, but for Wieand’s and Shelley’s discussions. If a stronger argument is required, however, I may add that my interpretation has the merit of squaring better with the rest of the text: as the quotations at the start of this section show, Hume defines the standard as neither a rule nor a decision, but as a set of sentiments. It also avoids the needless difficulty of trying to reconcile two conflicting definitions. According to Shelley, the puzzle does not only arise in virtue of this one sentence: Even if we ignore what I have been calling Hume’s definition of the standard of taste [the sentence just examined], at least one problem mentioned above will not vanish: it will still appear that Hume has not one standard of taste but two. Shortly after giving his definition, Hume identifies the standard with what he variously calls ‘rules of composition,’ ‘laws of criticism,’ and ‘rules of art,’ and which he defines as ‘general observations, concerning what has been universally found to please in all countries and in all ages’ (pp. 231–7). Later he claims that ‘the true standard of taste and beauty’ consists of the joint verdict of ‘true judges’ . . . (pp. 234–42). Hume’s reader, meanwhile, is left with the task of somehow reconciling these two standards, or, at least, of confirming one standard and condemning the other. (Shelley 1994, pp. 437–8) But the task Shelley here sets us is an easy one: the paragraph he himself quotes (and which I too quoted at the outset) confirms the second of these two standards, namely the joint verdict of the true judges. I  condemn Shelley’s first proposed standard, meanwhile, on the grounds that Hume nowhere identifies his standard of taste with the rules of art or criticism. I cannot guess what prompted the idea that the standard of taste consists of the rules of art. Wieand attributes it to Stuart Brown and Peter Jones, though I  can find in their articles no such claim (Wieand 1984, p. 129; citing Brown 1938 and Jones 1976a, 1976b). Brown and Jones do talk about the rules of art, but as far as I can tell they don’t equate them with the standard, any more than Hume himself does. These rules are certainly an important part of Hume’s overall story, and I will examine them in §12.4. But there is no danger that the mention of them might contradict Hume’s specification of the standard of taste itself. That standard is perfectly clear: the right sentiments are the natural ones, the ones felt in the sound state of the organ, or equivalently those felt by the true judges (whose organs are by definition sound). On Shelley’s interpretation, the joint verdict of the true judges is at least one of Hume’s standards. A  lot of what he says concerning the

196  Part II rules of art, moreover, seems to me perfectly correct. The chief difficulty with his interpretation is that it is more complicated than it needs to be, with these complications not being supported by the text. I am less sympathetic with Wieand’s account, which views the joint verdict merely as a guide to the right sentiments, insisting that the rules of art actually constitute the standard itself: The standard of taste cannot consist in the verdicts of true judges, because these judges may be wrong. This indicates that there is an independent standard of taste, namely the rules of art. But although Hume thinks that these rules are the standard, he also thinks that the verdicts of true judges are a good guide to what the rules are, and so function as a practical standard of taste. (Wieand 1984, p. 129) This simply does not agree with the text. The standard of taste must consist of the verdict of true judges, since Hume unambiguously states that it does. Does this mean we must invert Wieand’s reasoning in the passage just quoted, and conclude (in line with Shelley) that the true judges can never be wrong, and hence that they are an idealization that is never in fact realized? I do not think so; it seems to me this would be a peculiar and unmotivated way of speaking. We do not say that an honest person can never lie, or that a generous person can never be mean. It is of course true that they cannot do these things insofar as they are honest or generous, but it would be strange to insist that possession of such characteristics requires the perfect realization of them at all times. Just so, I  suggest we say that the true judges can never be wrong insofar as they are true judges, although obviously in practice they must all be allowed the occasional bad day. To be clear, I do not offer this last claim as a matter of textual interpretation (I am not aware of any text that would settle it one way or the other), but as a matter of common sense. In any case, it seems to me that the debate about whether Hume’s true judges are real or ideal is a largely verbal one, on which nothing of wider significance hangs. If one prefers to say that the true judge must always be a true judge, it will follow that the true judge can never be wrong, and hence that the notion is an unrealizable ideal. To repeat, that seems to me an odd way of speaking. But my objection to idealization interpretations runs no deeper than that.2

11.3.  Strong Sense, United to Delicate Sentiment The next question concerns the character of Hume’s true judges. It is usual for commentators to describe Hume as offering five defining characteristics, although to be ingenious I count only two: delicacy of taste

The Sentiments of Beauty  197 and good sense. The others are either ways of honing the former (practice and comparison), or components of the latter (freedom from prejudice). To be still more ingenious, there is also a third implicit characteristic, not commonly remarked upon, that I will draw out in the next section (namely sound moral judgement). But how we count the features is a matter of no real importance. Delicacy of taste or imagination, first, is a heightened sensitivity to the causes of the sentiment of beauty: Though it be certain, that beauty and deformity, more than sweet and bitter, are not qualities in objects, but belong entirely to the sentiment, internal or external; it must be allowed, that there are certain qualities in objects, which are fitted by nature to produce those particular feelings. Now as these qualities may be found in a small degree, or may be mixed and confounded with each other, it often happens, that the taste is not affected with such minute qualities, or is not able to distinguish all the particular flavours, amidst the disorder, in which they are presented. Where the organs are so fine, as to allow nothing to escape them; and at the same time so exact as to perceive every ingredient in the composition: This we call delicacy of taste, whether we employ these terms in the literal or metaphorical sense. (ST 16, p. 235) Hume illustrates what he has in mind here with a story from Don Quixote, in which Sancho Panza reports that two of his ancestors were able to discern a slight taste of iron and leather respectively in an otherwise fine wine. Sure enough, when the barrel was emptied, an iron key with a leather cord attached was found at the bottom. Hume’s delicacy of taste is intended to be the mental equivalent of this refined palate. Some people, Hume seems to think, will naturally have a more delicate taste than others. But the skill can be honed with sufficient practice: But though there be naturally a wide difference in point of delicacy between one person and another, nothing tends further to encrease and improve this talent, than practice in a particular art, and the frequent survey or contemplation of a particular species of beauty. (ST 18, p. 237) So as to fix the proper degree of beauty or deformity, furthermore, true judges must also be experienced in a range of works. Otherwise they will be in danger of rating too highly something that is merely quite good: A man, who has had no opportunity of comparing the different kinds of beauty, is indeed totally unqualified to pronounce an opinion with regard to any object presented to him. By comparison alone we fix

198  Part II the epithets of praise or blame, and learn how to assign the due degree of each. . . . The most vulgar ballads are not entirely destitute of harmony or nature; and none but a person, familiarized to superior beauties, would pronounce their numbers harsh, or narration uninteresting. . . . One accustomed to see, and examine, and weigh the several performances, admired in different ages and nations, can alone rate the merits of a work exhibited to his view, and assign its proper rank among the productions of genius. (ST 20, p. 238)3 As I noted at the outset, practice and comparison are not distinct characteristics of the true judge, but rather necessary exercises for honing one’s delicacy of taste. As Hume says, the true judge has a ‘delicate sentiment, improved by practice, perfected by comparison’ (ST  23, p.  241; my emphases); see also Jones (1976b, pp.  333–4) and Wieand (1984, pp. 135–6). The need for practice and comparison is not, for Hume, something peculiar to the sentiments of beauty and deformity, or to the appreciation of art. On the contrary, he maintains that it is a quite general feature of human psychology that our sentiments require experience and comparison to settle into something suitably ordered: [I]f a person full grown, and of the same nature with ourselves, were on a sudden transported into our world, he would be much embarrassed with every object, and would not readily determine what degree of love or hatred, of pride or humility, or of any other passion should be excited by it. The passions are often varied by very inconsiderable principles; and these do not always play with perfect regularity, especially on the first trial. But as custom or practice has brought to light all these principles, and has settled the just value of every thing; this must certainly contribute to the easy production of the passions, and guide us, by means of general established rules, in the proportions, which we ought to observe in preferring one object to another. (P 2.47, pp. 16–7; see also T 2.1.6.9, pp. 293–4) My main case for placing Of the Standard of Taste in the context of Hume’s philosophy of emotion will be given in the next chapter. But the present observation may serve as a hint of the fuller argument to come. The sentiments of the true judges determine the standard of taste; but it is the science of the passions that determines what the true judges must be like. The next feature of the true judge that Hume discusses is freedom from prejudice (though this is later subsumed under the more general characteristic of good sense, as we will see). In general, this amounts to an ability

The Sentiments of Beauty  199 to place oneself ‘in that point of view, which the performance supposes’ (ST 21, p. 239). But there are two notable ways, for Hume, in which one might fail to do this, and consequently we can distinguish two aspects of freedom from prejudice. I will call these impartiality and imaginative flexibility respectively. Though the terms are mine, the distinction is very clearly Hume’s. It is moreover an important one, for the failure of commentators to draw it before now has led to some unfortunate confusions. Impartiality, first, is a simple matter of avoiding any personal prejudice arising from one’s relationship with the author, and forgetting any interest one may have in the success of their work. Imaginative flexibility, second, is a matter of avoiding any cultural prejudices, and on the contrary entering (as far as possible) into the beliefs and prejudices of those for whom the work was originally intended. Here is Hume’s account of this dual characteristic (for clarity, I  have explicitly separated the two aspects): [General note:] We may observe, that every work of art, in order to produce its due effect on the mind, must be surveyed in a certain point of view, and cannot be fully relished by persons, whose situation, real or imaginary, is not conformable to that which is required by the performance. [Imaginative flexibility:] An orator addresses himself to a particular audience, and must have a regard to their particular genius, interests, opinions, passions, and prejudices; otherwise he hopes in vain to govern their resolutions, and inflame their affections. . . . A critic of a different age or nation, who should peruse this discourse, must have all these circumstances in his eye, and must place himself in the same situation as the audience, in order to form a true judgment of the oration. [Impartiality:] In like manner, when any work is addressed to the public, though I should have a friendship or enmity with the author, I must depart from this situation; and considering myself as a man in general, forget, if possible, my individual being and my peculiar circumstances. [General note:] A  person influenced by prejudice, complies not with this condition; but obstinately maintains his natural position, without placing himself in that point of view, which the performance supposes. [Imaginative flexibility:] If the work be addressed to persons of a different age or nation, he makes no allowance for their peculiar views and prejudices; but, full of the manners of his own age and country, rashly condemns what seemed admirable in the eyes of those for whom alone the discourse was calculated.

200  Part II [Impartiality:] If the work be executed for the public, he never sufficiently enlarges his comprehension, or forgets his interest as a friend or enemy, as a rival or commentator. (ST 21, p. 239; paragraph breaks mine) Emphasizing the requirement of imaginative flexibility (though not in those terms), Michelle Mason has remarked that ‘Hume’s true judge [is] less an impartial observer than a cultural chameleon’ (Mason 2001, p. 60). But the passage just quoted should make it clear that Hume’s true judge is both an impartial observer and a cultural chameleon. Peter Kivy has recently argued that there is a contradiction here, or at least a tension between these two aspects of the requirement for freedom from prejudice (Kivy 2011). He begins by calling attention to two ways of engaging with works of art from different times or cultures, calling these ‘the method of the unchanging human heart’ (following C. S. Lewis) and ‘the method of historicism’ respectively. In the former case, we are to concentrate on what is common to all human beings everywhere, stripping away the peculiarities of time and place to find what is of eternal value. In the latter case, we are instead to embrace those peculiarities. As Lewis nicely puts it (in his admirable defence of the historicist approach): ‘Instead of stripping the knight of his armour you can try to put his armour on yourself; instead of seeing how the courtier would look without his lace, you can try to see how you would feel with his lace’ (Lewis 1942, p. 64). Hume’s requirement of imaginative flexibility is precisely a c­ ommitment to historicism, as Kivy rightly notes. The difficulty arises because Kivy reads the requirement of impartiality as a commitment to anti-­historicism, to the method of the unchanging human heart. But there is no convincing basis for this reading, and it seems to have arisen from a failure to appreciate the different contexts of the two requirements. Imaginative flexibility is called for when engaging with works from a different age or nation (hence Hume’s is indeed a historicist view). I­ mpartiality, however, is necessary when engaging with works by artists one has a special relationship with, when there is a danger of this relationship colouring our judgement. If we conflate these different contexts, Hume’s remarks on impartiality might take on an anti-historicist appearance. When we keep them apart, however, as they clearly are in Hume’s text, the tension Kivy tries to bring out does not arise. In the next paragraph, Hume goes on to claim that the true judge must have good sense more generally, of which freedom from prejudice is just one (albeit particularly notable) component: It is well known, that in all questions, submitted to the understanding, prejudice is destructive of sound judgment, and perverts all operations of the intellectual faculties: It is no less contrary to good taste; nor has it less influence to corrupt our sentiment of beauty. It

The Sentiments of Beauty  201 belongs to good sense to check its influence in both cases; and in this respect, as well as in many others, reason, if not an essential part of taste, is at least requisite to the operations of this latter faculty. (ST 22, p. 240)4 The last sentence here indicates that ‘good sense’, for Hume, was just another term for ‘reason’, the faculty by which we judge of truth and falsehood. Hume seems to have viewed this faculty in normative and functional terms, as that aspect of our minds that gives rise to true beliefs, to sound judgements (see Millican 2009, especially section 4.2). But lest this suggestion be controversial, I  need only note here that good sense must undoubtedly be taken in this way; else it would not belong to good sense to check the perverting influence of prejudice. The true judge is not merely someone who has the ability to reason, but someone who has the ability to reason well. In addition to keeping our prejudices in check, good sense or reason also enables us to ‘perceive the consistence and uniformity’ of complex works of art; to judge whether a work is fitted to attain the ‘end or purpose, for which it is calculated’; and to discern whether ‘[t]he persons introduced in tragedy and epic poetry’ behave ‘suitably to their character and circumstances’ (ST  22, p.  240). The middle of these three points, incidentally, helps to explain Hume’s historicism, and why it is properly seen as a component of good sense. ‘The object of eloquence is to persuade’, Hume says, ‘of history to instruct, of poetry to please by means of the passions and the imagination’ (ST  22, p.  240). The true judge will deem a work good if it persuades or pleases those for whom it was intended, making allowances for their particular opinions and prejudices. Officially, then, this is the character of Hume’s true judge: a person with ‘strong sense, united to delicate sentiment’ (ST 23, p. 241), and all that these two qualities entail. There is however a complication. As I will argue in the next section, we must acknowledge a third characteristic of Hume’s true judge, though it is one Hume does not state explicitly.

