Spectres of False Divinity: Hume's Moral Atheism

Spectres of False Divinity This page intentionally left blank Spectres of False Divinity Hume’s Moral Atheism Thom...

Author: Thomas Holden

75 downloads 876 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Spectres of False Divinity

This page intentionally left blank

Spectres of False Divinity Hume’s Moral Atheism

Thomas Holden

1

1

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With ofﬁces in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Thomas Holden 2010 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2010 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Library of Congress Control Number 2009943753 Typeset by Laserwords Private Limited, Chennai, India Printed in Great Britain on acid-free paper by MPG Books Group, Bodmin and King’s Lynn ISBN 978–0–19–957994–5 10 9 8 7 6 5 4 3 2 1

For Priya

This page intentionally left blank

Contents Preface Abbreviations 1. Hume’s Moral Atheism

ix xv 1

2. Mitigated Skepticism and Hume’s Liminal Natural Theology

19

3. The Argument from Sentimentalism 1: Hume’s Critique of Religious Passions

49

4. The Argument from Sentimentalism 2: Religious Passions and the Deity’s Moral Status

95

5. The Argument from Motivation

115

6. The Arguments from Evil

145

7. The Arguments from Determinism

181

Conclusion

209

Bibliography Index

221 241

This page intentionally left blank

Preface This book presents a historical and critical interpretation of Hume’s rejection of the existence of a deity with moral attributes: his moral atheism, as I shall call it. In Hume’s view, no ﬁrst cause or designer responsible for the ordered universe could possibly have moral attributes; nor could the existence (or non-existence) of such a being have any real implications for moral practice or human life. Although Hume is justly famous for his skeptical critique of traditional forms of theological speculation, he is not a skeptic or agnostic on the question of the deity’s moral character: he does not simply suspend judgment regarding the moral attributes, but categorically rules them out. This exception to Hume’s usual embargo on theological speculation is principled, for while his epistemological strictures do condemn the traditional program of constructive natural theology, they do not preclude the sort of via negativa arguments that Hume advances in his own case for divine amorality. In the following pages I situate Hume’s own version of moral atheism in its historical and philosophical context (Chapter 1), offer an interpretation of his multi-layered case for divine amorality, and show how Hume can endorse moral atheism while consistently maintaining his skeptical attitude toward traditional forms of cosmological and theological speculation (Chapter 2). Hume has two main arguments for moral atheism, each of which questions whether certain preconditions for the possession of moral attributes could be met by any original cause or designer. The ﬁrst argument is driven by Hume’s sentimentalist metaphysics of morals together with his account of the limited reach of human sentimental attitudes (Chapters 3 and 4); the second, by his theory of motivation together with his account of the particular utility of human sentimental psychology for human life (Chapter 5). I also examine two further arguments for moral atheism that are sometimes attributed to Hume: an argument from the quantity of evil in the world (Chapter 6) and an argument from determinism and its implication that any deity is the remote cause of vice (Chapter 7). However, Hume does not actually think these latter arguments sound.

x

preface

Although versions of these arguments are set out in his works, they proceed from assumptions that Hume does not himself accept: in his hands this reasoning is strictly ad hominem, and perhaps even a sly parody of traditional forms of natural theology. Finally, I also offer some remarks on the signiﬁcance of Hume’s commitment to moral atheism, both for our wider understanding of his theoretical and practical philosophy, and for our understanding of the philosophical history of irreligion in the early modern period (see the conclusion). Historical interpretation rather than philosophical assessment is my primary goal, but a pure historical exposition of a philosopher’s arguments without some sort of rationalization and the attendant pressures of immanent critique is neither possible nor indeed particularly desirable. The contextualist historian will rightly insist on the importance of understanding an author’s intellectual milieu in order to get a clearer ﬁx on his or her motivations and minimize the danger of an anachronistic misreading. But context alone cannot tell us what a historical ﬁgure’s views are, and there is no doing without rational reconstruction altogether. Nor is there any need to apologize for looking at arguments as arguments. So while I employ contextualist methods and offer a historical understanding of Hume’s texts in their original eighteenth-century environment, I also adopt a more philosophically engaged approach: I look for arguments, ask about their validity and underlying assumptions, and reﬂect on their wider implications and overall plausibility. Such assessment and criticism as I offer takes place within the broad framework of early modern philosophy, however, and my aim is always to understand and evaluate Hume’s philosophy on its own terms. I take a broadly holistic approach to Hume’s writings, using one text to inform the interpretation of another and referring to doctrines from his philosophy of mind, general epistemology, and moral theory for the light they cast on his philosophy of religion. There has been a growing appreciation of the systematic nature of the Hume’s thought in recent years, including, for instance, a clearer understanding of the way in which his skepticism toward the traditional pretensions of reason complements rather than contradicts his naturalistic epistemology; the importance of his theory of the passions in grounding his moral theory and his account of human responsibility; the continuities between the transparently irreligious later works and the more covert

preface

xi

irreligion of the earlier Treatise; the mobilization of the ‘mitigated skepticism’ ratiﬁed in the ﬁrst Enquiry in the later Dialogues; and so on. However, there has been little work connecting Hume’s philosophy of religion to his science of human nature—including his cognitive psychology, his moral psychology, and his sentimentalist theory of moral judgment—and it is these connections that I particularly emphasize in exploring Hume’s case for moral atheism. The reader must judge the results, but the methodological principle should be clear: ‘There is no question of importance, whose decision is not compriz’d in the science of man; and there is none, which can be decided with any certainty, before we become acquainted with that science.’ (T Introduction 6) Hume’s later philosophical writings present his most sustained and explicit discussions of religion and natural theology. Here I draw particularly on the Enquiry concerning Human Understanding (1748), the Enquiry concerning the Principles of Morals (1751), ‘Of Suicide’ (1756, immediately suppressed, but subsequently authorized for posthumous publication), the Four Dissertations (1757)—including ‘The Natural History of Religion’ and ‘A Dissertation on the Passions’—and the Dialogues concerning Natural Religion (1779, posthumously; though Hume had an early draft of this work under preparation in the early 1750s). Since these works were all published or composed within a few years of one another, it is reasonable to proceed on the assumption that they form a tolerably coherent and systematic whole—or such at least will be my working hypothesis, subject to conﬁrmation as one text plausibly illuminates another. I also utilize the earlier Treatise of Human Nature (1739–40), the Essays, Moral and Political (1741–2), and two important letters from the early 1740s (to Francis Hutcheson on March 16, 1740, and to William Mure on June 30, 1743). However, the main doctrines that I emphasize in these earlier texts are also corroborated in the later period. The Dialogues concerning Natural Religion is of course a central text for any systematic interpretation of Hume’s philosophy of religion. Along with Hume’s original eighteenth-century reviewers and most current Hume scholars, I take it that the character Philo speaks for Hume’s philosophical views throughout the Dialogues. Of course, to say that Philo speaks for Hume is not to deny that the other characters occasionally make points that Philo (and hence Hume) concurs

xii

preface

with, particularly as the shifting three-cornered debate develops and Philo ﬁnds himself in a temporary alliance with one or the other participant. And the claim is not that everything that Philo says is perfectly straightforward, sincere, and unironic: that in itself would be quite un-Humean. While not seriously disputable, the identiﬁcation of Philo with Hume is still occasionally disputed, sometimes by the partisans of the constructive natural theologian Cleanthes, and sometimes (the more recent fashion) by those who refuse to identify any particular character with Hume himself, either because Hume spreads himself out among the various competing characters, or because the text is simply too quicksilver to permit any reliable inference to a ﬁxed and determinate authorial intention. However, these readings typically operate at too great a distance from the ﬁne-grained detail of the argument and cumulative philosophical power of Hume’s text, and—quite crucially—none of them has begun to adequately engage the totally conclusive, point-by-point case for identifying Philo with Hume set out in Norman Kemp Smith’s magisterial commentary (and ably reprised and expanded by J. C. A. Gaskin in the light of certain subsequent misguided challenges).¹ So although we will be seeing numerous parallels—including several previously unremarked—between Philo’s views and those expressed elsewhere in Hume’s own voice, I see no need to relitigate this interpretive question here, but will simply proceed on the assumption that Philo speaks for Hume. ‘Next to the ridicule of denying an evident truth, is that of taking much pains to defend it’ (T 1.3.16.1). Some of this material appears in print elsewhere. Sections 3.2 and 3.4 draw on an interpretation ﬁrst assayed in ‘Hume on Religious Affect’, Archiv f¨ur Geschichte der Philosophie 89 (2007), 283–06. Sections 7.2 and 7.3 draw heavily on ‘Religion and Moral Prohibition in Hume’s ‘‘Of Suicide’’ ’, Hume Studies 31 (2005), 189–210. I am grateful to the editors and publishers of both Hume Studies and Archiv f¨ur Geschichte der Philosophie for permission to reprint this material. ¹ Norman Kemp Smith, ‘Introduction’ to David Hume, Dialogues Concerning Natural Religion, ed. Norman Kemp Smith (Indianapolis: Bobbs-Merrill, 1947), 1–75: 57–74. J. C. A. Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, NJ: Humanities Press International, 1988), 209–18.

preface

xiii

I am greatly indebted to three anonymous referees for Oxford University Press, whose searching comments resulted in many improvements. Thanks are also due to my editor Peter Momtchiloff both for his encouragement of this project and his selection of such insightful referees, and to my philosophical mentors Simon Blackburn, Don Garrett, and Galen Strawson for all their unfailing help and support over the years. In thinking through the interpretive and philosophical puzzles addressed in this study I have particularly beneﬁted from conversations with Angela Coventry, James Dye, Lorne Falkenstein, Michael Gill, Lívia Guimar˜aes, Sean Greenberg, Nicholas Jolley, Daniel Kaufman, Patrick Miller, Peter Millican, Hans Muller, Robert Pasnau, Daniel Selcer, Lisa Shapiro, Matthew Smith, and Aaron Zimmerman. Apologies to those I may have forgotten. Finally, love and gratitude to Priya and Meha for ﬁring my animal spirits and passions.