11.4.  The Case Is Not the Same With Moral Principles Towards the end of the fourth dissertation, and in the context of ‘the celebrated controversy concerning ancient and modern learning’ (ST 32, p. 245), Hume places some moral limits on how flexible the true judge should be. Imaginative flexibility requires that we make allowances for ‘any innocent peculiarities of manners’ (ST  32, p.  245), he now clarifies, and for ‘speculative errors’ (ST 34, p. 247), including in particular those that concern religion. But the line is drawn with regard to moral errors: ‘where the ideas of morality and decency alter from one age to another, and where vicious manners are described, without being marked with the proper characters of blame and disapprobation; this must be allowed to disfigure the poem, and to be a real deformity’ (ST 32, p. 246).

202  Part II Characteristically for Hume, this enables him to take a middle ground in the ancient-modern debate. Ancient art should not be faulted for its innocent peculiarities and speculative errors, but can legitimately be criticized for its moral failings. The question prompted by this restriction, of course, is why moral differences are so special, when compared with differences of manners or speculative opinions. Of interpretative interest in its own right, this question is also an urgent one in the evaluation of Hume’s account. For there is a natural suspicion here, if not of inconsistency, then at least of capriciousness on Hume’s part: how can one be in favour of imaginative flexibility in general, but imaginative rigidity in the special case of moral principles? Richard Moran thinks Hume has no satisfactory answer to this question, offering merely ‘a few somewhat halfhearted attempts at explanation’, which moreover ‘do not appear to strike Hume himself as terribly persuasive’ (Moran 1994, p. 98). That would be an unfortunate state of affairs. Even more alarming is Christopher MacLachlan’s incredible diagnosis: that Hume’s position is so hopelessly contradictory on this score (and others) that the whole essay must be ironic (MacLachlan 1986, pp. 26–7). Reed Winegar has recently attempted to find in Hume’s text a clear and consistent explanation of why the true judges, though imaginatively flexible in most things, nevertheless hold firm in their moral principles (Winegar 2011). His account consists of two main points: first, that freedom from prejudice is a consequence or component of good sense; and second, that good sense requires us to overlook different manners and speculative opinions, but not to overlook moral differences. Therefore, ‘since a true judge is supposed to employ her good sense when evaluating a work, she should apply her own moral standards to artworks with morally flawed outlooks’ (2011, p. 29). The first of these points seems to me correct, and clearly supported by the text; I already endorsed it in the previous section, and am grateful to Winegar for drawing it to my attention. The second point, however, is less convincing, as I will argue once I have stated my positive alternative. The right interpretation of Hume on this matter, I suggest, is also the simplest: that the limits of imaginative flexibility imposed on the true judges are simply those discovered in all of us. According to Hume, it is simply a fact of human psychology that we cannot enter imaginatively into a different moral point of view; although the same, crucially, is not true regarding different speculative opinions: [W]hatever indulgence we may give to the writer on account of his prejudices, we cannot prevail on ourselves to enter into his [different moral] sentiments, or bear an affection to characters, which we plainly discover to be blameable. (ST 32, p. 246)

The Sentiments of Beauty  203 The case is not the same with moral principles, as with speculative opinions of any kind. . . . Whatever speculative errors may be found in the polite writings of any age or country, they detract but little from the value of those compositions. There needs but a certain turn of thought or imagination to make us enter into all the opinions, which then prevailed, and relish the sentiments or conclusions derived from them. But a very violent effort is requisite to change our judgment of manners, and excite sentiments of approbation or blame, love or hatred, different from those to which the mind from long custom has been familiarized. (ST 33, pp. 246–7) I take it to be a matter of no consequence, incidentally, whether this kind of moral flexibility is absolutely impossible (as Hume first suggests), or merely very difficult (as he later claims); for even in the latter case, works requiring us to enter into a different moral outlook will please the majority of true judges considerably less than they otherwise would, and that is all Hume’s account requires. Hume himself doesn’t go into any further detail on this point, but I like the suggestion of Michelle Mason, developed (apparently independently) at greater length by Eva Dadlez, that the problem with such moral mismatches is that they preclude imaginative and emotional engagement just when such engagement is called for. As Dadlez puts it: Fictional endorsements of conduct otherwise regarded as vicious constitute defects because they curtail the aesthetic experience. Hume’s claim, then, is as much about human psychology as it is about morality. We cannot divorce morality from narrative art because, for Hume, morality finds its basis in the very sentiments such art is intended to arouse. The content of the works under fire is objectionable because it prompts the imaginative disengagement of the reader or viewer just when engagement and emotional participation are necessary for full appreciation. (Dadlez 2002, pp. 143–4; see also Mason 2001, p. 66) This quotation is taken from the introduction of Dadlez’s article, and the idea naturally calls for elaboration and defence; but I defer to her for a fuller treatment.5 For present purposes, it suffices to say that this was Hume’s view, whatever deeper reasons may be given in support of it. To return to Winegar’s interpretation, the first thing to note is that merely stating that good sense entails imaginative rigidity with regard to one’s moral sentiments does not by itself solve the problem; we still need to be told why it has this curious feature. Winegar spends surprisingly little time on this point, though he does have some explanation to offer. His suggestion is that good sense requires us to enter into the intended

204  Part II audience’s prejudices only so as ‘to avoid being merely obstinate and rash’: ‘But a true judge who employs her own moral standards is not simply obstinate. Nor is her condemnation of immoral manners rash’ (2011, p. 30). Thus good sense does not require her to abandon these standards. The textual evidence offered in support of this proposal is Hume’s earlier claims that the prejudiced person ‘obstinately maintains his natural position’, and ‘rashly condemns what seemed admirable in the eyes of those for whom alone the discourse was calculated’ (ST 21, p. 239; Winegar’s emphases). But this is half an argument at best, and there is no complementary passage in the text where Hume says that holding fast to one’s moral standard is, by contrast, neither obstinate nor rash. What Hume in fact says, as we have seen, is that holding fast to one’s moral standard is psychologically inevitable. On Winegar’s interpretation, this claim—made at precisely the point where we would expect some explanation for the moral limits of imaginative flexibility—is irrelevant to that explanation; and good sense, which is not mentioned here at all, is apparently the key. In all, Winegar’s Hume is a curiously obscure presenter of his views, whereas my Hume tells us quite clearly what he thinks. Our reconstruction of this aspect of Hume’s account, however, is not yet complete. For as things stand, we simply have a relativistic impasse between true judges from (let us say) ancient Greece on the one hand, and modern Europe on the other, both of whom may be allowed to have good sense and a delicate taste, but who unavoidably have different preferences as a result of their different moral outlooks. But Hume does not make this concession to the relativist. Rather, he insists that modern art is better than ancient art, on account of its improved moral suppositions. The point about the psychological limits of imaginative flexibility by itself does not support this conclusion. Winegar seems at times to have grasped this problem, for instance, when he asks why Hume ‘should . . . not just say that only judges from Homer’s own culture who share his ancient Greek values can ever qualify as true judges of his poems’ (2011, pp. 26–7). He raises this question in criticism of Mason’s account, which he argues is not able to answer it. The criticism is fair, but it does not follow that Mason’s interpretation is wrong, merely that it is incomplete.6 His own interpretation, moreover, is equally unable to answer this question. In virtue of her good sense, Winegar tells us, a true judge will ‘apply her own moral standards’ in the assessment of artworks; and as a consequence of this, ‘a person who enters into vicious sentiments is not a true judge’ (2011, pp. 29, 30; my emphases). But the inference does not follow. What follows is merely that a person who enters into someone else’s moral sentiments is not a true judge. If that person has a flawed moral judgement to begin with, nothing yet explains why she should not qualify as having good sense. I can see no way around this difficulty except to acknowledge—as I  have already advertised—a tacit third characteristic of Hume’s true

The Sentiments of Beauty  205 judge, namely sound moral judgement. I hasten to add that this is only the outline of an answer, for it will immediately be wondered why the true judge of art is required to be a true judge of morals as well, something that is by no means obvious. But at least we have, for now, clearly identified the question. What is more, it is a question that can naturally be subsumed under the broader question of why it is the sentiments of Hume’s true judges—i.e. those of good sense, delicate sentiment, and (we now see) sound moral judgment—that set the standard for all of us. And this, I take it, is the fundamental question in the assessment of Hume’s account. I will turn to it in the next chapter. To end this chapter, however, a brief digression is in order to settle an old but hitherto unsettled objection.

11.5.  The Best Way of Ascertaining It The objection I  have in mind is the allegation of circularity in Hume’s account, whereby, as Kivy puts it, ‘good art is defined in terms of the good critic and the good critic in terms of good art’ (Kivy 1967, p. 60). This worry has not received much scholarly attention recently, but it is a staple of many undergraduate philosophy of art modules, and one of the three problems discussed in Christopher Williams’s recent Philosophy Compass article on Hume’s aesthetics (Williams 2007). (The other two problems are the role of rules in grounding the judges’ verdicts, and the relevance of their verdicts for the rest of us. These problems are closely related, in my view, and will be addressed in the next chapter.) It is completely groundless, but since I have never seen it satisfactorily answered in print, it seems worth spending some time setting the matter straight here. The problem is not with good sense or freedom from prejudice. No one, so far as I know, has ever thought these qualities of the true judge are in danger of landing Hume in a circle. The worry is exclusively with delicacy of taste, and with the practice and comparison that are needed to refine and perfect this quality. I will turn to delicacy itself in a moment. First, however, here is the case with regard to practice and comparison, as stated by Kivy: Practice Hume thinks of as ‘the frequent survey of a particular species of beauty’. Use of comparisons requires juxtaposing ‘the several species and degrees of excellence’. But we must be able to recognize the beautiful before we are able to determine whether a critic has or has not been engaged in ‘the frequent survey or contemplation of a particular species of beauty’. We must know what is excellent before we are able to determine whether or not a critic has compared ‘the several species and degrees of excellence’. (Kivy 1967, pp. 60–1)

206  Part II It is easy to see that there is no real difficulty here, for Kivy’s interpretations of practice and comparison are mistaken. The passages that he quotes might suggest, out of context, that Hume thought true judges must have more experience of exclusively good works of art. But by far the more plausible interpretation—both in itself, and when these passages are seen in context—is that he held true judges need simply to have more experience of works of art in general, both good and bad. In the sentences immediately following the first passage Kivy quotes, Hume talks of ‘beauties and defects’ (ST 18, p. 237; my emphasis), indicating that the ‘species of beauty’ in question is rather a species of attempted or intended beauty, a set that will include bad instances of the type alongside the good. We must give a similar interpretation to the ‘several species and degrees of excellence’, which are said to include ‘[t]he coarsest daubing’ and ‘[t]he most vulgar ballads’ (ST  20, p.  238) alongside the genuinely excellent works. There is thus no circularity here. Turning now to delicacy of taste itself, let us first recall Hume’s definition of this property, already quoted: ‘Where the organs are so fine, as to allow nothing to escape them; and at the same time so exact as to perceive every ingredient in the composition: This we call delicacy of taste’ (ST  16, p.  235). There is no reference to good works of art here, and hence no apparent circle. Why then would anyone have thought there is a problem? The worry stems from Hume’s subsequent claim, in the next paragraph, that ‘the best way of ascertaining’ whether someone has delicacy of taste ‘is to appeal to those models and principles, which have been established by the uniform consent and experience of nations and ages’ (ST  17, p.  237). Here Hume is appealing to good works of art, or principles linking works with the positive sentiments such as beauty, whence the threat of circularity.7 The first thing to say in constructing a response to this objection concerns the nature of the circle in question. As just noted, there is no hint of circularity in Hume’s definition of delicacy of taste. The difficulty, rather, concerns an evidential claim about how to ‘ascertain’ whether or not someone has this property. The second thing to say is that evidential circles are not in themselves problematic. On the contrary, we can acknowledge that there is an evidential circle here: being approved of by (someone who appears to be) a good critic is indeed evidence that something is a good work of art, and approving of (something that appears to be) a good work of art is indeed evidence that someone is a good critic, and hence in particular that they have delicacy of taste. The trouble would arise if this was the only evidence that we had in either direction, so that there was no way into the circle from the outside. What evidence, then, might get us into this circle? Hume also discusses delicacy of taste in his essay Of the Delicacy of Taste and Passion, in which—as one may guess from the title—he contrasts it with delicacy of passion. Whereas delicacy of taste, as we have seen, is a heightened sensitivity to (the causes of) the sentiments of beauty

The Sentiments of Beauty  207 and deformity, delicacy of passion is a corresponding sensitivity to the general ups and downs of life: it makes those suffering from it ‘extremely sensible to all the accidents of life, and gives them a lively joy upon every prosperous event, as well as a piercing grief, when they meet with misfortunes and adversity’ (DT 1, pp. 3–4). Kivy, although he thinks that practice and comparison do give rise to a vicious circle (as just discussed), argues that delicacy of taste itself is not thus problematic. His proposed solution is to suppose that there is a good correlation between delicacy of taste and delicacy of passion. In order to get into the circle, then, we need only look out for someone with a delicate passion (which we can do without reference to works of art at all, let alone good ones), for the presence of this characteristic will be good evidence that they have a delicate taste as well. Unfortunately, Kivy’s supposition of a correlation between delicacy of taste and delicacy of passion is in itself rather implausible (for a criticism, see Carroll 1984). More importantly in the present context, it is also something that Hume himself explicitly denies. The point of the essay Of the Delicacy of Taste and Passion is precisely to recommend cultivating the former but steering clear of the latter. This advice would hardly count for much if Hume thought that, as a matter of fact, you couldn’t have one without the other. We do not, it seems to me, need anything so ingenious as Kivy’s proposal to get ourselves into Hume’s evidential circle, and materials can already be found in Of the Standard of Taste itself. For instance, Hume did not think we necessarily needed good critics in order to establish the established models. These could be established instead by their ability to stand the test of time (as with Homer; ST 11, p. 233). Another option would be to find someone with good sense, who is free from prejudice (you may be such a person yourself), and then see how they (you) feel about works that don’t require a great deal of delicacy to appreciate, the ones whose features are presented ‘singly and in a high degree’ (ST 16, p. 235). However you do it, gather some evidence that such-and-such is a good feature of works of art. If you then find yourself enjoying works that possess that quality ‘in a smaller degree’ (ibid.), you have grounds for thinking that your taste is delicate. Yet another option, perhaps a more commonsensical one, would be for a judge to display the delicacy of their taste directly, by critically analysing a work of art, and pointing out its finer features—the equivalent of the key with the leather cord from the Don Quixote story—that others, perhaps, have missed. These are just a few possibilities that occur more or less immediately from reading Hume’s text. I dare say more evidential strategies could be proposed in a broadly Humean spirit. To repeat, it is not an objection to Hume that there is a circle here; evidence works both ways, and so the existence of a circle is only to be expected. The only worry would be if there was no way into the circle from the outside. But no one has given any reasons for thinking there is not, and while Kivy’s proposed route is

208  Part II a non-starter, the remarks above suggest several more promising points of entry. There is no significant problem of circularity in Hume’s account.

Notes 1 Hume’s language here is reminiscent of the problematic premise in his representative quality argument, which states that ‘[w]hen I am angry, I am actually possest with the passion, and in that emotion have no more a reference to any other object, than when I  am thirsty, or sick, or more than five foot high’ (T 2.3.3.5, p. 415). As I argued in §7.2, Hume’s point there is not that the passions have no intentional objects, but rather that they are not truth-apt representations of reality. It should be clear, I trust, that Hume means exactly the same thing here. This passage may indeed be considered further confirmation of my earlier claim. 2 For a more robust defence of a realist interpretation, see Stephanie Ross (2008). 3 See also Philo in the Dialogues: ‘[I]t is impossible for us to tell, from our limited views, whether this system [i.e. the universe] contains any great faults, or deserves any considerable praise, if compared to other possible, and even real systems. Could a peasant, if the Aeneid were read to him, pronounce that poem to be absolutely faultless, or even assign to it its proper rank among the productions of human wit; he, who had never seen any other production?’ (D 5.6, p. 167). 4 That freedom from prejudice is a component of good sense, rather than a distinct characteristic, seems to have passed most readers by, myself included. My thanks to Reed Winegar (2011) for having drawn it to my attention. In other respects, I am less persuaded by Winegar’s argument in this article, which I will have cause to examine in the next section. 5 I also recommend Weinberg and Meskin (2006) to the curious reader, which provides a modern discussion of the psychology of the imagination and its limits that is, to my mind, very Humean in spirit, and serves to complement Dadlez’s account nicely. 6 For the record, Winegar’s criticism of Mason’s proposed distinction between the true judge and the ‘man of learning and reflection’ (ST 31, p. 245)—where the latter supposedly can recognize his cultural prejudices, but cannot shake them off—also seems fair. As I read the relevant passages, the man of learning and reflection just is the man of good sense, as Winegar argues. In other respects, however, my account is closer in spirit to Mason’s. 7 It was this passage that caused S. G. Brown to pose the original problem, in the earliest statement of it that I have come across (Brown 1938).

Bibliography Brown, S. G. (1938). ‘Observations on Hume’s Theory of Taste’. English Studies 20(5), pp. 193–8. Carroll, Noel (1984). ‘Hume’s Standard of Taste’. The Journal of Aesthetics and Art Criticism 43(2), pp. 181–94. Dadlez, Eva (2002). ‘The Vicious Habits of Entirely Fictive People: Hume on the Moral Evaluation of Art’. Philosophy and Literature 26(1), pp. 143–56. Jones, Peter (1976a). ‘Hume’s Aesthetics Reassessed’. The Philosophical Quarterly 26(102), pp. 48–62.

The Sentiments of Beauty  209 ——— (1976b). ‘Cause, Reason, and Objectivity in Hume’s Aesthetics’. In Hume: A Re-Evaluation. Eds. by D. W. Livingston and J. T. King. New York: Fordham University Press, pp. 323–42. Kivy, Peter (1967). ‘Hume’s Standard of Taste: Breaking the Circle’. The British Journal of Aesthetics 7(1), pp. 57–66. ——— (2011). ‘Remarks on the Varieties of Prejudice in Hume’s Essay on Taste’. The Journal of Scottish Philosophy 9(1), pp. 111–4. Lewis, C. S. (1942). A Preface to Paradise Lost. Oxford: Oxford University Press. MacLachlan, Christopher (1986). ‘Hume and the Standard of Taste’. Hume Studies 12(1), pp. 18–38. Mason, Michelle (2001). ‘Moral Prejudice and Aesthetic Deformity: Rereading Hume’s “Of the Standard of Taste” ’. The Journal of Aesthetics and Art Criticism 59(1), pp. 59–71. Millican, Peter (2009). ‘Hume on Induction and the Faculties’. Draft article. Available at https://davidhume.org/scholarship/papers/millican Moran, Richard (1994). ‘The Expression of Feeling in Imagination’. The Philosophical Review 103(1), pp. 75–106. Ross, Stephanie (2008). ‘Humean Critics: Real or Ideal?’ The British Journal of Aesthetics 48(1), pp. 20–8. Shelley, James (1994). ‘Hume’s Double Standard of Taste’. The Journal of Aesthetics and Art Criticism 52(4), pp. 437–45. Weinberg, Jonathan M. and Aaron Meskin (2006). ‘Puzzling over the Imagination: Philosophical Problems, Architectural Solutions’. In Shaun Nichols, ed. The Architecture of the Imagination: New Essays on Pretence, Possibility, and Fiction. Oxford: Oxford University Press, pp. 175–204. Wieand, Jeffrey (1984). ‘Hume’s Two Standards of Taste’. The Philosophical Quarterly 34(135), pp. 129–42. Williams, Christopher (2007). ‘Some Questions in Hume’s Aesthetics’. Philosophy Compass 2(2), pp. 157–69. Winegar, Reed (2011). ‘Good Sense, Art, and Morality in Hume’s “Of the Standard of Taste” ’. The Journal of Scottish Philosophy 9(1), pp. 17–35.

12 The Laws of Criticism

In the previous chapter, we stated the problem: how can there be a standard of taste, which pronounces some sentiments right and others wrong, given the sceptical principle that beauty (value, desirability) is not a quality of objects themselves? We saw the basic structure of Hume’s reply: the right sentiments are those felt by the true judges, i.e. people with good sense, a delicate taste, and sound moral judgement. The pressing question, I take it, is why these people should set the standard for the rest of us. Why, in the first instance, should we prefer Hume’s true judges to some other ‘false’ alternative? Or, to put it in less contentious terms, why suppose that Hume’s judges are indeed the ‘true’ ones? Second, and more profoundly, why suppose that any judges are the true ones, instead of embracing a more liberal relativism in which everyone sets their own standard? These questions both admit of two answers, an easy one and a hard one. I will begin, naturally enough, with the easy answers, which may be thought satisfactory up to a point (§12.1). The hard answers go deeper, however, and not only provide a stronger case against the relativist (§12.3 and §12.4), but also help to answer two other questions regarding Hume’s account (§12.4 and §12.5). Crucially, they also require us to place that account in the broader context of Hume’s philosophy of emotion (§12.2).

12.1.  The Side of Common Sense and Reason Assuming for the moment that we have already been persuaded to reject relativism, and to embrace some standard of taste, our first question is why we should suppose that the true standard is the one Hume proposes, or that his judges are indeed the ‘true’ ones. The easy answer to this question is that Hume’s true judges are, quite simply, the only credible choice. For are we really to suppose instead that the right sentiments are those felt by dull-witted, ignorant, insensitive, biased, and unimaginative people, with a perverted sense of morals and little or no experience of the art they are engaging with? That is surely not a position anyone could seriously defend.

The Laws of Criticism  211 If need be, we can firm up this easy answer by saying that the characteristics of the true judge are all relevant virtues in the appreciation of art. That they are virtues ought to be uncontroversial. At any rate Hume takes it to be so: ‘that such a character is valuable and estimable will be agreed in by all mankind’ (ST 25, p. 242). By calling them relevant, meanwhile, I simply mean that they have an influence on our responses to works of art. Partiality may cause us to overestimate the work of a friend, or to underestimate that of an enemy (ST 21, p. 239); good sense is necessary to judge whether characters in a drama behave in accordance with their nature and circumstances (ST  22, p.  240); without delicacy of sentiment, various features of a work of art will simply pass us by (ST 17, p. 236); without imaginative flexibility, our appreciation will be limited to a very small set of local, recent works (ST 32, pp. 245–6); and given the psychological limits of this flexibility, in particular with regard to moral sensibilities, it is impossible for us to fully enjoy works that presuppose or require a different moral outlook (ST 32–3, pp. 245–7). The more profound question, however, is why we should prefer the standard set by Hume’s true judges to the relativist alternative in which everyone sets their own standard. If someone is wrong, it may be allowed that Hume’s true judges provide the obvious standard by which to measure success, but why suppose that anyone is wrong? Why not suppose instead that all sentiments are right for those who have them, and leave it at that? The easy answer to this question is that we can suppose there is a standard if we like, and that we must do so if we want to use the language of beauty and engage in discussions or debates about the quality of art. This is because the language of beauty and debates about the beautiful presuppose that there is something here that is not merely a matter of personal taste. They presuppose, that is to say, that there is something independent of anyone’s particular preferences and prejudices, something about which everyone can and should agree. There is little direct evidence of this easy answer in Of the Standard of Taste itself, but the parallel idea in the case of ethics is perhaps clearer in the moral Enquiry: When a man denominates another his enemy, his rival, his antagonist, his adversary, he is understood to speak the language of selflove, and to express sentiments, peculiar to himself, and arising from his particular circumstances and situation. But when he bestows on any man the epithets of vicious or odious or depraved, he then speaks another language, and expresses sentiments, in which, he expects, all his audience are to concur with him. He must here, therefore, depart from his private and particular situation, and must chuse a point of view, common to him with others. (M 9.6, p. 272)

212  Part II When we read Hume’s fourth dissertation in the light of this passage, it is easy to imagine that he thought the same point applied. ‘It is natural’, he writes, ‘for us to seek a Standard of Taste; a rule, by which the various sentiments of men may be reconciled; at least, a decision afforded, confirming one sentiment, and condemning another’ (ST 6, p. 229). This is a suitably modest ambition for someone who embraces the sceptical principle. If a group of Sceptics all want such a standard, meanwhile, then Hume’s is likely the one on which they will agree. Of course, this easy answer will do nothing to silence the hard-line relativist who doesn’t want a standard, or who refuses to accept that there is one until given some compelling reason to do so. But how much should this trouble the Humean? Drawing on the moral Enquiry again, we might think the appropriate strategy with such obstinate thinkers is simply to ignore them: Those who have denied the reality of moral distinctions, may be ranked among the disingenuous disputants [who really do not believe the opinions they defend]; nor is it conceivable, that any human creature could ever seriously believe, that all characters and actions were alike entitled to the affection and regard of every one. . . . Let a man’s insensibility be ever so great, he must often be touched with the images of RIGHT and WRONG; and let his prejudices be ever so obstinate, he must observe, that others are susceptible of like impressions. The only way, therefore, of converting an antagonist of this kind, is to leave him to himself. For, finding that no body keeps up the controversy with him, it is probable he will, at last, of himself, from mere weariness, come over to the side of common sense and reason. (M 1.2, pp. 169–70) As someone with a Humean temperament myself, I confess I have a good deal of sympathy with this attitude. Since many do not, however, it may be just as well that Hume (and the Humean) can offer a deeper answer to this question. As it happens, this also constitutes a deeper answer to the first question as well, and even helps us to answer two additional questions regarding Hume’s account. But before we can offer this answer, we need to place Hume’s views on art in the context of his philosophy of emotion. It is to this task that I now turn.