This page intentionally left blank

Abbreviations References to Hume’s texts DNR

Dialogues concerning Natural Religion, in Dialogues concerning Natural Religion: And Other Writings, ed. Dorothy Coleman (Cambridge: Cambridge University Press, 2007), followed by part and paragraph number

DP

‘A Dissertation on the Passions’, in A Dissertation on the Passions; The Natural History of Religion: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2008), 1–29, followed by section and paragraph number

E

Essays, Moral, Political, and Literary, ed. E. F. Miller (Indianapolis: Liberty Fund, 1987), followed by page number

EHU

An Enquiry concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000), followed by section and paragraph number

EPM

An Enquiry concerning the Principles of Morals: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2003), followed by section and paragraph number

HE

History of England, ed. William B. Todd, 6 vols. (Indianapolis: Liberty Fund, 1983), followed by volume and page number

L

The Letters of David Hume, ed. J. Y. T. Grieg, 2 vols. (Oxford: Clarendon Press, 1931), followed by volume and page number

xvi

abbreviations

LG

A Letter from a Gentleman to his Friend in Edinburgh, in A Treatise of Human Nature: A Critical Edition, ed. D. F. Norton and M. J. Norton, 2 vols. (Oxford: Clarendon Press, 2007), 419–31, followed by paragraph number

NHR

‘The Natural History of Religion’ in A Dissertation on the Passions; The Natural History of Religion: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2008), 31–87, followed by section and paragraph number

NL

New Letters of David Hume, ed. Raymond Klibansky and Ernest C. Mossner (Oxford: Clarendon Press, 1954), followed by page number

T

A Treatise of Human Nature: A Critical Edition, ed. D. F. Norton and M. J. Norton, 2 vols (Oxford: Clarendon Press, 2007), followed by book, part, section, and paragraph number

TA

An Abstract of a book lately published, entitled, A Treatise of Human Nature, in A Treatise of Human Nature: A Critical Edition, ed. D. F. Norton and M. J. Norton, 2 vols (Oxford: Clarendon Press, 2007), i. 403–17, followed by paragraph number

1 Hume’s Moral Atheism 1.1. Introduction ‘[I]f we had no certain and settled notion of the goodness and justice, and truth of God, he would be altogether an unintelligible being; and religion, which consists in the imitation of him, would be utterly impossible.’ Thus Archbishop John Tillotson in his popular Sermons (1695–1704; published posthumously), distilling a necessary condition for any religion with real implications for practice or conduct.¹ George Berkeley, Dean of Derry and later Bishop of Cloyne, endorses much the same requirement in his study of irreligion and free-thought, the 1732 dialogue Alciphron. ‘[A]t bottom the being of God is a point in itself of small consequence, and a man may make this concession without yielding much,’ the more cynical of Berkeley’s characters remarks. So long as the moral attributes of the deity remain in doubt, ‘nothing can be inferred from such an account of God, about conscience, worship, or religion.’² Berkeley’s other characters implicitly admit the point: without the ﬁxed star of a just and good divinity, the bare admission that God exists offers no guidance for moral or religious practice. This view of the preconditions for a practically meaningful religion would be widely shared by the philosophers and theologians of Hume’s generation. It is invoked, for instance, by Bishop William Warburton of Gloucester, ¹ John Tillotson, Sermons on several subjects and occasions, 12 vols (London: 1742–4), vii. 2280. This particular sermon (‘Concerning the Perfection of God’) was ﬁrst published after Tillotson’s death in 1694, in Tillotson, Sermons, ed. Ralph Barker, 14 vols (London: 1695–1704). ² George Berkeley, Alciphron, in David Berman (ed.), Alciphron, or the Minute Philosopher in Focus (London: Routledge, 1993), 17–161: 105, 107.

2 hume ’s moral atheism the ﬁrebrand controversialist and a savage critic of Hume, in his anti-deistic polemics of the 1750s. For Warburton, since the existence of an amoral deity would have no implications for moral conduct, belief in such a being is merely the ‘Ape of Religion.’³ Hume himself registers the point through his character Cleanthes, the spokesman for the traditional program of constructive natural theology in the Dialogues concerning Natural Religion (1779; published posthumously). Absent some conception of the moral attributes of the deity, Cleanthes suggests, the cosmological speculations of natural theology are for all practical purposes inert: ‘For to what purpose establish the natural attributes of the Deity, while the moral are still doubtful and uncertain?’ (DNR 10.28). The central claim of this essay is that Hume positively rejects the existence of a god with moral attributes: that he is (what we might call) a moral atheist. Whatever beliefs he might have concerning the existence or non-existence of a ﬁrst cause or designer, for Hume nothing approximating to this traditional conception of the divine could possibly have moral characteristics. Hume is not merely an agnostic on this point, as might at ﬁrst seem entailed by his skeptical attitude toward speculation about the divine attributes. Nor is his moral atheism merely tentative or provisional in spirit: it is not simply a probationary hypothesis hazarded from a position of little real information. Rather, Hume has what he takes to be a decisive case against the being of a god with moral attributes—an irreligious version of via negativa reasoning, concluding in the inconsistency of moral characteristics with the central, identifying features of the deity. As is well known, Hume maintains in the ﬁrst Enquiry and again in the Dialogues that natural theology—the program of using natural human reason to reach conclusions about the being or attributes of the deity—is quite incapable of conﬁrming that the deity has a moral character. If there is a ﬁrst cause or designer, this being ‘discovers himself only by some faint traces and outlines, beyond which we have no authority to ascribe to him any attribute or perfection’ (EHU 11.27). So we cannot disprove moral atheism through unaided human ³ William Warburton, A View of Lord Bolingbroke’s philosophy; in four letters to a friend. Letters ﬁrst and second (London, 1754), 69.

hume ’s moral atheism

3

reason: the speculative program of natural theology fails to rule out an amoral ﬁrst cause.⁴ Still, this familiar Humean thesis may seem to leave open the possibility that the deity has moral attributes, notwithstanding our inability to prove the point. By denying knowledge, it may even seem to make way for faith. But what this assessment misses is the fact that Hume has further arguments in hand: arguments that add up to a positive case for moral atheism. As Hume himself was quite aware, his theory of moral motivation, his determinism, and his sentimentalist metaphysics of morals all point to the exclusion of any original cause of all from the moral sphere. Not only is it quite impossible to prove that the ﬁrst cause or designer has moral attributes, Hume further holds that no such being could possibly have a moral character. In this essay, I bring out this aspect of Hume’s challenge to traditional theism, and examine the relationship between his case for moral atheism on the one hand, and his well-known skeptical critique of natural theology on the other. My claim is that Hume’s case for moral atheism is a central plank of his naturalistic and irreligious agenda. It complements his critique of theistic metaphysics in the Enquiry and Dialogues, and threatens to rule out any practically meaningful religion—any religion at all, that is, in Tillotson’s, Berkeley’s, and Warburton’s morally freighted sense. (I set aside the question of whether Hume accepts the existence of any original cause or organizing principle as beyond the scope of this essay.⁵) In this introductory chapter I ﬁx some basic points of terminology and set out my interpretive thesis in a little more detail. I deﬁne ⁴ On this point, see J. C. A. Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, NJ: Humanities Press International, 1988), 53–58; Nicholas Capaldi, David Hume: The Newtonian Philosopher (Boston: G. K. Hall & Co., 1975), 188–97; Stephen Buckle, Hume’s Enlightenment Tract (Oxford: Clarendon Press, 2001), 279–80. For discussion of the implications of this Humean thesis for religious practice, see Simon Blackburn, ‘Playing Hume’s Hand,’ in D. Z. Phillips and Timothy Tessin (eds), Religion and Hume’s Legacy (New York: St Martin’s Press, 1999), 3–16. ⁵ Hume clearly holds that there is no prospect of demonstrating the existence of an original cause or ultimate organizing principle (T 1.3.3.1–8, 1.3.15.1, TA 11, LG 26–7, EHU 12.29, DNR 9.5–6). But most of the time he seems inclined to grant that there is some sort of ﬁrst cause (LG 34, E 145, DNR 2.3, 11.15) or ultimate organizing principle (LG 27, 30, NHR Introduction 1, 1.5, EHU 11.11, DNR 12.6, 12.7, 12.33) behind the ordered universe. Of course, the precise meaning, basis, and purpose of these admissions deserve close interpretive scrutiny. For commentary on some of the main issues, see Paul Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion (New York: Oxford University Press, 2008), 113–28; J. C. A. Gaskin, Hume’s Philosophy of Religion, 74–93, 219–21.

4 hume ’s moral atheism moral atheism, identify Hume as a particularly strong sort of moral atheist (Section 1.2), and relate this doctrine to the early modern period’s own ways of thinking about religion, irreligion, theism, and atheism (Section 1.3). Finally, I close out the chapter by introducing two possible large-scale objections to the interpretation of Hume as any sort of moral atheist (Section 1.4). This will help to put some important interpretive issues on the table and set the stage for my big-picture survey of Hume’s overall case for divine amorality in Chapter 2.