12.2.  Minute Observations of Grammar and Criticism The immediate purpose of Hume’s fourth dissertation is, obviously enough, to offer a standard of taste with which to silence the relativist. But Hume also has a wider purpose in this work, one that emerges more clearly when we see it in the context of his philosophy of emotion. Of the

The Laws of Criticism  213 Standard of Taste is a manifesto for an experimental science of criticism, a science that Hume barely began, but which he considered to be both possible and desirable. Work on the science of criticism was first promised in the advertisement to the Treatise: The subjects of the understanding and passions make a compleat chain of reasoning by themselves; and I was willing to take advantage of this natural division, in order to try the taste of the public. If I have the good fortune to meet with success, I shall proceed to the examination of morals, politics, and criticism; which will compleat this Treatise of human nature. (Ad1739 1, p. xii) As is well known, Hume abandoned this project not long afterwards. Nevertheless, a third book on morals was published, as well as the two Enquiries and a number of essays on political subjects. Regarding criticism, however, we have only scraps; though Hume continued to be conscious of the need for more, as we learn from this comment in his essay Of Civil Liberty: Men, in this country, have been so much occupied in the great disputes of Religion, Politics, and Philosophy, that they had no relish for the seemingly minute observations of grammar and criticism. And though this turn of thinking must have considerably improved our sense and our talent of reasoning; it must be confessed, that, even in those sciences above-mentioned, we have not any standard-book, which we can transmit to posterity: And the utmost we have to boast of, are a few essays towards a more just philosophy; which, indeed, promise well, but have not, as yet, reached any degree of perfection. (CL 8, p. 92) Of Civil Liberty was first published in 1741, 16 years before Of Tragedy and Of the Standard of Taste appeared. Which essays Hume had in mind as moving ‘towards a more just philosophy’ I do not know, but presumably Addison and Shaftesbury (to whom he had referred four paragraphs earlier) cannot have been far from his mind.1 In understanding Hume’s conception of the science of criticism, it is important to appreciate how closely related it is to the science of the passions. One of the main purposes of art, for Hume, is to excite our passions; thus those familiar with the laws governing the interplay of the passions will be well placed to create good art (or to theorize about the merits of existing art). Consider Hume’s praise of his friend John Home’s tragedy Douglas, given in the dedication of the Four Dissertations:2

214  Part II [T]he unfeigned tears which flowed from every eye, in the numerous representations which were made of it on this theatre; the unparalleled command, which you appeared to have over every affection of the human breast: These are incontestible proofs, that you possess the true theatric genius of Shakespear and Otway, refined from the unhappy barbarism of the one, and licentiousness of the other. (Ad1757 5) Thus, though Hume never pursued the science of criticism fully or in detail, his work on the passions may be taken to provide much of its foundation. The dissertation Of Tragedy clearly lies in the overlap between these two sciences. The view that command of the passions was high among the artist’s skills was by no means uncommon in the eighteenth century. Alexander Gerard, for example, in his Edinburgh Society prize-winning Essay on Taste, states that ‘[a] very great part of the merit of most works of genius arises from their fitness to agitate the heart with a variety of passions’ (Gerard 1759, p. 87). Furthermore, in a passage that strikes an obvious note of agreement with Hume’s essay, Gerard writes: Genuine criticism . . . investigates those qualities in its objects which, from the invariable principles of human nature, must always please or displease; describes and distinguishes the sentiments which they in fact produce; and impartially regulates its most general conclusions according to real phaenomena. (1759, p. 181) Or consider also this passage from chapter  2 of Kames’s Elements of Criticism (the second longest chapter in the work by a long way, and devoted exclusively to the passions): The design  .  .  . of this chapter is to delineate that connection [between passions and the fine arts], with the view chiefly to ascertain what power the fine arts have to raise emotions and passions. To those who would excel in the fine arts, that branch of knowledge is indispensable; for without it the critic, as well as the undertaker, ignorant of any rule, have nothing left but to abandon themselves to chance. Destitute of that branch of knowledge, in vain will either pretend to foretell what effect his work will have upon the heart. (Kames 1762, vol. 1, p. 32) It should thus come as no surprise that Hume first published Of the Standard of Taste alongside his Dissertation on the Passions. This juxtaposition would not have seemed remotely unusual to his readers at the time.

The Laws of Criticism  215 Another thing to note about Hume’s conception of the science of criticism is that it primarily concerns, not art in general, but only those arts involving words, such as poetry and theatre. (Notice how, in the quotation given earlier from Of Civil Liberty, criticism and grammar go hand in hand.) It is not, I am sure, that Hume explicitly discounted music and the visual arts from its remit—certainly he never did so in print, and I  doubt that he did in his own mind either—but more that he simply showed no interest in them, and tended to ignore them or unconsciously set them aside. This focus on poetry adds weight to the claim that understanding the passions was crucial to Hume’s science of criticism, especially when we consider that, for Hume, ‘[t]he object . . . of poetry [is] to please by means of the passions and the imagination’ (ST 22, p. 240). Thus the hunt was on for the ‘rules of composition’, ‘rules of art’, ‘rules of criticism’, or ‘rules of beauty’ (ST  9, p.  231; ST  10, p.  232; ST 16; p. 235): general causal laws governing what features give rise to which passions, and in particular which properties give rise to the positive sentimental responses, thus rendering the works that possess these features good or beautiful.3 Hume does not venture any of these rules in Of the Standard of Taste, but he does not doubt their existence. Our question now is: how do these rules of art relate to the standard of taste, and to Hume’s attempt to refute the relativist?

12.3.  The General Rules of Art There is no a priori guarantee that a science of criticism, of the sort that Hume envisaged, will be possible. A science of this sort is concerned with uncovering general principles, but what if the relevant phenomena are simply too diverse to admit of any substantial generalizations? What if individual tastes are simply too different to be accounted for by any general rules of art or criticism? This is the thesis of (let us say) particularism, and Hume’s science of criticism is premised on its falsehood. Particularism about art these days typically means a particularism about properties, the view that no general rules exist conferring value on properties of works of art: the most we can say is that a property makes this particular work of art valuable, not that it renders valuable, to a degree, any work possessing that feature. This is not the view that I have in mind here (though I take it that Hume was implicitly opposed to this view as well).4 The view that I have in mind here is rather a particularism about people; particularists of this sort may allow that a certain property is always valuable in any work of art relative to a particular person or group of people. What they deny is that there are any general rules of this nature governing all people (or all groups of people). Some people like this sort of thing, and other people like that sort of thing, and that is all we can say at the general level.

216  Part II It will be apparent, I trust, how close the relationship is between particularism (in my sense) and relativism. But the two are not the same. Crucially, particularism is a purely descriptive doctrine. It claims that different people are, in matters of taste, too different to allow for any general science of taste or criticism that makes substantial claims about all of them. Perhaps we can do a general science of these people’s taste, or those people’s taste; but a general science of human taste will not be possible. The relativist, by contrast, makes an evaluative claim: that different people have different tastes, and that we cannot legitimately criticize others for having a taste that is different from our own, or pronounce any one taste superior to any other. Hume is firm in his conviction of the possibility of uncovering general rules of art. He acknowledges, of course, the great variety of taste in the world, but he urges that these differences are not simply brute, inexplicable facts, as the particularist maintains. Rather, there are general principles of human nature governing everyone’s sentimental responses, with the divergent sentiments that we actually experience being the result of causal interference from other principles which operate in different people to different degrees: But though all the general rules of art are founded only on experience and on the observation of the common sentiments of human nature, we must not imagine, that, on every occasion, the feelings of men will be conformable to these rules. Those finer emotions of the mind are of a very tender and delicate nature, and require the concurrence of many favourable circumstances to make them play with facility and exactness, according to their general and established principles. The least exterior hindrance to such small springs, or the least internal disorder, disturbs their motion, and confounds the operation of the whole machine. (ST 10, p. 232) The organs of internal sensation are seldom so perfect as to allow the general principles their full play, and produce a feeling correspondent to those principles. They either labour under some defect, or are vitiated by some disorder; and by that means, excite a sentiment, which may be pronounced erroneous. (ST 23, p. 241) As I said in §12.1, Hume holds that the defining characteristics of his true judges are virtues, i.e. qualities that we all agree are valuable. They are moreover relevant virtues in the appreciation of art. I have given some motivation for these claims already, but we are now in a position to see Hume’s main justification for them, or rather the general backdrop of his particular justifications. The central idea is that these virtues enable us to

The Laws of Criticism  217 remove the causal ‘noise’ that prevents us from experiencing the ‘catholic and universal beauty’ (ST 10, p. 233), from experiencing the sentiments that arise in accordance with the general principles of taste that (Hume believes) lie in each and every one of us. We must of course ask the very important question of why Hume thinks he is entitled to assert that there are these general principles in all of us, or on what basis he concludes that a general experimental science of criticism is indeed possible. This question is all the more urgent given his failure to pursue this science in very much detail (and indeed his acknowledgement, in Of Civil Liberty, that he and his predecessors had yet to take it very far). I will return to this worry presently. In the meantime, suffice it to say that if such a science is possible, then it does seem legitimate to erect a standard of taste on the back of these rules. There can be no suspicion of arbitrariness in choosing to fix on the sentiments of the true judges rather than those of any other group, since there will be an important sense in which these are the sentiments of all of us. It is simply that in many cases various factors prevent us from feeling them when and where we otherwise would.

12.4.  An Entire or Considerable Uniformity of Sentiment The success of Hume’s response to the relativist thus depends upon the prospects for a general science of criticism. To put it another way, it depends upon the empirical claim that there is or will be a sufficient uniformity in the sentimental responses of the true judges: ‘If, in the sound state of the organ, there be an entire or a considerable uniformity of sentiment among men, we may thence derive an idea of the perfect beauty’ (ST 12, p. 234). If not, however, it would seem the Humean sceptic is forced to acknowledge particularism, to abandon the science of criticism, and to accept the relativist’s conclusion. This, I take it, is the point of Hume’s claim that the standard of taste consists of the joint verdict of the true judges, rather than the sentiments of any one such judge in isolation. The standard is premised on the extent to which they all agree. Hume does not insist on complete uniformity of sentiment among his true judges, and nor need he in order to establish his science of criticism; as in the passage just quoted, a ‘considerable’ uniformity would suffice. Towards the end of the dissertation, moreover, he explicitly denies that there is complete uniformity, thereby conceding some ground to the relativist: But notwithstanding all our endeavours to fix a standard of taste, and reconcile the discordant apprehensions of men, there still remain two sources of variation, which are not sufficient indeed to confound all the boundaries of beauty and deformity, but will often serve to produce a difference in the degrees of our approbation or blame.

218  Part II The one is the different humours of particular men; the other, the particular manners and opinions of our age and country. The general principles of taste are uniform in human nature: Where men vary in their judgments, some defect or perversion in the faculties may commonly be remarked; proceeding either from prejudice, from want of practice, or want of delicacy; and there is just reason for approving one taste, and condemning another. But where there is such a diversity in the internal frame or external situation as is entirely blameless on both sides, and leaves no room to give one the preference above the other; in that case a certain degree of diversity in judgment is unavoidable, and we seek in vain for a standard, by which we can reconcile the contrary sentiments. (ST 28, pp. 243–4) Hume proceeds to offer examples of such blameless differences. ‘A young man, whose passions are warm, will be more sensibly touched with amorous and tender images, than a man more advanced in years’ (ST  29, p. 244); ‘One person is more pleased with the sublime; another with the tender; a third with raillery’ (ST 30, p. 244); and so on. To this extent, he allows, the relativist is perfectly correct: ‘Such preferences are innocent and unavoidable, and can never reasonably be the object of dispute, because there is no standard, by which they can be decided’ (ibid.). Still, according to Hume this diversity is not so great as to undermine the whole project. But now we come to what is perhaps the most serious difficulty with Hume’s response to relativism. The difficulty has both an internal and an external facet. On the external side, the worry is simply that there may be more variation in our sentimental responses to art than Hume supposed. In other words, particularism may in fact be true. But even if particularism is false, the internal worry is that Hume himself failed to provide an adequate justification for this foundational tenet of his position. Thus Matthew Kieran finds it ‘hard to see why we should grant the assumption that ideal critics will, even for the most part, converge’ (Kieran 2005, p. 228). For Malcolm Budd, similarly, the ‘principal weakness’ of Hume’s account is his ‘blithe optimism about the uniformity of response of his true judges’ (Budd 1995, p. 19). I will not attempt to defend Hume against the external aspect of this worry. The truth or falsehood of particularism is an empirical matter on which I am not qualified to judge, and which would in any case take us far beyond the scope of this study. The internal objection, by contrast, is comfortably within my present remit, and I will do my best to plead on Hume’s behalf. Before I do so, however, I should like to concede up front that I think there is much truth to the objection. The situation with regard to the rules of art is similar to that concerning the conversion principle. As I admitted at the end of Chapter 10 (again expressing some

The Laws of Criticism  219 sympathy with Budd’s criticisms), Hume’s confidence in that principle seems to outrun his evidence for it, which consists of phenomena only loosely analogous to each other, and not obviously all instances of the very same principle. We can certainly excuse Hume for this: he was in the vanguard of those trying to apply the burgeoning scientific method to the study of the human mind. This group was filled with optimism, but inevitably lacking in scientific maturity. But if we are able to attain such maturity now, it is only thanks to their pioneering work. (And who can say how naïve we will appear to psychologists 300 years from now?) Nevertheless, the criticism still stands. This being said, it would be unfair to say that Hume had nothing to offer in support of his optimism. Of the Standard of Taste does present at least one argument for the existence of rules of composition, based on the fact that some works of art—he chooses Homer as his example—have been appreciated throughout the ages: The same Homer, who pleased at Athens and Rome two thousand years ago, is still admired at Paris and at London. All the changes of climate, government, religion, and language, have not been able to obscure his glory. . . . It appears then, that, amidst all the variety and caprice of taste, there are certain general principles of approbation or blame, whose influence a careful eye may trace in all operations of the mind. (ST 11–12, p. 233) In a recent article, Shelley has defended Hume against Budd and Kieran by appealing to this same argument (Shelley 2013, pp.  148–9). In one respect, which I will come to in a moment, I feel Shelley overstates the case. In another respect, however, he says less than he might have. For although this may be the only evidence presented in Of the Standard of Taste itself, we do well to point out the surrounding evidence from the other three dissertations. Taken together, the Four Dissertations in fact provide a whole host of arguments against particularism, and moreover arguments of precisely the right sort: Hume seeks to establish the viability of his interlocking sciences of the passions and of criticism by directly pursuing the former, by attempting to construct general principles out of the observable phenomena. The naïvety of some of these principles, I have said, is certainly a weakness, but an entirely forgivable one. And it is certainly not the case that Hume had no arguments at all. We should however admit, as Shelley does not, that the argument from Of the Standard of Taste stands in need of some strengthening. That there are in human nature certain general principles of taste is certainly one potential explanation for the fact that some works have stood the test of time, perhaps even the most natural one. But it is by no means the only one, and for all Hume says there might be other explanations

220  Part II available that are consistent with particularism. Perhaps the common appreciation of Homer, however widespread, is still peculiar to a certain set of people, those brought up with a prejudice in favour of the Hellenic tradition. Paris and London are one thing, but how does Homer fare in Tokyo or Pyongyang? It is safe to say that Hume had no idea. More interestingly, there is the possibility that general principles concerning the passions might themselves offer some support for particularism specifically regarding our responses to art. For example, it is a well-­established principle that we tend to prefer things we are familiar with, and consequently that we come to like things more, the more we are exposed to them (see e.g. Zajonc 2001). For all Hume says, the extended success that Homer enjoys might owe more to this aspect of our nature than to the intrinsic qualities of Homer’s writing. And there would of course be a reinforcing loop here, since we will naturally expose our children and students more to the artists we ourselves most value. To be clear, I am not saying this explanation of Homer’s continued appeal is right. The empirical question is not one I am qualified to answer, nor would it be appropriate to try and do so in the present context.5 The point is that the success of Hume’s account ultimately depends on how these things pan out, and that Hume himself—albeit perfectly u ­ nderstandably—­offered only the very beginnings of a solid empirical case.

12.5.  I Find Your Questions Very Perplexing Seeing Hume’s standard of taste in the context of his science of the passions, then, enables us not only to offer a deeper answer to the question of why his true judges set the standard for all of us, but also to justify (up to a point) the crucial claim that the sentiments of these judges will coincide to a significant extent. By way of conclusion, I shall now argue that it also helps us to respond to a challenge made recently by Jerrold Levinson, which he provocatively labels ‘the real problem’ with Hume’s standard of taste (Levinson 2002). The difficulty Levinson highlights is not with saying why the sentiments of the true judges are right, but rather why—even assuming that they are—this fact should be any concern of ours: Why are the works enjoyed and preferred by ideal critics characterized as Hume characterizes them ones that I should, all things being equal, aesthetically pursue? Why not, say, the objects enjoyed and preferred by critics—call them izeal critics—who are introverted, zany, endomorphic, arrogant, and left-handed? . . . To put the question in its most egoistic form, why think you will be aesthetically better off if you become ideal, rather than izeal? . . . Why should one be moved by the fact that such and such things are approved [of] or preferred by ideal critics, if one is not [an ideal critic] oneself? (2002, pp. 229–30)

The Laws of Criticism  221 A quick response to this worry on Hume’s behalf would be to say that this was not a question he himself asked or attempted to answer. Hume’s question, clearly enough, was whether we could find some set of judges who had the right sentiments; it is a further question why anyone should want to share their tastes. Levinson himself is unmoved by this response. It was already raised by Wieand (2003), and in a subsequent paper Levinson emphasized that he did not mean to raise a problem specifically for Hume, but merely a difficulty that strikes people of a certain turn of mind upon encountering Hume’s essay (Levinson 2003). Not only is Levinson’s question one that Hume himself didn’t raise, however, it is also one that a Humean may be inclined to think has no answer, at least not one that applies to all human beings in general. Perhaps there are some reasons applicable to some of us, other reasons applicable to others, and few or no reasons at all for those who quite legitimately have better things to do: Do you come to a philosopher as to a cunning man, to learn something by magic or witchcraft, beyond what can be known by common prudence and discretion?—Yes; we come to a philosopher to be instructed, how we shall chuse our ends, more than the means for attaining these ends: We want to know what desire we shall gratify, what passion we shall comply with, what appetite we shall indulge. . . . I am sorry then, I have pretended to be a philosopher: For I find your questions very perplexing. (Sc 6–7, p. 161) However, to satisfy Levinson (and to continue this passage from The Sceptic in paraphrase), I shall deliver my opinion upon the matter, and shall only desire you to esteem it of as little consequence as I do myself. In the first place, it should be stressed that the question, for Hume, is not so much why we should spend time with the works of art true judges enjoy (even though we find them dull or impenetrable ourselves), but rather why we should cultivate these judges’ defining characteristics (whereupon we will find ourselves appreciating what they enjoy, and Levinson’s question will vanish). And in answer to this, it is perhaps more readily apparent what one should say. Good sense, first of all, is obviously a useful quality to possess, for all sorts of reasons (by no means limited to the appreciation of art). And Hume makes a point of saying, second, that delicacy of taste is agreeable: A very delicate palate, on many occasions, may be a great inconvenience both to a man himself and to his friends: But a delicate taste of wit or beauty must always be a desirable quality; because it is the source of all the finest and most innocent enjoyments, of which human nature is susceptible. (ST 17, p. 236)

222  Part II In Of the Delicacy of Taste and Passion, moreover, Hume argues at length for the agreeableness of this quality (and disagreeableness of the parallel delicacy of passion). Perhaps scepticism is appropriate about the extent to which we should free ourselves from prejudice (at least of the morally blameless and cognitively harmless kind). Hume after all considers some degree of prejudice to be quite acceptable in the appreciation of art: ‘it is almost impossible not to feel a predilection for that which suits our particular turn and disposition. Such preferences are innocent and unavoidable, and can never reasonably be the object of dispute’ (ST 30, p. 244). But in general, the character of the true judge is surely a desirable one, quite independently of that character’s emotional responses to works of art. Levinson insists, however, that these observations do not get to the heart of the matter: It will be remarked immediately that the traits of Humean ideal critics, in contrast to the traits of izeal critics, are inherently desirable and widely admired. But that does not in itself show why it will be to your aesthetic benefit to acquire them and to follow up the preferences of perceivers who have them. It will next be recalled that the traits of an ideal critic, unlike those of an izeal critic, are not only desirable or admirable in themselves, but ones that enable their possessors to have superior aesthetic reactions from works with the capacity to afford them. But how do we know that? That is, how do we know that the traits of ideal critics put them in a better position overall to have aesthetic experiences from works of art? What assures us that those traits, and not others, optimize capacity for aesthetic response? (2002, pp. 229–30) The point here, I take it, is that Levinson is looking for reasons to have the character traits of true judges that are concerned exclusively with the appreciation of art. To put it another way, while it may be agreed that we all need good sense to help get us to the theatre, Levinson is asking why we should want it for the duration of the play. The quotation just given also highlights that there are in fact two aspects to this problem, though Levinson does not always clearly distinguish them. The first is why it is in my interests to become like the true judge. But the second is what evidence I have for thinking this is in my interests. Levinson’s goal is not merely critical, and he himself offers a solution to his problem. Structurally, he gives the only possible answer to the first question: he claims that the works true judges enjoy afford experiences that are ‘ultimately more worth having’ (2002, p. 233). I will return to this. His answer to the second question, meanwhile, draws on a canon of masterworks that have stood the test of time. We have reason to

The Laws of Criticism  223 believe, he says, that these works are aesthetically rewarding to a particularly high degree; and the fact that true judges understand and appreciate these works gives us reason to believe that the other works they appreciate will be similarly rewarding (2002, pp. 232–4). He does not claim this solution is Hume’s, but does claim it is at least Humean,6 and commends Mary Mothersill for having emphasized the ‘subtext’ of Hume’s essay that involves reference to such a canon of masterworks (Mothershill 1989; see Levinson 2002, pp. 231–2). As I read Hume, however, the point of this ‘subtext’ is entirely missed by Levinson. Levinson sees the appreciation of masterworks as a key identifying characteristic of the true judge, without which we have no reason to think their taste is one worth cultivating. As he says in response to Wieand: Wieand holds to the traditional reading of Hume, according to which true judges are identified by the five famous marks of such judges, thus leaving no role in their identification for works of art that have passed the test of time. But Wieand does not see that without an appeal to something beyond the five marks, those marks can seem rather arbitrary and unsupported—leaving aside worries about circularity and counterproductiveness—as almost all commentators agree. (Levinson 2003, p. 399) Now in fact I agree that works of art that pass the test of time do help to support the characteristics of the true judge. But the way in which they do so is indirect, and very different from the story Levinson tells. What grounds the characteristics of the true judge—as I have explained earlier—­is the science of criticism, and the rules of art that are uncovered in the pursuit of that science; and what encourages us to think that such a science will be possible is (among other things) the fact that some works have stood the test of time, and have been admired across different ages and cultures. If the account I have pieced together in this chapter is right, then Levinson’s solution to his problem is certainly not Hume’s, and has little claim to being considered Humean either. As to the first part of Levinson’s solution, namely that the aesthetic experiences true judges enjoy are ‘ultimately more worth having’ than others (2002, p. 233), it is here that the scepticism I voiced earlier comes into play. They are certainly more worth having, trivially, if what you care most about is having the right sentiments. Are they more worth having if what you care most about is having the most pleasant experiences? Perhaps (as Levinson himself seems to maintain); though I certainly do not see it as a requirement on a standard of taste that it should entail that this is so. And what if you care more about maintaining your high opinion of your friend’s frankly middling compositions? In that case the

224  Part II balance of reasons for you will presumably be in favour of retaining your partiality. The reason the ‘real’ problem arises for Levinson is that he holds (what he terms) a capacity theory of aesthetic value, the view that ‘the primary artistic value of a work of art, what Hume calls its beauty or excellence, is plausibly understood in terms of the capacity or potential of the work, in virtue of its form and content, to afford appreciative experiences worth having’ (2002, p.  233). It is a consequence of this view that the best works of art are those that are ‘capable of affording better, or ultimately preferable, aesthetic experiences’ (2002, p. 231). Thus we are faced with the problem of showing that the works recommended by true judges are indeed the ones that afford us such experiences. But Hume held a quite different view of aesthetic value, and this at bottom is why Levinson’s article misses its mark.7 For Hume, the beauty of a work of art—its ‘real’ beauty (ST 7, p. 230), its ‘catholic and universal’ beauty (ST 10, p.  233)—consists of its capacity to afford, not the best positive experiences, but positive experiences to the best people. Thus the Humean response to Levinson’s problem is to point out that on Hume’s account it does not arise.