1.2. Hume’s Moral Atheism: The Interpretive Proposal To ﬁx my interpretive proposal and prepare the way for an examination of Hume’s arguments, a more precise deﬁnition of moral atheism is required.⁶ Thus far I have characterized the position, somewhat loosely, as the denial of the existence of a god with moral attributes.⁷ But to understand just what this denial amounts to, we need to understand just what is intended by the term ‘god,’ and what exactly ‘moral attributes’ are. First, in Hume’s usage, ‘God’ (along with synonyms like ‘the Deity’ and ‘the Divine Being’) picks out the ultimate cause of the universe, or at least the ultimate cause of the order found in the universe. Rather than specifying its meaning by way of a list of deﬁnitive intrinsic properties (traditionally, omniscience, omnipotence, and the rest), Hume identiﬁes the referent of ‘God’ in relational terms: God is the being or principle, whatsoever its intrinsic nature, that produced ⁶ J. C. A. Gaskin ﬁrst introduced this term of art in the characterization of Hume’s philosophy in his ‘Hume, Atheism, and the ‘‘Interested Obligation’’ of Morality,’ in David Fate Norton, Nicholas Capaldi, and Wade L. Robison (eds), McGill Hume Studies (San Diego: Austin Hill Press, 1976), 147–59, where a moral atheist is ‘anyone who denies the existence of any god having moral attributes’ (151). I am adopting Gaskin’s term but also stipulating its meaning more precisely for my own purposes. ⁷ In this usage a ‘moral atheist’ is then neither an atheist who lives according to moral standards, nor someone who denies the force and reality of moral claims. (For this latter usage, see Charles Kingsley, His Letters and Memories of His Life, ed. Frances Eliza Kingsley, 2 vols (London: H. S. King and Co., 1877), ii. 75.)

hume ’s moral atheism

5

the universe, or at least produced the order of the universe.⁸ For Hume, the idea of God is thus an instance of his category of ‘relative ideas’: it picks out its referent by way of its relational place rather than (as with a ‘positive idea’) by specifying its intrinsic character.⁹ Hume’s character Philo (widely and in my view correctly regarded as the spokesman for Hume’s own views) introduces this approach early in the Dialogues, with his interlocutors apparently accepting this ⁸ There are two exceptions to this usage in Hume’s texts, but neither of them amounts to a serious counterexample to the current interpretation. (1) In the ﬁrst Enquiry Hume writes of ‘The idea of God, as meaning an inﬁnitely intelligent, wise, and good Being’ (EHU 2.6). But here in section 2 (‘Of the Origin of Ideas’) Hume is invoking the traditional monotheistic conception of a supremely perfect being simply as an illustration of his empiricist thesis that the ‘creative power of the mind amounts to no more than the faculty of compounding, transposing, augmenting or diminishing the materials afforded us by the senses and experience,’ even when forming such a sublime idea as this (EHU 2.5). So Hume wants to allow that we can construct some such idea of a supremely good and intelligent being (and in focusing on this particular idea as a test case for empiricism he is directly following Locke (John Locke, An Essay concerning Human Understanding, ed. P. H. Nidditch (Oxford: Oxford University Press, 1975), book 2 chapter 23 sections 33–6).) But the term ‘God’ does not have this connotation when Hume addresses natural theology directly. (2) The Natural History of Religion deﬁnes religion as ‘[t]he belief in an invisible, intelligent power’ (NHR introduction 1; see also NHR 2.2, 2.5, 4.2), and Hume goes on to refer to such putative invisible agents as so many ‘deities’ and ‘gods’ (NHR 2.3, 2.4, 3.2, 3.3). So here Hume seems to have an entirely different deﬁnition of the meaning of ‘god’ in play, one that picks out its referent by way of the deﬁning properties of invisibility, intelligence, and potency rather than (as with his usual usage) picking out its referent by way of the relational property of standing to the ordered universe as some sort of original cause or ultimate organizing principle. However, even here in the Natural History Hume goes out of his way to emphasize that belief in invisible intelligent powers must be clearly distinguished from belief in an ultimate ‘author of nature’ or ‘original cause of all things,’ i.e. a god or deity in his usual sense of the word (NHR 4.1, 5.2). Thus the title of section 4 declares that the polytheistic divinities of ancient myth were ‘not considered as creators or formers of the world,’ and Hume goes on to explain that for this reason ‘the gods of all polytheists are no better than the elves or fairies of our ancestors, and merit as little any pious worship or veneration. These pretended religionists are really a kind of superstitious atheists, and acknowledge no being, that corresponds to our idea of a deity. No ﬁrst principle of mind or thought: No supreme government and administration: No divine contrivance or intention in the fabric of the world’ (NHR 4.2; see also 4.7). In terminological strictness, then, ‘[i]t is a fallacy, merely from the casual resemblance of names, without any conformity of meaning, to rank such opposite opinions [as ancient polytheism and modern theism] under the same denomination’ (NHR 4.1). ⁹ On Hume’s distinction between relative and positive ideas, see Daniel E. Flage, ‘Hume’s Relative Ideas’, Hume Studies 7 (1981), 55–73, and ‘Relative Ideas Revisited: A Reply to Thomas,’ Hume Studies 8 (1982), 158–71. (The essential points are also distilled in Flage, ‘Relative Ideas Re-Viewed’, in Rupert Read and Kenneth A. Richman (eds), The New Hume Debate: Revised Edition (New York: Routledge 2007), 138–55). See also Galen Strawson, The Secret Connexion (Oxford: Clarendon Press, 1989), 50–3, 122–3.

6 hume ’s moral atheism characterization of the subject matter of their debate: ‘Nothing exists without a cause; and the original cause of this universe (whatever it be) we call God’ (DNR 2.3). Similarly, Hume regularly uses the deﬁnite description ‘the original cause of all things’—along with correlates like ‘the original source of all things,’ ‘the ﬁrst causes of the universe,’ or ‘the original cause’—interchangeably with phrases like ‘God’ and ‘the Deity’ (E 145, EHU 8.32, NHR 5.2, DNR 2.20, 11.14, 11.15). In other places Hume identiﬁes the deity by way of its role in producing, not the universe simpliciter, but the order of the universe. Thus Philo takes the core of natural theology to concern the character of ‘the cause or causes of order in the universe,’ and presents controversy about the divine attributes to be disagreement over the nature of ‘the principle which ﬁrst arranged, and still maintains order in this universe’ (DNR 12.33, 12.7). So if there is either an ultimate cause of the universe or an ultimate source of the order in the universe—a being or principle that might serve as the terminus of either the cosmological or the teleological argument, for instance—then we have a deity.¹⁰ Hume’s relational characterization of the referent of ‘God’ reﬂects one core feature of the traditional conception of the divine: both the god of scripture and the theoretically articulated god of the philosophers are each traditionally understood as the cause or organizing principle behind the ordered universe, and hence each would satisfy Hume’s standards for the application of the term. But in picking out the referent of ‘God’ purely by way of its place in a system of relations rather than by any particular speciﬁcation of intrinsic properties, Hume’s usage is highly ecumenical, allowing the widest possible debate over the actual nature of this primary being or principle. At the same time, it allows Hume to piously afﬁrm the deity’s existence without committing to any substantive conception of the divine attributes, and thus without surrendering hostages to any particular traditional religious world views. Indeed, as Philo notes in the Dialogues, many thinkers traditionally regarded as atheists could agree that there is some such primary being, and hence ¹⁰ This disjunctive characterization is quite explicit in the ﬁrst Enquiry, where in the dialogue ‘Of a Particular Providence and a Future State’ the character playing Epicurus frames the debate by identifying ‘the gods’ as ‘the authors of existence or order of the universe’ (EHU 11.14, my emphasis).

hume ’s moral atheism

7

that there is a god in Hume’s spare sense, while of course rejecting the theist’s particular elaboration of this being’s intrinsic character (DNR 12.7).¹¹ For the purposes of this essay, I will be adopting Hume’s usage of ‘God’ and its various correlates to refer to the original cause or organizing principle of the ordered universe. Like Hume, I use the capitalized ‘God’ as a proper name of whatever being occupies this relational place; but I also use the lower case ‘god’ (and ‘deity,’ and so on) as an abbreviation for the deﬁnite description ‘the original cause or organizing principle of the ordered universe.’ For the sake of brevity, I usually shorten the deﬁnite description ‘the original cause or organizing principle of the ordered universe’ simply to ‘the original cause’ (or ‘the primary cause,’ ‘the ﬁrst cause,’ and so on)—although the reader should bear in mind that this is just shorthand for the more complex disjunctive characterization. So moral atheism is the position that, if there is an original cause or organizing principle behind the ordered universe, then this being or principle lacks moral attributes. But what exactly does this denial of ‘moral attributes’ amount to? There are two ways of understanding such a denial, and hence two versions of moral atheism. The weaker position is the denial of a morally praiseworthy god.¹² As I will use this predicate phrase, to qualify as morally praiseworthy one needs an overall character that is morally laudable when all things are considered, not simply a character that is morally laudable in this or that respect. (No doubt Caligula has a character that is morally laudable in some respects, but this is not enough for him to qualify as morally praiseworthy in my sense.) According to weak moral atheism, then, if there is a deity, then it lacks a morally praiseworthy character, and the traditional theistic conception of this primary being as some sort of paragon of virtue or moral worth is simply mistaken. Numerous commentators both in ¹¹ Hume’s early memoranda notes indicate that he encountered this way of conceptualizing the matter in Pierre Bayle, the seventeenth-century skeptic. Thus a note drawn from his reading of ‘Baile’ runs: ‘The Center of Unity of all Men with relation to Religion is: That there is a First Cause. As you augment the Proposition you ﬁnd Non-Conformists, Atheists, Epicureans, Idolaters, those who maintain the Extension, Composition, Necessity of the First Cause etc.’ (See E. C. Mossner, ‘Hume’s Early Memoranda, 1729–1740: The Complete Text,’ Journal of the History of Ideas 9 (1948), 292–518: 500 (note 2.8).) ¹² This is apparently the sense of ‘moral atheism’ intended by J. C. A. Gaskin in ‘Hume, Atheism, and the ‘‘Interested Obligation’’ of Morality,’ 151, 155.

8 hume ’s moral atheism Hume’s day and our own have taken him to be at least tentatively endorsing this weaker form of moral atheism. This interpretation is typically defended with reference to Hume’s treatment of the problem of evil in the Dialogues and his apparent preference there for the hypothesis that ‘the original source of all things . . . has no more regard to good above ill than to heat above cold’ (DNR 11.14). But advocates of this line of interpretive argument can at best only hope to show that Hume tentatively inclines towards weak moral atheism, for his comments about the likely indifference of the deity are presented in the form of a cautious speculative hypothesis, and quite explicitly fall short of a deﬁnitive afﬁrmation. (Hume is perfectly clear that we are not entitled to assert the absolute logical incompatibility of the deity’s moral goodness and the existence of evil, a conclusion ‘too presumptuous for creatures so blind and ignorant’ (DNR 11.12).) Furthermore, there may be a deeper problem with this line of interpretive argument, for there is an alternative way of understanding the texts where Hume moves from observations about the world’s evils to the hypothesized moral indifference of the creator. This inference proceeds in apparent deﬁance of Hume’s usual doubts about our ability to infer anything substantive regarding the deity’s distinctive intrinsic nature from the character of the observed world, suggesting that he may not be altogether in earnest. Perhaps, in drawing up this speculative hypothesis about the deity’s presumed moral indifference from the evidence of the world around us, Hume is simply parodying the overreaching style of traditional natural theology.¹³ In any case, there is a stronger version of moral atheism in Hume’s texts, or so I shall argue. This stronger version of moral atheism denies the existence, not merely of a morally praiseworthy god, but of a morally assessable god. On this view, if there is a ﬁrst cause or organizing principle, this being or principle is not a proper object of moral assessment or evaluation one way or the other. Here the rejection of divine ‘moral attributes’ is taken in its broadest possible sense. It is not merely that the deity lacks positive moral virtues. Rather, it stands outside the scope of moral assessment altogether. It follows, of course, that the deity is not morally praiseworthy: strong moral atheism entails ¹³ I discuss this interpretive issue and defend this alternative reading in section 6.4.

hume ’s moral atheism

9

weak moral atheism. But while weak moral atheism is consistent with a morally reprehensible god, or with a deity that is balanced neutrally between praiseworthy and blameworthy tendencies, for the strong moral atheist the deity stands outside the moral order entirely. The deity is not then amoral in the sense importing culpability, the sense in which a heartless villain might be described as amoral. (Such amorality is of course simply a certain type of immorality.) Rather it is amoral in the sense in which a stone or a snowstorm might be described as amoral—as something removed from the moral realm altogether. Hereafter, whenever I use the phrase ‘moral atheism’ without qualiﬁcation, I mean to refer to this stronger version of the doctrine. The thesis that Hume is a moral atheist—the central claim of this essay—should then be understood as the thesis that Hume denies the existence of a morally assessable ﬁrst cause or organizing principle.

1.3. Atheism and Moral Atheism Neither weak nor strong moral atheism amounts to a form of atheism by the lights of our twenty-ﬁrst-century use of the term. The moral atheist is simply rejecting the existence of one particular sort of deity, and this leaves open the possibility of other (amoral) types of god. But in Hume’s era, the position that there is no god with moral attributes was characterized as a form of atheism, not simply in the vituperative usage loosely applied to any form of unorthodox or supposedly impious belief, but also in a more precise and philosophically rigorous sense. In the current section I sketch this early modern terminological background. Early modern controversialists often used ‘atheism’ and its cognates in a loose and abusive sense, imputing either some form of heterodox religious belief or attitude, or simply licentious (and thus ‘godless’) behavior—along, of course, with the writer’s revulsion at such impiety. The word functioned, as one historian has put it, ‘as a majestic term of reproach and condemnation.’¹⁴ Catholics were atheists to ¹⁴ D. C. Allen, Doubt’s Boundless Sea: Skepticism and Faith in the Renaissance (Baltimore: Johns Hopkins Press, 1964), 1.

10 hume ’s moral atheism Protestants; Protestants were atheists to Catholics. Noncomformists and Independents were atheists to sober-minded Anglicans; the dissolute and libertine were atheists to all. Even Milton’s rebel angels, in the midst of their open war against the creator, are, apparently, an ‘Atheist crew.’¹⁵ Francis Bacon’s Essays (1601) registers one version of this inexact and abusive usage. There are, he admits, some who are properly called atheists. But polemicists multiply their numbers by applying the word indiscriminately, and ‘all that Impugne a received Religion, or Superstition, are by the adverse Part, branded with the Name of Atheists.’¹⁶ However, as Bacon allows, a more exact sense is also current in this period, a sense that provides the word’s core meaning and lends the wider vituperative uses their rhetorical and condemnatory force. First, in this stricter usage, atheism is a matter of theoretically articulated doctrine, not merely the alleged unbelief sometimes taken to be implied by immoral behavior in apparent deﬁance of God’s laws. (Atheism in the strict and philosophical sense is thus ‘contemplative’ or ‘speculative atheism,’ not simply the ‘practical atheism’ of a wanton life.) And second, in this core sense used by early moderns, atheists are those who reject the traditional monotheistic doctrine of a creator god that is unitary, intelligent, personal, moral, and providential: a denial that might involve the outright denial of his being, or simply the rejection of one or more of these deﬁning attributes. Epicurean anti-providentialists and Spinozistic pantheists are atheists by these standards. Deists may or may not be, depending on their speciﬁc account of the divine nature.¹⁷ The traditional polytheistic systems of pre-Christian Europe will also qualify as forms of ‘superstitious atheism,’ as Hume himself remarks in the Natural History of Religion ¹⁵ John Milton, Paradise Lost (1667), ed. Christopher Ricks (New York: New American Library, 1968), 184 (book 6 line 370). ¹⁶ Francis Bacon, ‘Of Atheisme,’ in The Essayes or Counsels, Civill and Moral, ed. Michael Kiernan (Oxford: Clarendon Press, 1985), 51–4: 52–3. ¹⁷ Consider this from the Boyle lecturer John Harris’s The Atheistical Objections against the Being of a God, and his Attributes, fairly considered and fully refuted (1698): ‘[Deists] . . . though in Words they may profess to believe and honour a God, yet in Reality they deny him, and have no Manner of Notion of his true Nature and Perfections. . . . [I]f [a man] have not such a Belief of God, as implies in it a Knowledge of the Perfections of his Nature, he may call himself by as ﬁne and fashionable Names as he pleases, and pretend to Deism and natural Religion; but in reality he is an Atheist, and so ought to be esteemed by all Mankind’ (quoted in Paul Russell, The Riddle of Hume’s Treatise, 51).

hume ’s moral atheism

11

(NHR 4.2).¹⁸ The theologian and philosopher Samuel Clarke’s celebrated 1704 Boyle lecture A Demonstration of the Being and Attributes of God nicely demonstrates this stricter usage, as Clarke begins his discussion by emphasizing the atheism of denying the various traditional divine attributes and explicitly deﬁnes atheists as those ‘who either disbelieve the being of God or would be thought to do so or, which is all one, who deny the principal attributes of the divine nature.’¹⁹ Also illustrative are the comments of the religious writer and would-be epic poet Richard Blackmore prefacing his counterblast to Lucretius, Creation: A Philosophical Poem Demonstrating the Existence and Providence of God (1712): there are two sorts of Men, who without Injustice have been call’d Atheists; those who frankly and in plain terms have deny’d the Being of God; and those, who tho’ they asserted his Being, denied those Attributes and Perfections, which the Idea of a God includes; and so while they acknowledge the name, subverted the thing. These are as real Atheists, as the former, but less sincere.²⁰

Or, for another example, we might consider the deﬁnition of speculative atheism deployed by the Scottish ‘common sense’ philosopher James Beattie in the speciﬁc context of charging Hume with atheism in the savagely polemical Essay on the Nature and Immutability of Truth (1770): Perhaps it will be asked, what I mean by the word Atheist? I answer, A reasonable creature, who disbelieves the being of God, or thinks it inconsistent with sound reason, to believe, that the Great First Cause is perfect in holiness, power, wisdom, justice, and beneﬁcence,—is a speculative Atheist.²¹

Not just any form of heterodox religious opinion will qualify as atheistic by these more exact standards: only those that go so far as to deny an attribute regarded as an essential aspect of the traditional theistic god. But moral atheism certainly qualiﬁed as a form of atheism ¹⁸ I cite the relevant passage in note 8 above. ¹⁹ Samuel Clarke, A Demonstration of the Being and Attributes of God, and Other Writings, ed. Ezio Vailati (Cambridge: Cambridge University Press, 1998), 3. ²⁰ Richard Blackmore, Creation: A Philosophical Poem. Demonstrating the Existence and Providence of God, 2nd edn (London 1712), ix. For commentary, see David Berman, A History of Atheism: From Hobbes to Russell (London: Routledge, 1988), 98. ²¹ James Beattie, An Essay on the Nature and Immutability of Truth, in opposition to sophistry and scepticism (1770) (facsimile reprint: Stuttgart-Bad Cannstatt: Verlag, 1974), 488 note.