Notes 1 Hugh Blair, in his Lectures on Rhetoric and Belles Lettres (first published in 1783, but given to students at Edinburgh at least 15  years earlier), echoes Hume’s lament in Of Civil Liberty that ‘[w]e are far from having yet attained to any system concerning this subject’ (volume 1, p. 52), but mentions Addison’s Essay on the Pleasures of the Imagination (Spectator, volume 6) as the first attempt at systematization (ibid.). The lectures also refer to Hume’s Of the Standard of Taste (volume 1, p. 20), and Blair’s response to relativism is very much in the same spirit as Hume’s. 2 Evidence, perhaps, that Hume was guilty of partiality, and hence not himself a true judge in this instance. 3 Sometimes the prompting of negative sentiments, for Hume, can be valuable in a work (as in the case of tragedy), but only because they increase the predominant pleasurable sentiments. 4 See the exchange between Dickie (2003), a generalist, and Shelley (2002, 2004), a particularist (in the modern sense). Shelley argues, convincingly to my mind, that Hume was a generalist in the sense that he himself rejects. 5 Some intriguing research conducted recently by Aaron Meskin, Mark Phelan, Margaret Moore, and Matthew Kieran (Meskin et al. 2013) suggests that mere exposure to (what true judges deem to be) bad art in fact tends to make us like it less, indicating that this explanation of Homer’s enduring fame is not right. So much the better for the modern-day Humean; but the argument, no doubt, does not end here. 6 ‘[T]he materials for a solution to [the] problem are at hand in Hume’s essay, even if not deployed by him to that end’ (Levinson 2003, p. 399). 7 I am not concerned here with its merits as an independent development or defence of the capacity theory. For a criticism of it on this score, which also

The Laws of Criticism  225 defends an interpretation of Hume very much in line with the one I am here proposing, see Shelley (2011).

Bibliography Blair, Hugh (1783). Lectures on Rhetoric and Belles Lettres, in 3 volumes. Dublin. Budd, Malcolm (1995). Values of Art: Pictures, Poetry, and Music. London: The Penguin Press. Dickie, George (2003). ‘James Shelley on Critical Principles’. The British Journal of Aesthetics 43(1), pp. 57–64. Gerard, Alexander (1759). An Essay on Taste. London. Kames, Lord [Henry Home] (1762). Elements of Criticism, in three volumes. Edinburgh. Quotations and page references from Peter Jones, ed. (2005). Elements of Criticism, in two volumes, by Henry Home, Lord Kames. Ed. by Peter Jones. Indianapolis: Liberty Fund. Kieran, Matthew (2005). Revealing Art. London: Routledge. Levinson, Jerrold (2002). ‘Hume’s Standard of Taste: The Real Problem’. The Journal of Aesthetics and Art Criticism 60(3), pp. 227–38. ——— (2003). ‘The Real Problem Sustained: Reply to Wieand’. The Journal of Aesthetics and Art Criticism 61(4), pp. 398–9. Meskin, Aaron et al. (2013). ‘Mere Exposure to Bad Art’. The British Journal of Aesthetics 53(2), pp. 139–64. Mothershill, Mary (1989). ‘Hume and the Paradox of Taste’. In Aesthetics: A  Critical Anthology. Eds. by George Dickie, Richard Sclafani and Ronald Roblin. New York: St Martin’s Press, pp. 269–86. Shelley, James (2002). ‘The Character and Role of Principles in the Evaluation of Art’. The British Journal of Aesthetics 42(1), pp. 37–51. ——— (2004). ‘Hume’s Principles of Taste: A Reply to Dickie’. The British Journal of Aesthetics 44(1), pp. 84–9. ——— (2011). ‘Hume and the Value of the Beautiful’. The British Journal of Aesthetics 51(2), pp. 213–22. ——— (2013). ‘Hume and the Joint Verdict of True Judges’. The Journal of Aesthetics and Art Criticism 71(2), pp. 145–53. Wieand, Jeffrey (2003). ‘Hume’s Real Problem’. The Journal of Aesthetics and Art Criticism 61(4), pp. 395–8. Zajonc, R. B. (2001). ‘Mere Exposure: A Gateway to the Subliminal’. Current Directions in Psychological Science 10(6), pp. 224–8.

Conclusion

The current scholarly emphasis on the Treatise, at the expense of Hume’s later works, has become so familiar that it now scarcely raises an eyebrow. In truth, however, it ought to be very puzzling. Not only does common sense suggest that Hume’s thought would have got better over time, but we also know that Hume himself deeply regretted publishing the Treatise, and in later life had a clear preference for the Enquiries and Dissertations. In October 1775, most notably, he wrote to William Strahan asking that the following advertisement be prefixed to any remaining copies of volume 2 of his Essays and Treatises (in the event it was prefixed to volume 2 of the posthumous 1777 edition): MOST of the principles, and reasonings, contained in this volume, were published in a work in three volumes, called A Treatise of Human Nature: A work which the Author had projected before he left College, and which he wrote and published not long after. But not finding it successful, he was sensible of his error in going to the press too early, and he cast the whole anew in the following pieces, where some negligences in his former reasoning and more in the expression, are, he hopes, corrected. Yet several writers, who have honoured the Author’s Philosophy with answers, have taken care to direct all their batteries against that juvenile work, which the Author never acknowledged, and have affected to triumph in any advantages, which, they imagined, they had obtained over it: A  practice very contrary to all rules of candour and fair-dealing, and a strong instance of those polemical artifices, which a bigoted zeal thinks itself authorised to employ. Henceforth, the Author desires, that the following Pieces [the two Enquiries, the Dissertation on the Passions, and the Natural History of Religion] may alone be regarded as containing his philosophical sentiments and principles. (Ad1777 1, p. 2) Hume had in mind here the criticisms of him by Thomas Reid (1764) and James Beattie (1770); in the letter he writes that the advertisement

Conclusion  227 ‘is a compleat Answer to Dr Reid and to that bigoted silly Fellow, Beattie’ (HL  2, p.  301). But his dissatisfaction with the Treatise goes back much further. In the published Letter from a Gentleman to his Friend in Edinburgh (1745), for example, written in response to a pamphlet accusing the Treatise of endorsing ‘Universal Scepticism’ (L 14, p. 425), ‘Principles leading to downright Atheism’ (L 15, p. 425), and ‘With sapping the Foundations of Morality’ (L 19, p. 425), Hume admitted, after denying the charges, ‘that the Author had better delayed the publishing of that Book; not on account of any dangerous Principles contained in it, but because on more mature Consideration he might have rendered it much less imperfect by further Corrections and Revisals’ (L 41, p. 431). The same attitude emerges from his private correspondence. In March 1740, even before Book 3 had been published, Hume was already writing thus to Hutcheson: ‘I wait with some impatience for a second Edition principally on Account of Alterations I  intend to make in my Performance’ (HL 1, p. 38). Of course, he never did produce a second edition, but wrote instead the Enquiries and Dissertations. Indeed, in a letter to John Stewart some time later, he wrote that the ‘positive Air’ of the Treatise ‘so much displeases me, that I have not Patience to review it’ (HL  1, p.  187). Once the first Enquiry was in print, moreover, he expressed a clear preference for this work in a letter to Gilbert Elliot: I believe the philosophical Essays contain every thing of consequence relating to the Understanding, which you would meet with in the Treatise; & I give you my Advice against reading the latter. . . . The Philosophical Principles are the same in both: But I was carry’d away by the Heat of Youth & Invention to publish too precipitately. So vast an Undertaking, plan’d before I was one and twenty, & compos’d before twenty five, must necessarily be very defective. I have repented my Haste a hundred, & a hundred times. (HL 1, p. 158) The evidence is unambiguous, and there is no doubt that Hume came to dislike his own Treatise almost immediately after publishing it. While this much is uncontroversial, however, the more difficult question is why Hume was so unhappy with his first performance. There are, essentially, two possible answers. The less interesting one holds that the basis of Hume’s attitude was solely or at least predominantly stylistic. There is certainly some prima facie evidence in support of this view. In the Letter from a Gentleman, Hume admits that he had better delayed the publishing of the Treatise, but not because of its ‘dangerous Principles’. In the 1777 advertisement, he tells us that ‘[m]ost of the principles, and reasonings’ from his later work were already to be found in the Treatise, and while he acknowledges ‘some negligences in his former reasoning’, it appears that he is more unhappy with the ‘expression’. Again, the

228 Conclusion letter to Elliot states, with regard to Book 1 and the first Enquiry, that ‘[t]he Philosophical Principles are the same in both’. And in his autobiographical My Own Life, Hume said he felt that the failure of the Treatise ‘proceeded more from the manner than the matter’ (MOL 8, p. xxxv) of the work. This answer, I take it, is the one that most scholars today favour, a view that goes some way to explaining the general preference for the Treatise over Hume’s later works. But the evidence of Hume’s explicit remarks on the matter is also consistent with a more intriguing possibility: that his dissatisfaction with the Treatise was not merely stylistic, but also grounded in some of its content. Most of the principles are not all of the principles, after all; negligences in reasoning, however reticently acknowledged, are certainly matters of substance; and Hume can hardly have thought that he answered Reid’s and Beattie’s philosophical criticisms merely by improving his writing style.1 Needless to say, it is this latter answer that I favour: as I have argued throughout this book, there are several substantial differences between Treatise Book 2 and the Four Dissertations. I have tended to speak neutrally of differences between Hume’s earlier and later works, but I needn’t be coy. It has no doubt been apparent from very early on that I believe the philosophy of emotion we find in the Four Dissertations to be not only different from that of Treatise Book 2, but also better. It used to be thought that the Treatise was Hume’s principal philosophical work, and that, prompted by its lack of immediate success, he dumbed down his ideas in the later Enquiries so as to reach a wider audience (while throwing in some provocative irreligious arguments for good measure). No one today, I take it, holds this uncharitable interpretation of Hume’s literary and intellectual motives.2 Old habits die hard, however, and the Treatise is still considered by most scholars to be the ‘heavyweight’ of Hume’s philosophical texts. Success for my argument would entail nothing less than the relegation of Book 2 from its current exalted position as the pinnacle of Hume’s thought on the passions, to the more humble—but philosophically and historically more plausible— status of a first draft.

Notes 1 This last point is made by John Nelson (1972), who may also have been the first person (I haven’t done a thorough search) to ask the present question clearly in these terms. For a detailed attempt to explain how Hume’s first Enquiry ‘answers’ Reid’s criticisms of the Treatise, see Millican (2006). 2 The classic refutations of it are Kemp Smith (1941, pp. 526–30) and Mossner (1950).

Bibliography Beattie, James (1770). An Essay on the Nature and Immutability of Truth; in Opposition to Sophistry and Scepticism. Edinburgh.

Conclusion  229 Kemp Smith, Norman (1941). The Philosophy of David Hume. Basingstoke: Palgrave Macmillan. Millican, Peter (2006). ‘Hume’s “Compleat Answer to Dr Reid” ’. Hume Conference, University of Koblenz, Germany. Available at https://davidhume.org/ scholarship/papers/millican Mossner, E. C. (1950). ‘Philosophy and Biography: The Case of David Hume’. The Philosophical Review 59(2), pp. 184–201. Nelson, John O. (1972). ‘Two Main Questions Concerning Hume’s Treatise and Enquiry’. The Philosophical Review 81(3), pp. 333–50. Reid, Thomas (1764). An Inquiry into the Human Mind, on the Principles of Common Sense. Edinburgh.