12 hume ’s moral atheism properly so-called, for the essential attributes of the theistic divinity included a morally praiseworthy character. As we saw in the extracts from Tillotson, Berkeley, and Warburton opening this chapter, the doctrine of an amoral ﬁrst cause was widely identiﬁed by theologians of Hume’s day as a particularly insidious form of unbelief. For these three philosophical prelates at least, the denial of a creator with moral attributes was a form of ﬁfth column atheism, implicit in the dark doctrines of Hobbes and Spinoza, and the threatened upshot of the attack on scriptural religion presented by the rising tide of deists and other so-called ‘freethinkers.’ Indeed, in British intellectual culture before Hume and Darwin, with the existence of some sort of ultimate designing intelligence apparently beyond serious questioning, moral atheism seems to have been regarded as the most serious philosophical challenge to orthodox belief.²² In the great seventeenth- and eighteenth-century bestiaries of irreligion, the atheism of denying the deity’s moral praiseworthiness is typically bound up with the atheism of denying a morally ordered providence—denying, that is, that the deity governs the world according to moral principles. In early modern Christian thought, these two doctrines tend to merge: the orthodoxy that the deity is morally praiseworthy is assimilated with the doctrine that the deity is a providential agent, an active superintendent watching over human history. Divine goodness is not simply a Form of Forms, an abstract transcendental posit or regulative principle mandated by theoretical or practical reason. Rather, divine goodness is manifested quite concretely in the deity’s beneﬁcent care for and just government of this world. The denial of a morally structured providential order in our universe is thus tantamount to the denial of God’s moral praiseworthiness, and each of these scandalous positions is classiﬁed as a form of atheism ²² On the use of ‘atheism’ and its cognates in the seventeenth and eighteenth centuries, see David Berman, A History of Atheism in Britain; Isabel Rivers, Reason, Grace, and Sentiment: A Study in the Language of Religion and Ethics in England, 1660–1780, 2 vols (Cambridge: Cambridge University Press, 1991–2000), ii. 9–11; Paul Russell, The Riddle of Hume’s Treatise, ch. 5; A. P. Martinich, The Two Gods of Leviathan: Thomas Hobbes on Religion and Politics (Cambridge: Cambridge University Press, 1992), 19–30; David Wootton, Paolo Sarpi: Between Renaissance and Enlightenment (Cambridge: Cambridge University Press, 1983), 4–5; D. C. Allen, Doubt’s Boundless Sea.

hume ’s moral atheism

13

properly so called.²³ Thus Ralph Cudworth’s compendious attack on irreligious philosophy The True Intellectual System of the Universe (1678) characterizes the denial of a morally ordered providence as an express form of atheism.²⁴ The classicist Richard Bentley endorses a similar categorization in his 1692 Boyle lectures, as does Bishop Francis Gastrell in the lectures of 1697, and Bishop Joseph Butler in his monumental Analogy of Religion (1736).²⁵ But the classiﬁcation comes out most vividly in the polemical writings of Warburton, where it appears repeatedly in the scourging of suspected deists and freethinkers. For Warburton, any version of ‘naturalism’—that is, ‘the belief of a God, the Creator and Physical Preserver, but not Moral Governor of the World’²⁶—is a clear and particularly repellent form of atheism: For tho’ the principles may be called naturalism, yet if Scripture has deﬁned an atheist right, to be one who has no hope, and is without god in the ²³ According to George T. Buckley, the ﬁrst use of ‘atheism’ in English, in an essay by Sir John Cheke from around 1540, carries just this sense of a rejection of a moral providence rather than the outright rejection of any sort of god. George T. Buckley, Atheism in the English Renaissance (New York: Russell and Russell, 1965), 64. More recently, Michael J. Buckley has noted that Cheke’s essay is a free translation of Plutarch’s On Superstition: Michael J. Buckley, At the Origins of Modern Atheism (New Haven: Yale University Press, 1987), 9. ²⁴ Ralph Cudworth, The True Intellectual System of the Universe, wherein all the reason and philosophy of atheism is confuted and its impossibility demonstrated (London: 1678), 79. ²⁵ For Bentley, ‘the Divine Inspection into the affairs of the World doth necessarily follow from the Nature and Being of God. And he that denies this, doth implicitly deny his Existence . . . the Existence of God and his Government of the World do mutually suppose and imply one another’ (Richard Bentley, The Folly and Unreasonableness of Atheism (London: 1692), 5–6). Similary for Gastrell, the atheist is one who ‘says there is no God that governs the world, and judgeth the earth’ (Francis Gastrell, The Certainty and Necessity of Religion in general (London: 1697)); Joseph Butler, The Analogy of Religion (1736), in The Works of Bishop Butler, ed. J. H. Bernard, 2 vols. (London: Macmillan and Co., 1900), ii. 269. The same categorization can of course also be found in early modern Continental philosophers, as when the great German moral philosopher and jurist Samuel Pufendorf writes that ‘Whoever wholly violates and breaks through this Obligation [to revere God’s majesty and obey his commandments and laws], stands guilty of the most heinous Charge of Atheism; because he must at the same time deny either the Existence of God, or his Care of human Affairs. Which two sins, with regard to their moral Consequences and Effects, are equivalent to each other; and either of them overthrows all Religion’ (Samuel Pufendorf, Of the Law of Nature and Nations, tr. Basil Kennet et al. with the notes of Jean Barbeyrac (London, 1729), second emphasis mine). ²⁶ William Warburton, Remarks on Mr. David Hume’s Essay on the natural history of religion (London 1757), 9.

14 hume ’s moral atheism world, our Professor of Naturalism comes within the description. For tho’ he acknowledges the being of a God, yet as he is without God in the world, that is a Being who presides over it, as the moral Governor of it, which is the foundation in which all Religion stands, Religionists will seek no other title for him. And surely he will be properly deﬁned. For tho’ the abstract term Atheism carries, as it’s principal idea, a relation to God’s being: yet Atheist, the concrete, seems to have it’s chief relation to his government.²⁷

In sum, while moral atheism may not be a form of atheism by our twenty-ﬁrst-century criteria for the use of the term, in Hume’s day it was regarded as a clear form of atheism, both meeting the proper standards for the application of the word, and deserving the opprobrium and condemnatory force associated with it.²⁸

1.4. The Objection from Textual Insufﬁciency Before examining the details of Hume’s case for moral atheism I need to preempt two possible big-picture objections to the interpretation ²⁷ William Warburton, A View of Lord Bolingbroke’s philosophy; in four letters to a friend. Letters ﬁrst and second (London, 1754), 72–3. Although Warburton is attacking Bolingbroke rather than Hume here, he also characterizes Hume as a ‘naturalist’ in other contexts. See Warburton, Remarks on Mr. David Hume’s Essay, 9, and Letters of a late Eminent Prelate to One of His Friends, ed. Richard Hurd (Kidderminster, 1808), 175. ²⁸ Although the threat of moral atheism was regarded as a particularly pressing challenge in Hume’s day, the issue is of course a perennial of natural theology. Consider, for example, Cicero’s dialogue De Natura Deorum, a work that Hume knew well and borrowed from stylistically when shaping his own Dialogues concerning Natural Religion. Cicero’s entire discussion takes off from the question of moral atheism and its implications for religious practice. Thus ‘the crux and center of the argument is the question whether the gods do nothing, care for nothing, and take their ease detached from all concern with the care and government of the world . . . / There are and there have always been some philosophers who believe that the gods have no concern whatever with the affairs of men. But if this belief is true, what becomes of piety, of reverence and of religion?’ The worry is that ‘there can be no divine guidance of human affairs if the gods make no distinction between good and evil’ (Marcus Tullius Cicero, The Nature of the Gods, tr. Horace C. P. McGregor (London: Penguin, 1972), 69–70, 230). On Hume’s familiarity with and borrowings from De Natura Deorum, see Norman Kemp Smith, ‘Introduction’ to David Hume, Dialogues Concerning Natural Religion, ed. Norman Kemp Smith (Indianapolis: Bobbs-Merrill, 1947), 1–75: 60–1; Peter Jones, Hume’s Sentiments: Their Ciceronian and French Context (Edinburgh: Edinburgh University Press, 1982), 29–40; Peter S. Fosl, ‘Doubt and Divinity: Cicero’s Inﬂuence on Hume’s Religious Skepticism,’ Hume Studies 20 (1994), 103–120; Christine Battersby, ‘The Dialogues as Original Imitation: Cicero and the Nature of Hume’s Skepticism,’ in Norton, Capaldi, and Robison (eds), McGill Hume Studies, 239–52.

hume ’s moral atheism

15

of Hume as any sort of moral atheist. According to the ﬁrst (which I address in the current section), Hume never explicitly endorses moral atheism in any of his writings, and so my proposed interpretation lacks an appropriate textual basis. According to the second (which I address in Chapter 2 below), Hume could not possibly endorse moral atheism given his own well-known skepticism regarding all theological speculation. Each of these objections raises large-scale strategic and methodological questions about the interpretation of Hume that are best addressed at this preliminary stage. The ﬁrst objection is that Hume never openly endorses moral atheism in any of his various writings, and so positive textual evidence for this reading is quite lacking. Indeed, one might go further, for in two texts Hume seems to afﬁrm a perfectly contrary view, openly crediting the deity with a morally praiseworthy character. First, in the Dialogues, immediately following his relational characterization of the meaning of ‘God,’ Hume’s spokesman Philo asserts that he ‘piously ascribe[s] to [this being] every species of perfection’ (DNR 2.3). That sounds as if it might include a positive moral character, and in the closing part of the Dialogues Philo makes the connection explicit, apparently ruling out moral atheism of all stripes: ‘as the Supreme Being is allowed to be absolutely and entirely perfect, whatever differs most from him departs the farthest from the supreme standard of rectitude and perfection’ (DNR 12.8). Second, in his private correspondence with his close friend William Mure, Hume makes the following admission: ‘It must be acknowledg’d that Nature has given us a strong Passion of Admiration for whatever is excellent, & of Love & Gratitude for whatever is benevolent and beneﬁcial, & that the Deity possesses these Attributes in the highest Perfection’ (L i. 51). Thus we have two texts suggesting that the deity has a morally laudable character, and no text that expressly afﬁrms moral atheism. So the proposed interpretation of Hume as some sort of moral atheist can look like a non-starter. But any student of Hume’s philosophy of religion knows that this is a little too quick. Hume often presents his irreligious views in an oblique manner, and his real position sometimes needs to be pieced together both from what is said and what is left unsaid. Moreover, Hume regularly lays down a smokescreen of pious language that is quite insincere. Given the very real possibility of social ostracism,

16 hume ’s moral atheism blackballing, or even prosecution for blasphemy, the motivation for this layering of messages is most obviously prudential, though Hume clearly also enjoys its ironic possibilities. On some occasions he urbanely deadpans his own irreligious views in the language of ﬁdeism and holy mystery.²⁹ At other times, Hume piously appeals to one aspect of orthodox religion while attacking another, only to turn the tables in another context.³⁰ Such faux-pious indirection is quite in keeping with the literary and social conventions of Hume’s day, and was well understood by his eighteenth-century audience: it is not so much that the irreligious implications of his writings are particularly esoteric or well-hidden (the camouﬂage is often indeed ‘ﬂamboyantly disingenuous’³¹), but the respectable language permits at least a modicum of plausible deniability should it be required.³² In these respects, Hume’s case for moral atheism is really no different from his other more familiar irreligious arguments, including his attack on the credibility of miracles, or his critique of the argument from design. Understanding his implied moral atheism thus requires a willingness to follow out his arguments where they lead, and a critical attitude toward the surface pieties of his texts. In addition, some work needs ²⁹ Most notoriously, having argued that there is insufﬁcient historical evidence to justify belief in the miracles of scripture, Hume blithely asserts that ‘the christian religion not only was at ﬁrst attended with miracles, but even at this day cannot be believed by any reasonable person without one. Mere reason is insufﬁcient to convince of its veracity: And whoever is moved by Faith to assent to it, is conscious of a continued miracle in his own person, which subverts all the principles of his understanding, and gives him a determination to believe what is most contrary to custom and experience’ (EHU 10.41). ³⁰ Consider Hume’s avowed deference to revelation when attacking the credentials of natural theology in the Dialogues (DNR 12.33, see also LG 25), or here in ‘Of the Immortality of the Soul’: ‘Nothing could set in a fuller light the inﬁnite obligations which mankind have to divine revelation; since we ﬁnd, that no other medium could ascertain this great and important truth [i.e. the doctrine of immortality]’ (E 598). Such deference is of course belied by Hume’s attack on revealed religion in ‘Of Miracles’ in the ﬁrst Enquiry (EHU 10). ³¹ George Botterill, ‘Hume on Liberty and Necessity,’ in Peter Millican, ed., Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 277–300: 289. ³² For discussion of these literary and social conventions, see David Berman, ‘David Hume and the Suppression of ‘‘Atheism’’,’ Journal of the History of Philosophy 21 (1983), 375–87, and ‘Deism, Immortality, and the Art of Theological Lying,’ in J. A. Leo Lemay (ed.), Deism, Masonry, and the Enlightenment (London and Toronto: Associated University Presses, 1987), 61–78; Isabel Rivers, Reason, Grace and Sentiment, ii. 31–50. For the early reception of Hume’s philosophy in the light of these conventions, see James Fieser, ‘Hume’s Concealed Attack on Religion and his Early Critics,’ Journal of Philosophical Research, 20 (1995), 83–101.

hume ’s moral atheism

17

to be done to see his various writings as providing the materials for an integrated case for moral atheism, for Hume never gathers all of these together in one place. But Hume’s works do provide such a case, and his correspondence shows that he was quite aware of this—or so at least I shall argue. Of course, Hume’s apparent pieties cannot be simply ignored. They place a burden of proof on any irreligious interpretation of his philosophy, and even if they are to be ultimately discounted, still they must be explained. But in fact the texts where Hume appears to grant that God has moral attributes all ﬁt the model of pious indirection to a tee. First, Philo’s claim that he ‘piously ascribe[s] to [God] every species of perfection’ is immediately punctured by the following declaration: ‘But as all perfection is entirely relative, we ought never to imagine that we comprehend the attributes of this divine Being, or to suppose that his perfections have any analogy or likeness to the perfections of a human creature’ (DNR 2.3). The original pious ascription is thus rendered quite meaningless. The same point holds for Philo’s heavily ironic reference to the deity as providing a ‘supreme standard of rectitude and perfection’—for, as Philo himself immediately emphasizes, this is quite consistent with the deity’s ‘benevolence’ and ‘justice’ bearing no meaningful analogy to our own conception of these things (DNR 12.8; see also 11.16). As for Hume’s admission in the letter to Mure that God is ‘benevolent and beneﬁcent . . . in the highest Perfection’: either we similarly read this (in the words of the ﬁrst Enquiry) as so much meaningless ‘ﬂattery and panegyric’ (EHU 11.27; compare also DNR 4.1), or the claim that we know any such thing is quite contradicted by the actual arguments of the Enquiry and Dialogues. And in fact this admission has the place of a concession made simply for the sake of the argument, for Hume immediately proceeds to his main point in the letter: his argument that, even if God is benevolent, we still cannot have positive feelings toward him, for ‘he is not the natural Object of any Passion or Affection’ (L i. 51).³³ ³³ Furthermore, while this declaration of God’s benevolence appears in a letter written to Hume’s conﬁdante Mure, it occurs in a section of the letter providing comments on William Leechman’s recently published sermon On the Nature, Reasonableness, and Advantages of Prayer (Glasgow: 1743), and these comments were intended to be passed

18 hume ’s moral atheism In the following chapter I address the second big-picture objection to the characterization of Hume as some sort of moral atheist: the charge that he could not indulge in such negative dogmatics without contradicting his own pervasive skepticism toward all ‘airy sciences’ (EHU 1.12), and in particular his own powerful epistemological critique of theological and cosmological speculation. along to Leechman, Mure’s ex-tutor. Leechman was soon to be Professor of Divinity and eventually Principal at Glasgow University, and is just the sort of establishment ﬁgure before whom Hume would have been well advised to dilute his irreligious views somewhat. In fact, Hume should probably have been more guarded still in this letter, for its claim that God is not a natural object of human affective attitudes is clearly highly subversive—and the irreverent tone and irreligious substance of the letter cannot have helped when Leechman subsequently decided to join Francis Hutcheson in opposing Hume’s candidacy for academic positions at Edinburgh and Glasgow. For details of the exchange with Leechman, see E. C. Mossner, The Life of David Hume, 2nd edn (Oxford: Clarendon Press, 1980), 148–9. (I discuss the argument of this letter in Chapter 3.)

2 Mitigated Skepticism and Hume’s Liminal Natural Theology 2.1. Introduction In the previous chapter I ﬁxed some important points of terminology, introduced my interpretive thesis that Hume is a strong moral atheist, and addressed a possible objection from the lack of explicit textual evidence conﬁrming this interpretation. But there is a second bigpicture objection that I also need to preempt before turning to the speciﬁcs of Hume’s case for divine amorality. While the previous objection emphasized Hume’s various protestations of piety, this new objection focuses our attention on the deeper currents of skepticism that run beneath such conventional disclaimers. The charge will be that Hume cannot endorse moral atheism without contradicting his own underlying skepticism concerning all theological and cosmological speculation: he could be a moral agnostic, perhaps, but not a moral atheist. According to my interpretative proposal, Hume is a strong moral atheist: he holds that if there is a cause or designer of the ordered universe, this being or principle is not morally assessable. Like a stone or a snowstorm, the deity is neither praiseworthy nor blameworthy; it is simply not a proper object of moral assessment one way or the other. But this amounts to a speculative theory about the nature of the divinity, or so at least it appears. After all, we say that stones and snowstorms are not proper objects of moral assessment only because of what we know about the nature of stones and snowstorms—for

20 mitigated skepticism and hume , s natural theology instance, that they lack intentional agency, a necessary precondition for moral praise and blame to be appropriate. And moral atheism, it seems, must be based on analogous sorts of claims about the divine nature. But (the current objection reminds us) the clear tendency of Hume’s philosophy of religion is to rule out all speculation about the divine attributes as straying quite beyond the reach of our faculties. Moral atheism, whether weak or strong, can thus appear quite inconsistent with Hume’s usual rejection of speculative theorizing about the original cause of all. This objection raises questions about the relationship between Hume’s skeptical epistemology and his critique of natural theology that are as fundamental as any for the systematic interpretation of his philosophy of religion. These questions will loom over each of the particular arguments for moral atheism that I examine in the subsequent chapters. In addition, they also provide much of the deeper subtext and philosophical substance behind some of the most familiar interpretive controversies over Hume’s philosophy of religion. (For instance: is Hume a total skeptic about all religious questions, or is he perhaps some sort of deist? Or: what should we make of Philo’s so-called ‘reversal’ and apparent concessions to the advocate of the design hypothesis at the end of the Dialogues?) It is well, then, to tackle these questions head-on at this preliminary stage. The chapter proceeds as follows. Having ﬁrst set out the objection from Hume’s skepticism in a little more detail (Section 2.2), I then show that, for all his skeptical critique of cosmological and theological speculation, Hume does engage in some forms of argument concerning the character of the ﬁrst cause of all, and advances at least some nonskeptical conclusions about the divine nature (Section 2.3). The apparent tension between Hume’s avowed skepticism and his actual argumentation can be resolved, I argue, by way of a distinction between two kinds of natural theology: a ‘core’ form of natural theology that aims at knowledge of the deity’s distinctive intrinsic character, and a ‘liminal’ form of natural theology that stops short of such ambitions. Hume’s skepticism rules out core natural theology; but the liminal sort of theological speculation need not violate his epistemological strictures, as I show in a series of examples taken from Hume’s own texts (Section 2.4). Since these examples include, inter alia, each of Hume’s arguments for moral atheism, this section

mitigated skepticism and hume , s natural theology

21

also provides a preview of his overall strategy for establishing divine amorality. With the objection from Hume’s skepticism cleared and a prospectus of his case for moral atheism in hand, we will then be ready to proceed (in the following chapters) to the details of the arguments.

2.2. The Objection from Hume’s Skepticism In the Treatise of Human Nature (1739–40) Hume’s skeptical critique of theological and cosmological speculation is largely left between the lines,¹ but it is foregrounded quite clearly in the Enquiry concerning Human Understanding (1748) and the posthumous Dialogues concerning Natural Religion (1779). The Enquiry serves as a particularly useful text for our purposes. Its publication inaugurates the crucial tenyear period in which Hume produced all his major works bearing directly on religion: in addition to the two Enquiries (1748, 1751), an early ﬁrst draft of the Dialogues (part of which Hume sent to his friend Gilbert Elliot in 1751), the caustic Stuart volumes of the History of England (1754, 1756), the essays ‘Of the Immortality of the Soul’ and ‘Of Suicide’ (which were prepared for publication in 1756, but then suppressed), and the Natural History of Religion (1757). The Enquiry also provides the deﬁnitive statement of Hume’s ‘philosophical sentiments and principles’—or so at least he declares in the famous ‘Advertisement’ drawn up in 1775²—and it presents us with his most forthright attack on those ‘airy sciences’ that would take us beyond the modest reach of human understanding (EHU 1.12). In its concluding crescendo, the ﬁnal section of the Enquiry (‘Of the Academical or Sceptical Philosophy’) draws on the critique of human understanding developed throughout that work to present a manifesto for an epistemological attitude Hume labels ‘mitigated scepticism.’ This stance encompasses two separate ‘species’ or forms of skeptical attitude, ¹ See Paul Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion (New York: Oxford University Press, 2008) for a full account of the extensive, though mostly implicit, irreligious ramiﬁcations of the Treatise. ² David Hume, An Enquiry concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000), 1. (On the circumstances and history of the ‘Advertisement,’ see Beauchamp’s editorial comments at xiv–xv, and annotations at 125.)