Appendix 1 The Meaning of All the Terms

The following tables show my reconstruction of Hume’s taxonomy of the passions in the Treatise, and definitions of his terms for particular passions. See Chapter  1 and Chapter  3 for the relevant discussion. To be clear, I offer no corresponding presentation of the view I attribute to Hume after the Treatise. This is not because I believe his view remained the same, but because I believe he abandoned the taxonomic project altogether. See Chapter 4 and Chapter 7. Impressions Original Impressions/ Impressions of Sensation

Secondary Impressions/Impressions of Reflection Calm Emotions (i.e. the moral and critical sentiments)

Violent Emotions (i.e. passions) Direct Passions (examples follow)

Indirect Passions (examples follow)

Token passions, both direct and indirect, can also be relatively ‘calm’ or ‘violent’

Direct Passions arise from pleasure and pain without any other principles Desire Aversion Joy Grief/Sorrow Hope Fear Security Despair

primitive primitive primitive primitive a mixture of joy and sorrow, in which the joy predominates; paradigmatically felt when an imagined good is likely a mixture of joy and sorrow, in which the grief predominates; paradigmatically felt when an imagined evil is likely undefined (perhaps a special case of joy, or a mixture like hope); paradigmatically felt when an imagined evil is unlikely undefined (perhaps a special case of sorrow, or a ­mixture like fear); paradigmatically felt when an imagined good is unlikely

Appendix 1  231 Indirect Passions arise from pleasure and pain together with other principles Pride Humility Ambition

Vanity

Love

Hatred

Esteem/Respect Contempt Benevolence

Anger

Gratitude

a positive feeling about oneself; arises from the double relation of impressions and ideas a negative feeling about oneself; arises from the double relation of impressions and ideas a. the pleasure of considering our power over others; b. the corresponding desire for this pleasure, and aversion to its contrary; arises from a favourable comparison with the ­ person we have power over a.  a synonym for pride; b. a special case of pride, namely the pleasure arising from the love, respect, or praise of others; c. the corresponding desire for this pleasure, and aversion to its contrary; arises from sympathy with the other person’s good opinion a positive feeling about another person, analogous to pride; arises from the double relation of impressions and ideas a negative feeling about another person, analogous to humility; arises from the double relation of impressions and ideas a mixture of love with humility a mixture of hatred with pride a. the pleasure arising from the happiness of our friends (i.e. those we love); b. the corresponding desire for the happiness of our friends, and aversion to their misery; arises from love, through an original instinct a. the pain arising from the happiness of our ­enemies (i.e. those we hate); b. the corresponding desire for the misery of our enemies, and aversion to their happiness; arises from hatred, through an original instinct a special case of benevolence, arising in response to services rendered (Continued)

232  Appendix 1 Indirect Passions arise from pleasure and pain together with other principles Resentment Pity/ Compassion Generosity Malice

Envy

a special case of anger, arising in response to injuries received a.  a sympathetically communicated sorrow; b. the corresponding desire for the happiness of another, and aversion to their misery; arises from sympathy with another’s suffering a sympathetically communicated joy; arises from sympathy with another’s joy a.  a delight in the sufferings of another; b. the corresponding desire for the misery of another, and aversion to their happiness; arises from a favourable comparison of another’s situation with our own a pain in the happiness of another; arises from an unfavourable comparison of ­another’s situation with our own

Miscellaneous

Avarice

The Amorous Passion

Curiosity

the (excessive) delight in or desire for wealth; often mentioned, but never explicitly defined or discussed, in the Treatise or elsewhere (the early essay Of Avarice is moralizing rather than descriptive or psychological) a mixture of the love of beauty, lust (the bodily ­appetite for generation), and benevolence; perhaps technically an indirect passion, this is nevertheless something of an odd one out, and includes an original impression (the bodily appetite) as one of its ingredients the love of truth; explicitly stated to lie outside the general ­classificatory scheme (T 2.3.10.2, p. 448)

Appendix 2 Comparison of Ideas

The following table shows the sections and paragraphs of the Dissertation on the Passions alongside the paragraphs from Treatise Book 2 from which they were presumably derived. In producing this table, all paragraphs of the Dissertation were assumed new until proven otherwise; if anyone notices a ‘new’ paragraph that does, after all, have a precursor in the Treatise, I would gladly be corrected. ‘Copies’ are sometimes shorter than their originals, and often contain a handful of stylistic changes. A ‘cf.’ entry means material too different to count as a copy, but with clear echoes of the earlier work. Typically these are summaries or paraphrases of Treatise material.1 Dissertation

Treatise Book 2

SECT. 1

≈ T 2.3.9

§1 §2

§3

§4 §5 §6

§7 §8

P 1.1 P 1.2 P 1.3 P 1.4 P 1.5 P 1.6 P 1.7 P 1.8 P 1.9 P 1.10 P 1.11 P 1.12 P 1.13 P 1.14 P 1.15 P 1.16 P 1.17 P 1.18 P 1.19 P 1.20

T 2.3.9

T 2.3.9

T 2.3.9 T 2.3.9 T 2.3.9

T 2.3.9 T 2.3.9

new new new copy of T 2.3.9.5 copy of T 2.3.9.6 copy of T 2.3.9.7 copy of T 2.3.9.9 copy of T 2.3.9.10 copy of T 2.3.9.11 copy of T 2.3.9.12 copy of T 2.3.9.18 copy of T 2.3.9.19 copy of T 2.3.9.20 copy of T 2.3.9.21 copy of T 2.3.9.22 copy of T 2.3.9.23 copy of T 2.3.9.24 copy of T 2.3.9.25 copy of T 2.3.9.27–28 copy of T 2.3.9.29

(Continued)

234  Appendix 2 Dissertation §9

Treatise Book 2 P 1.21 P 1.22 P 1.23 P 1.24 P 1.25

SECT. 2 §1 §2

§3

§4 §5 §6 §7

§8

§9 §10

§11

T 2.3.9

copy of T 2.3.9.14 copy of T 2.3.9.15 copy of T 2.3.9.16 copy of T 2.3.9.17 new

≈ T 2.1 P 2.1 P 2.2 P 2.3 P 2.4 P 2.5 P 2.6 P 2.7 P 2.8 P 2.9 P 2.10 P 2.11 P 2.12 P 2.13 P 2.14 P 2.15 P 2.16 P 2.17 P 2.18 P 2.19 P 2.20 P 2.21 P 2.22 P 2.23 P 2.24 P 2.25 P 2.26 P 2.29 P 2.30 P 2.31 P 2.32 P 2.33 P 2.34 P 2.35 P 2.36 P 2.37 P 2.38 P 2.39 P 2.40 P 2.41

T 2.1.2, T 2.2.1

T 2.1.4

T 2.1.5 T 2.1.5 T 2.1.7 T 2.1.8

T 2.1.9

T 2.1.10 T 2.1.11

T 2.1.6, T 2.1.8

new new cf. T 2.1.2 & T 2.2.1 new; but cf. T 2.1.2 & T 2.2.1 cf. T 2.1.2 & T 2.2.1 copy of T 2.1.4.2 copy of T 2.1.4.3 copy of T 2.1.4.4 copy of T 2.1.4.5 cf. T 2.1.5 cf. T 2.1.5 cf. T 2.1.5, T 2.1.7.1 cf. T 2.1.5, T 2.1.7.1 copy of T 2.1.7.2+5 copy of T 2.1.7.6 copy of T 2.1.7.7 copy of T 2.1.8.1 cf. T 2.1.8.2–3 copy of T 2.1.8.5 copy of T 2.1.8.6 copy of T 2.1.9.1 copy of T 2.1.9.6 copy of T 2.1.9.7 copy of T 2.1.9.8 copy of T 2.1.9.9 copy of T 2.1.9.10 copy of T 2.1.9.3 cf. T 2.1.10.1 copy of T 2.1.10.2 cf. T 2.1.10.3 new new copy of T 2.1.11.10 copy of T 2.1.11.11 new new copy of T 2.1.11.12 new copy of T 2.1.6.6

(Continued)

Appendix 2  235 Dissertation

Treatise Book 2

SECT. 2

≈ T 2.1 copy of T 2.1.6.7 copy of T 2.1.6.4 copy of T 2.1.8.8 copy of T 2.1.8.9 copy of T 2.1.6.8 copy of T 2.1.6.9

P 2.42 P 2.43 P 2.44 P 2.45 P 2.46 P 2.47 SECT. 3 §1 §2 §3 §4 §5 §6 §7

≈ T 2.2 P 3.1 P 3.2 P 3.3 P 3.4 P 3.5 P 3.6

T 2.1.1, T 2.1.2

P 3.7 P 3.8 P 3.9 P 3.10 P 3.11 P 3.12 P 3.13

T 2.2.7, T 2.2.8

T 2.2.4 T 2.2.6

T 2.2.9

T 2.2.10, T 2.2.11

P 3.14 SECT. 4 §1 §2 §3 §4 §5

§6 §7

cf. T 2.2.11 ≈ T 2.2.2, T 2.2.8

P 4.1 P 4.2 P 4.3 P 4.4 P 4.5 P 4.6 P 4.7 P 4.8 P 4.9 P 4.10 P 4.11 P 4.12

T 2.2.2 T 2.2.2 T 2.2.2 T 2.2.8

T 2.2.8

P 4.13 SECT. 5 §1 §2 §3 §4

cf. T 2.1.1–2 cf. T 2.1.1–2 cf. T 2.2.4.2 cf. T 2.2.4.2 cf. T 2.2.4.3 cf. T 2.2.6 esp. T 2.2.6.3 cf. T 2.2.7 cf. T 2.2.8 cf. T 2.2.9 cf. T 2.2.9 new new cf. T 2.2.10

new cf. T 2.2.2.12 cf. T 2.2.2.13 cf. T 2.2.2.14–16 cf. T 2.2.2.18 copy of T 2.2.8.13 copy of T 2.2.8.15 copy of T 2.2.8.16 copy of T 2.2.8.17 copy of T 2.2.8.18 copy of T 2.2.8.19 new; cf. T 2.3.6.1 (notice the contrast) copy of T 2.2.2.8

≈ T 2.3.3 P 5.1 P 5.2 P 5.3 P 5.4

T 2.3.3 T 2.3.3 T 2.3.3 T 2.3.3

cf. T 2.3.3.1–9 cf. T 2.3.3.1–9 cf. T 2.3.3.1–9 copy of T 2.3.3.10

(Continued)

236  Appendix 2 Dissertation

Treatise Book 2

SECT. 6

≈ T 2.3.4–8

§1

§2 §3 §4 §5 §6 §7 §8 §9

P 6.1 P 6.2 P 6.3 P 6.4 P 6.5 P 6.6 P 6.7 P 6.8 P 6.9 P 6.10 P 6.11 P 6.12 P 6.13 P 6.14 P 6.15 P 6.16 P 6.17 P 6.18

T 2.3.4

T 2.3.4 T 2.3.4 T 2.3.4 T 2.3.4 T 2.3.4 T 2.3.4 T 2.3.5 T 2.3.6–8

new copy of T 2.3.4.2 copy of T 2.3.4.3 new copy of T 2.3.4.4 copy of T 2.3.4.5 copy of T 2.3.4.6 copy of T 2.3.4.7 copy of T 2.3.4.8 copy of T 2.3.4.9 copy of T 2.3.4.10 copy of T 2.3.5.2 copy of T 2.3.6.1 copy of T 2.3.6.5 copy of T 2.3.6.6 copy of T 2.3.6.7 copy of T 2.3.6.8 cf. T 2.3.7–8

Afterword P 6.19

new

Note 1 This table was first printed in Hume Studies 35(1&2), 2009, pp. 202–6. It is reproduced here with the kind permission of the editors.

Index

abstract ideas 135 – 6 Abstract of the Treatise, the 130 Addison, Joseph 109n9, 213, 224n1 ambition 11, 67n4; its origin 61, 80 – 1 anger 11, 60, 63 – 5, 159; its classification 25n6, 73; its object 131; its origin 63 – 5, 80 animal spirits 26 – 7n17, 179 – 80 argument from common consent 124, 127n10 association 11 – 12, 20, 55, 57, 60, 67n2, 130, 137, 140 – 2, 180; applied to comparison 12, 60; applied to sympathy 11, 55, 57, 67n2; see also the double-relation theory associative qualities 11, 49, 54 atomism 132 – 3, 136 Augustine 19 – 20, 145 – 6 authority, principle of 56, 61 aversion: its classification 10; its object 43; its origin 50 – 4, 85 – 7, 130 Beattie, James 226 – 8 belief 10, 25n4, 50; causes of vs. reasons for 94 – 6 benevolence 57, 79, 88n4, 159; contrasted with generosity 62 – 3; in Hutcheson 41 – 3, 152; its classification 11, 25n6, 73; its origin 63 – 7, 80; vs. self-love 45 Berkeley, George 26n17 Blair, Hugh 224n1 bodily appetite 73 Bolingbroke, Viscount [Henry St John] 105 – 7, 109n9, 111 Burnet, Gilbert 42, 47n6, 151