22 mitigated skepticism and hume , s natural theology each of which Hume commends, and each of which he claims is likely to take hold of those who reﬂect on the numerous inﬁrmities and limitations of our cognitive powers (EHU 12.24–5). The ﬁrst species of mitigated skepticism applies to all domains of human enquiry, and simply urges a moderate and fallibilistic attitude to one’s own beliefs: although some opinions may certainly be preferred to others, all are revisable, and ‘there is a degree of doubt, and caution, and modesty, which in all kinds of scrutiny and decision, ought forever to accompany a just reasoner’ (EHU 12.24). The second species of mitigated skepticism, by constrast, does draw a distinction between different domains of enquiry. It urges us to refrain from speculating in those particular domains that are beyond the reach of our faculties, and counsels instead ‘the limitation of our enquiries to such subjects as are best adapted to the narrow capacity of human understanding.’ It is this second form of mitigated skepticism that is our real concern here, since it might seem to rule out all speculation about the attributes of the ﬁrst cause of all. As Hume puts it, once his sobering reevaluation of our cognitive powers is taken onboard, ‘A correct Judgment . . . [will avoid] all distant and high enquires, [and conﬁne] itself to common life, and to such subjects as fall under daily practice and experience.’ Properly chastened, ‘[t]hose who have a propensity to philosophy’ will accept this humbling contraction of philosophy’s traditional remit and will never be tempted to go beyond common life, so long as they consider the imperfection of those faculties which they employ, their narrow reach, and their inaccurate operations. While we cannot give a satisfactory reason, why we believe, after a thousand experiments, that a stone will fall, or ﬁre burn; can we ever satisfy ourselves concerning any determination, which we may form, with regard to the origin of worlds, and the situation of nature to, and from eternity? (EHU 12.25)

Hume admits that humans will be perennially tempted toward this sort of cosmological speculation, for the imagination is naturally ‘delighted with whatever is remote and extraordinary’ (EHU 12.25, see also T 1.4.7.13). But with the faculties that we have, there is no hope of establishing any real knowledge beyond the sphere of everyday human experience—and this will exclude not only matters ‘concerning the origin of worlds,’ but also ‘the œconomy of the intellectual system or region of the spirits.’ All such metaphysical

mitigated skepticism and hume , s natural theology

23

exotica ‘lie entirely beyond the reach of human capacity’ (EHU 8.1). The ‘mitigation’ of this second form of mitigated skepticism thus appears to refer merely to its limited scope—to the fact that it disallows only ‘distant and high enquiries,’ while permitting reasoning concerning common life—rather than to any dilution of the force of its doubts within the sphere of cosmological speculation.³ Right at the outset of the Dialogues it is Philo’s defense of this same species of epistemic humility that marks him most clearly as Hume’s spokesman. In the Enquiry Hume had asserted that our faculties are better suited to the world of ‘daily practice and experience’ (‘common life,’ in the phrase he employs repeatedly in both the Enquiry and Dialogues) rather than remote cosmological speculation. And similarly while Philo is prepared to offer a qualiﬁed defense of philosophical reasoning in common life (DNR 1.9), this does not extend to cosmological and theological topics: when we look beyond human affairs and the properties of surrounding bodies: When we carry our speculations into the two eternities, before and after the present state of things; into the creation and formation of the universe; the existence and properties of spirits; the powers and operations of one universal spirit existing without beginning and without end; omnipotent, omniscient, immutable, inﬁnite, and incomprehensible: We must be far removed from the smallest tendency to skepticism, not to be apprehensive, that we have here got quite beyond the reach of our faculties. (DNR 1.10)

Philo’s general concern here about the limits of cognitive powers is sharpened when he considers the force of skeptical arguments against the speculative ambitions of natural theology. When reasoning within the sphere of common life, we can always appeal to the immediate bedrock evidence of sense and experience, but when we philosophize about transcendental matters we forfeit this sort of insurance against skeptical attack (DNR 1.11). In the end, whatever the constructive ³ For commentary on Hume’s mitigated skepticism, see Stephen Buckle, Hume’s Enlightenment Tract: The Unity and Purpose of An Enquiry concerning Human Understanding (Oxford: Clarendon Press, 2001), 314–22; David Fate Norton, ‘Of the Academical or Sceptical Philosophy,’ in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 371–92; Don Garrett, ‘A Small Tincture of Pyrrhonism: Skepticism and Naturalism in Hume’s Science of Man,’ in Walter Sinnott-Armstrong (ed.), Pyrrhonian Skepticism (Oxford: Oxford University Press, 2004), 68–98; Russell, The Riddle of Hume’s Treatise, 204–222.

24 mitigated skepticism and hume , s natural theology powers of reason may be in the sphere of common life, the skeptic is entirely victorious in the domain of theological speculation: All religious systems, it is confessed, are subject to great and insuperable difﬁculties. Each disputant triumphs in his turn; while he carries on an offensive war, and exposes the absurdities, barbarities, and pernicious tenets of his antagonist. But all of them, on the whole, prepare a complete triumph for the sceptic; who tells them, that no system ought ever to be embraced with regard to such subjects . . . A total suspense of judgment is here our only reasonable resource. And if every attack, as is commonly observed, and no defence, among theologians, is successful; how complete must be his victory, who remains always, with all mankind, on the offensive, and has himself no ﬁxed station or abiding city, which he is ever, on any occasion, obliged to defend? (DNR 8.12, see also 11.5)

In the image Hume borrows from Bayle, and which Kant will famously invoke in the Critique of Pure Reason, the skeptic enjoys a ‘complete triumph’ so long as he remains a rootless and nomadic marauder, refusing to commit to any ‘ﬁxed station or abiding city’—refusing, that is, any afﬁrmative doctrine concerning the divine attributes.⁴ So the problem for the interpretation of Hume as any form of moral atheist is easy to see. Both in propria persona and through his spokesman Philo, Hume appears quite doubtful of our ability to establish any point of substance about the nature of the ﬁrst cause. And if he is really counseling ‘a total suspense of judgment’ concerning the divine attributes, Hume ought to refrain from all speculative commitments regarding the deity’s moral character, positive or negative. A strict agnosticism on this question would seem his only possible position. Indeed, were we to read Hume as arguing for moral atheism, then for all his famous skeptical critique of the program of natural religion he would in fact be engaging in his own (admittedly rather unorthodox) brand of speculative natural theology. But this, it might be argued, cannot be the Hume of the ﬁrst Enquiry and Dialogues. The objection is all the more pressing if we recall that my interpretive proposal is that Hume regards moral atheism as a provable certainty, and not simply a provisional hypothesis. Were we to read Hume as merely ⁴ Pierre Bayle, Historical and Critical Dictionary, ed. and tr. Richard H. Popkin (Indianapolis: Hackett, 1991), 27 (the article on ‘Arriaga,’ note B). Immanuel Kant, Critique of Pure Reason, tr. and ed. Werner S. Pluhar (Indianapolis: Hackett, 1996), A ix.

mitigated skepticism and hume , s natural theology

25

tentatively or provisionally leaning toward moral atheism, this might, perhaps, seem easier to reconcile with his skeptical critique of natural theology. But a reading of Hume as quite afﬁrmatively committed to moral atheism appears plainly inconsistent with the skeptical attitude he recommends toward cosmological and theological speculation.

2.3. Theological Argument in Hume’s Texts The ﬁrst point to make in the face of this objection is that, for all Hume’s strong pronouncements concerning the overreaching character of natural theology, in practice he does engage in at least some forms of reasoning about the divine nature. When addressing the traditional arguments of natural theology, for instance, he does not simply dismiss the whole program from the outset, reiterating a global epistemological critique of all speculation beyond the sphere of common life. Rather, he enters into a sustained debate concerning the merits and demerits of the various particular arguments. Of course, Hume’s assessment of these arguments is largely negative: the classic arguments are either invalid or sustain much weaker conclusions than had traditionally been thought. But he only reaches these critical conclusions after examining the arguments in detail—considering, for instance, whether an inﬁnite past needs explanation in terms of an external cause (DNR 9.9); whether the order of the world can be traced to a designing intelligence without introducing a regress of questions about the origins of the order in that designing mind (DNR 4.7, 4.9); and whether correct analogical reasoning would permit us to conceive of the original cause as some sort of biological principle of generation or vegetation before it would permit us to conceive of it as a rational principle of mind or intelligence (DNR 7.3–8). In fact Hume had explicitly endorsed the propriety of at least some limited reasoning on theological topics in a letter to his close friend Gilbert Elliot of 18 February 1751 (just a couple of weeks before sending him an early draft sample of the Dialogues), writing that in Politics & natural Philosophy, whatever Conclusion is contrary to certain Matter of Fact must certainly be wrong, and there must some Error lie somewhere in the Argument, whether we be able to show it or not. But in

26 mitigated skepticism and hume , s natural theology Metaphysics or Theology, I cannot see how either of these plain & obvious Standards of Truth can have place. Nothing there can correct bad Reasoning but good Reasoning: and Sophistry must be oppos’d by Syllogism. (L i. 151, emphasis mine)