Butler, Joseph 18 – 19, 37, 43 – 6, 46n4, 87 – 8n1, 88 – 9n5; his influence on Hume 3, 24, 67n2, 70 – 2, 75, 77 – 80, 82 – 5, 87; his influence on Kames 172 – 3; model for Cleanthes 22 causation 26n10, 181 – 2 charity 36 – 8, 57, 67 Cicero 21 – 2, 26n14, 26n15, 173, 184 – 5 Clarke, John 30 Clarke, Samuel 18, 42, 87 – 8n1, 157, 191; compared with Shaftesbury 35 – 6; correspondence with Butler 43; his moral rationalism 145, 148 – 51, 153, 159; model for Demea 22 Clayton, Robert 30 comparison, principle of 11 – 2, 26 – 7n17, 54, 59, 67n3; in relation to art 197 – 8; in relation to envy and malice 59 – 61, 80 – 1 compassion/pity 35, 61 – 2, 174n2; its classification 10 – 11, 25n6; its origin 11, 56 – 9, 80 contempt 63 – 4; its classification 11 conversion principle 16 – 17, 170, 174n1, 218; its interpretation 176 – 88, 189n12; its theoretical context 162 – 4 copy principle 10, 25n4, 140 Cudworth, Ralph 107, 111, 151 Cumberland, Richard 42 curiosity 15 – 17, 21; its relevance to religion 16 – 17, 101, 103, 106, 113, 119 deism 19, 36, 106, 107, 111 Descartes, René 19, 26n15, 146, 179, 188n7

238 Index design argument, the 19 – 20, 43, 103, 115, 126n4; Hume’s endorsement of 102, 107, 119 desire: calm vs. violent 15, 162; in Butler 44 – 6, 72, 77; in Hume 50 – 5, 72 – 3, 80, 85 – 7, 130; in Hutcheson 43, 58, 82 – 5, 152 – 3; in Locke 31, 84; its classification 9; its relation to sympathy 58 – 9 despair 15, 51, 173 Dialogues concerning Natural Religion 2, 13, 22, 94, 97 – 9, 109n10, 126n4, 208n3 Dissertation on the Passions, A 1 – 2, 4n3, 130, 155 – 6, 226; compared with the moral Enquiry 81 – 2, 114 – 15, 152; compared with the Treatise 2, 11 – 12, 14 – 18, 24n1, 80, 85 – 7, 113, 129, 138 – 42, 156, 157 – 60, 178; its relation to Of the Standard of Taste 214; its relation to Of Tragedy 162 – 4; its relation to the Natural History 104 – 5, 115 double-relation passions, the (pride, humility, love, and hatred) 10 – 11, 25n5, 50, 198; their cause and object distinct 133 – 4; their definition 139; their indefinability 134 – 5; their origin 130, 140 – 2 double-relation theory, the 11, 129, 140 – 2, 181 – 2 Du Bos, Jean-Baptiste 2 – 3, 164 – 72 egoism 18 – 19, 30 – 1, 82 – 4, 171; Butler’s argument against 44 – 6; Hobbes’s 31, 46n3, 46n4; Hume’s endorsement of 23 – 4, 50 – 9, 62 – 7, 73 – 4, 77 – 8, 81; Hume’s rejection of 75 – 9, 82 – 4, 87; Locke’s 31 – 3; Mandeville’s 23, 38 – 40, 55 Enquiry concerning Human Understanding, An 1, 4n1, 12, 15, 226 – 8, 228n1; compared with the Treatise 13, 142 – 3n2, 77 – 8, 129, 139 – 40; its Epicureanism 20; its relation to Bolingbroke 106; its relation to Hume’s philosophy of religion 94, 98 – 9, 108n4, 121; its relation to the Natural History 96, 118, 125n1, 126n7; its Scepticism 21 Enquiry concerning the Principles of Morals, An 1, 3, 4n1, 13, 115, 160n4, 226 – 8; compared with

the Dissertation 81 – 2; compared with the Treatise 3, 23 – 4, 52, 67n2, 78 – 81, 84, 97, 142 – 3n2; its relation to Hume’s philosophy of art 192, 211 – 12; its relation to Hume’s philosophy of religion 94, 121 – 2; on moral approbation 152, 154, 158 – 60 enthusiasm 20, 95 – 6, 99, 108 – 9n6, 119 envy 10 – 11, 140 – 2; its object 132 – 3; its origin 59 – 62, 80 Epicurean, The 18, 179 Epicureanism 18 – 20, 42, 171 – 2; Hume’s 21 – 3, 55, 60, 67, 73, 100, 108, 174n2; Mandeville’s 37 Essays, Moral and Political 1, 77, 109n9 esteem/respect 11, 56, 61 – 4, 81 fear 15 – 16; its origin 52, 115 – 17, 126n6, 163 – 4; its relevance to religion 16 – 17, 101 – 4, 114 – 15, 117, 119 – 20 Fontenelle, Bernard Le Bovier 2, 164, 166 – 70 Four Dissertations 2 – 4, 25n9, 26n11, 96, 105, 142 – 3n2, 213; the dedication 25n8, 213; their difference from the Treatise 18, 87, 164, 228; their unity 12 – 18, 164, 219 Galen 179, 188n5, 188n6 Galvani, Luigi 179 generosity 10 – 11, 39; contrasted with benevolence 62 – 3; its origin 62, 131 Gerard, Alexander 214 gratitude 99, 107 grief/sorrow 10, 15, 60; its origin 51 – 2, 141; its relation to hope and fear 115 – 16, 126n5; its relation to pity 57 – 9 Harvey, William 188n6, 188n7 hatred 159 – 60; its relation to anger 25n6, 63 – 5, 73; see also the double-relation passions hedonism 19, 30 – 1, 44 – 5; Hume’s endorsement of 3, 46, 49 – 51, 53 – 4, 74, 80, 86; Hume’s rejection of 66, 72, 74 – 5, 86 – 7; Hutcheson’s 43, 58, 82, 84; Locke’s 31 – 2 History of England, The 94, 99

Index  239 Hobbes, Thomas 34, 36, 46n2, 58; his account of religion 100 – 5; his egoism 31, 46n3, 46n4; his Epicureanism 18 – 19, 171 – 2 Home, John 25n8, 213 Homer 204, 207, 219 – 20, 224n5 hope 15 – 16; its origin 50 – 2, 115 – 17, 163 – 4; its relevance to religion 16 – 17, 95 – 6, 99, 104, 117 humanity, principle of 67n2, 154 Hume’s fork 155 – 6, 160n3 humility 55, 132, 136 – 8; see also the double-relation passions Hurd, Richard 105 – 7 Hutcheson, Francis 40 – 5, 51, 227; contrasted with Hume 58, 60, 63 – 4, 66; his anti-rationalism 19, 42 – 3, 77, 151 – 5; his hedonism 43, 45, 58, 82 – 3, 88 – 9n5; his influence on Hume 3, 20 – 1, 23 – 4, 63, 77, 82 – 5, 129 – 30, 155 – 6; his Stoicism 18 joy 15, 134; its origin 51 – 2, 130; its relation to generosity 62; its relation to hope and fear 115 – 16; its relation to malice 61, 63 justice 20, 33, 62 – 3, 76 Kames, Lord [Henry Home] 71, 131, 172 – 3, 174n2, 214 Letter from a Gentleman, A 160n4, 227 liberty and necessity 15, 26n10, 50 Locke, John 18 – 19, 30 – 4, 40, 45, 84, 157; his influence on Hume 3, 9 – 10, 20, 24, 49 – 50, 53 – 4, 66, 116, 126n5 love 159 – 60; its relation to benevolence 25n6, 63 – 6, 73; see also the double-relation passions Lucretius 60, 172 lust 20, 38, 72, 85, 150 – 1 Malebranche, Nicolas 19, 22, 35, 145 – 8, 153 – 7, 179; his influence on Hume 26 – 7n17, 180 malice 10 – 12, 25n6, 43, 63 – 5; its origin 59 – 62, 80 Mandeville, Bernard 37 – 42, 55, 66 – 7, 76; his Epicureanism 18; his influence on Hume 3, 20 – 1, 23 – 4, 53, 56 – 7, 60, 64, 81 – 2 Millar, Andrew 12, 25n9, 26n11

miracles 19 – 20, 96, 118 – 19 monotheism see polytheism motivation 2 – 3; distinct from approbation 41 – 2, 151 – 2, 154 – 5, 158 – 9; Hume’s theory of 12, 49 – 55, 74, 81, 153 – 7, 191 – 2; moral 32 – 3, 55, 62, 76, 121 – 2; see also egoism; hedonism; motivational pluralism motivational pluralism 45, 67n2, 75, 82, 172 Natural History of Religion, The 2, 12 – 13, 126n4, 130; its empirical nature 109n8, 111 – 14, 126 – 7n9; its endorsement of the design argument 97, 119; its philosophical implications 94 – 6, 120 – 5; its relation to Bolingbroke 105 – 7; its relation to the Enquiries 118 – 19, 121 – 2, 125n1; its relation to Hume’s philosophy of emotion 4n5, 16 – 18, 95 – 6, 112 – 15, 125 – 6n2; meaning of its title 100 – 1, 103 – 5 necessary connection 10, 15, 25n4, 50 Of Civil Liberty 213, 215, 217, 224n1 Of the Delicacy of Taste and Passion 206 – 7, 222 Of the Dignity or Meanness of Human Nature 75 – 8 Of the Immortality of the Soul 12, 25n8, 94, 98, 108n5, 125 Of the Standard of Taste 2, 12 – 14, 25n9, 213; its relation to Hume’s philosophy of emotion 193, 198, 214; its response to relativism 192 – 3, 211 – 12, 215 – 20; the threat of circularity 205 – 8; two standards 194 – 6 Of Suicide 12, 25n8, 94, 108n5, 121, 125 Of Superstition and Enthusiasm 20, 94, 95 – 6, 99, 121 Of Tragedy 2, 12 – 13, 213; its criticism of Du Bos and Fontenelle 168 – 70, 172; its positive account 170 – 1, 173 – 4, 176 – 88; its relation to Hume’s philosophy of emotion 15 – 18, 164, 214 original instinct/principle 65 – 6, 81 – 2, 137

240 Index pity see compassion/pity Platonism 18 – 23, 26n15, 26 – 7n17, 35, 43, 99 Platonist, The 18, 22 Political Discourses 13 polytheism 102, 104, 106 – 7, 111 – 15, 117 – 22 Pope, Alexander 106 pride 55, 132, 136 – 8; its relation to vanity 38, 82; its relevance to religion 96; see also the doublerelation passions providence 19 – 20, 97, 118 – 19 rationalism 20, 35 – 7, 42, 145 – 51, 160n1; Hume’s rejection of 3, 77 – 9, 143n4, 155 – 60, 191 – 2; Hutcheson’s rejection of 19, 151 – 3; Locke’s 33 reason 155, 191, 201; causing the passions 116 – 17; motivationally ineffective 65, 131 – 2, 151 – 3; vs. passion 15, 17, 50, 52, 146 – 51, 154 – 60, 191 – 2; potential foundation of morals 35 – 6, 42, 44; potential foundation of religion 113 – 14, 118 – 20, 122 – 3 Reid, Thomas 226 – 8, 228n1 relativism 192 – 4, 204, 210 – 20, 224n1 representative quality argument, the 131, 156 – 9, 208n1 resentment 15, 45, 65 – 6 respect see esteem/respect Sceptic, The 18, 158, 192, 221 sceptical principle, the 26n13, 192 – 3, 212 Scepticism 13, 18, 21, 106, 222 – 3, 227 security 15, 51, 126n6, 171 self-hatred 43, 152 self-love see egoism sentimentalism 19 – 20, 36 – 7, 44, 160n1, 160n5 separability principle 135, 139 – 40 Shaftesbury, Lord [Anthony Ashley Cooper] 18 – 19, 24, 34 – 43, 47n5, 213

sorrow see grief/sorrow Stanhope, Lord [Philip Stanhope] 12, 25n9 Stewart, Dugald 105 Stewart, John 227 Stoic, The 18 Stoicism 18 – 23, 34 – 5, 43, 99, 174n2 Strahan, William 12, 226 superstition 20, 103 – 4, 106, 108 – 9n6, 113; vs. monotheism 120 – 5; vs. true religion 95, 97 – 100, 108n5, 119 – 20, 125 sympathy 3, 11, 20, 54 – 9, 62 – 4, 67n2, 80 – 1, 174n2 Treatise of Human Nature, A: Butler’s influence on it 70 – 2, 78; compared with Hume’s later work 77 – 82, 84 – 7, 128 – 9, 135 – 40, 142, 157 – 60, 168 – 9, 172; Hobbes’s influence on it 46n2, 171; Hume’s dissatisfaction with it 79, 226 – 8; Hume’s reworking of Book 2 11 – 12, 15 – 18; Hutcheson’s influence on it 83 – 5, 156; its antirationalism 78, 148, 155 – 7; its atomism 132 – 5; its egoism and hedonism 3, 23 – 4, 46, 49 – 67; its incoherence 72 – 5, 78; its irreligious implications 93 – 4, 108; its scepticism 21; its structure 9 – 12, 14 – 15; on space and time 13 – 14, 26n11; see also the representative quality argument Trenchard, John 103 – 6, 109n7 unpublished dissertation on geometry, the 12 – 15, 17, 25n9, 26n11, 113, 193 vanity 11, 115; its origin 55 – 6, 80 – 2; its relation to virtue 38 – 9, 67, 76 volition/the will 10, 25n4, 50, 155 – 8 Walpole, Robert 105 Warburton, William 105 – 7 will, the see volition/the will Wollaston, William 42, 151