So Hume cannot mean to preclude literally all philosophical theorizing and argumentation beyond the sphere of common life, or (more speciﬁcally) to preclude all argumentation concerning the original cause of all. Of course, it is one thing to enter into a philosophical debate that touches on the original cause of all, and quite another to promote speciﬁc conclusions concerning the nature of this ultimate being or principle. So while Hume does engage in some forms of reasoning concerning the ﬁrst cause, still it might be suggested that he holds to his strict skeptical principles in refusing to endorse any particular view of its attributes. Most likely (it might be said) he is discussing the original cause simply in order to refute the various traditional proofs regarding its nature, thereby conﬁrming our inability to establish any genuine knowledge concerning the attributes of this ultimate being or principle. Interpreted this way, Hume is indulging the natural theologian’s desire to speculate about cosmological questions beyond common life, at least up to a point, but purely in order to embarrass natural theology on its own terms.⁵ And on this view, the apparent tension between Hume’s mitigated skepticism and his discussion of cosmological questions far beyond the sphere of common life might seem to evaporate. He is only entertaining such questions in order to enforce skeptical conclusions, and so he is not so much neglecting the strictures of mitigated skepticism as he is vindicating them when he engages in philosophical argument concerning the ﬁrst cause of all. There is no doubt that this picture captures what is often going on when Hume is examining theological and cosmological issues beyond the sphere of common life. But it cannot be the complete story, which brings me to my second point. Contrary to the preceding proposal, Hume does not engage with theological and cosmological ⁵ At one point in the Dialogues, Philo explicitly states that his speculative reasonings should be understood as merely ad hominem, at least in that speciﬁc place, where he is simply ‘argu[ing] with Cleanthes in his own way [to show] him the dangerous consequences of his tenets’ (DNR 2.11).

mitigated skepticism and hume , s natural theology

27

topics simply in order to draw out object lessons in skepticism and the limits of human understanding. On occasion he also positively endorses speciﬁc conclusions about the divine attributes—conclusions that deﬁne or delimit the divine attributes in one way or another, and not simply skeptical conclusions that place the divine attributes beyond human comprehension. I provide examples in the next section, where I document three kinds of conclusion characterizing the original cause, each of which Hume allows. But for the moment one particularly well-known case will illustrate the point. At the close of the Dialogues, Hume concedes (through his spokesman Philo) that ‘the cause or causes of order in the universe probably bear some remote analogy to human intelligence’ (DNR 12.33). As he immediately makes plain, this is a very meager concession. It is largely vitiated by the fact that there is ‘a certain degree of analogy among all the operations of nature’ (DNR 12.7), and is quite insufﬁcient to license any ‘inference that affects human life, or can be the source of any action or forbearance’ (DNR 12.33). It is certainly of no real use to the partisans of Christianity or other traditional theistic systems. But none of this tells against the point that Hume is afﬁrming some sort of positive thesis (however meager or inconsequential) about the likely properties of the ﬁrst cause of all.⁶ There is no point in being coy about it: Hume is advancing some sort of natural theology here. So there must be some way of reconciling Hume’s mitigated skepticism with this foray into theological speculation—unless, of course, he is simply forgetting the strictures of his own epistemology, and consistency in this part of his philosophical system is but a will-o’-the-wisp. In the remainder of this chapter I document the various sorts of cases where Hume advances non-skeptical conclusions about the divine attributes, and thereby map the outlines of the modest sort of natural theology that Hume does in fact permit. I also work back from these speciﬁc cases to reverse-engineer the general distinction, implicit in Hume’s practice, between the (‘liminal’) sort of natural theology that he allows, and the (‘core’) sort of natural theology that he totally rejects. What is needed here is an interpretation that can accommodate both Hume’s actual theological argumentation and his repeated disavowals of speculation beyond the sphere of common life. ⁶ I return to the interpretation of this concession in Section 4.2 below.

28 mitigated skepticism and hume , s natural theology It will not do simply to emphasize one of these aspects of Hume’s philosophy, as is common enough in the literature, while neglecting or ignoring the other.

2.4. Hume’s Liminal Natural Theology How then is Hume’s practice to be reconciled with his skeptical doubts concerning cosmological and theological speculation? In this section I identify three different kinds of conclusion concerning the divine nature that Hume does permit, and (I argue) can permit without violating the spirit of mitigated skepticism. The ﬁrst and third categories here will prove particularly important for our purposes, since each of them provides Hume with an opening to advance a different sort of argument for moral atheism. So here I will be mapping three forms of liminal natural theology that Hume does in fact allow, and along the way I will also be previewing Hume’s case for the amorality of any original cause or designer. Before I begin, I should stress that I have no interest in either explaining away or discounting Hume’s commitment to mitigated skepticism. At least by the period of the Enquiry and Dialogues, Hume clearly is a mitigated skeptic, as is his spokesman Philo: section 12 of the Enquiry and part 1 of the Dialogues put this past all reasonable doubt.⁷ Moreover, this mitigated skepticism has real teeth. As I will argue, it completely rules out the traditional program of (what I will call) core natural theology: the program of employing natural reason to work our way to species-speciﬁc knowledge of the intrinsic character of the original cause—knowledge that would ﬁll out our conception of the deity’s intrinsic nature in some distinctive way, enabling us to afﬁrm things of it that we could not equally afﬁrm of any other unknown being. For Hume, the empirical evidence ⁷ Various commentators have argued that Hume also endorses mitigated skepticism (without using the phrase) in the earlier Treatise. See for instance, Don Garrett, ‘Reasons to act and believe: naturalism and rational justiﬁcation in Hume’s philosophical project,’ Philosophical Studies 132 (2007), 1–16: 3; Russell, The Riddle of Hume’s Treatise, 215–17; Miriam McCormick, ‘A Change in the Manner: Hume’s Scepticism in the Treatise and the First Enquiry,’ Canadian Journal of Philosophy 29 (1999), 431–48. But we need not broach this larger question here.

mitigated skepticism and hume , s natural theology

29

that is available to our human faculties—which is the only sort of evidence that can help us with matters of fact and existence (EHU 4.1–13, 12.28–9)—is quite insufﬁcient to establish any such species-speciﬁc knowledge of the deity’s intrinsic character. In this sense, the mitigated skeptic regards the ﬁrst cause of all as permanent terra incognita—and it is this result that is borne out in devastating detail in the Dialogues, where Hume systematically identiﬁes the various overreaching premises and invalid inferences that compromise the traditional arguments of core natural theology. So the challenge that mitigated skepticism poses to traditional natural theology goes far—farther, in fact, than is sometimes appreciated. For instance, it will rule out not only the sort of ‘experimental theism’ (DNR 5.2) that would carry us from premises about the order and adjustment of empirical phenomena to the conclusion that the original cause bears a close and religiously signiﬁcant resemblance to mind or intelligence, but also the sort of ‘evidential’ argument from evil that would carry us to the conclusion that the original cause is most likely indifferent to human suffering.⁸ Nevertheless, Hume’s mitigated skepticism does not rule out all speculative conclusions about the divine attributes, as I now argue.

2.4.1. General Arguments Concerning Intrinsic Properties The ﬁrst kind of conclusion that Hume is willing to allow concerns the intrinsic properties of the original cause of all. I just claimed that Hume’s mitigated skepticism rules out any species-speciﬁc knowledge of the deity’s intrinsic properties—that is, knowledge of its distinctive intrinsic character, knowledge that would enable us to afﬁrm things of it that we could not equally afﬁrm of any other unknown being. But this still leaves open the possibility of our being able to draw certain highly general conclusions about the intrinsic character of any being or principle, and thus, inter alia, about the intrinsic character of any original cause of all. In the ﬁrst place, we can know in advance that the intrinsic properties of the original cause will not involve any conceptual impossibilities or (Hume might say) ‘repugnancies’: contradictions ⁸ This latter argument is often attributed to Hume himself, but I ﬁnd this reading unpersuasive. I examine this interpretive controversy in Section 6.4.

30 mitigated skepticism and hume , s natural theology that can be shown merely by inspecting the relations among our own ideas. We can know in advance that the ultimate cause of all is not both square and circular, since we see by reﬂection on the relations among our ideas that nothing could be both square and circular. And not all such aprioristic conceptual reasoning is as harmless for traditional theological views as the prohibition on square circles. For instance, Hume’s character Cleanthes insists that there is a conceptual incoherence in the idea of a mind that is totally immutable, thereby placing a limitation on how we can conceive of the ﬁrst cause: [T]hough it be allowed, that the deity possesses attributes, of which we have no comprehension; yet ought we never to ascribe to him any attributes, which are absolutely incompatible with that intelligent nature, essential to him. A mind, whose acts and sentiments and ideas are not distinct and successive; one, that is wholly simple, and totally immutable; is a mind which has no thought, no reason, no will, no sentiment, no love, no hatred; or in a word, is no mind at all. It is an abuse of terms to give it such an appellation; and we may as well speak of limited extension without ﬁgure, or of number without composition. (DNR 4.3; compare also NHR 6.5)

Since Cleanthes holds that the original cause of all is some sort of mind, he cannot then allow that this ultimate being or principle is immutable. But Hume’s own point is surely that, without committing to the positive view that the ﬁrst cause is in fact a mind, we can say with certainty that it is not both a mind and immutable—and this is an a priori knowable limitation on the intrinsic character of any ﬁrst cause. Consider a second example. In part 9 of the Dialogues, Cleanthes argues, against Demea’s a priori proof of a ‘necessarily existent being, who carries the reason of his existence in himself’ (DNR 9.3), that the notion of a necessarily existent being is conceptually incoherent. [T]here is an evident absurdity in pretending to demonstrate a matter of fact, or to prove it by any arguments a priori. Nothing is demonstrable, unless the contrary implies a contradiction. Nothing, that is distinctly conceivable, implies a contradiction. Whatever we conceive as existent, we can also conceive as non-existent. There is no being, therefore, whose non-existence implies a contradiction. . . . It is pretended that the Deity is a necessarily existent being; and this necessity of his existence is attempted to be explained by asserting, that, if we knew his whole essence or nature, we should perceive it to be as impossible for him not to exist, as for twice two not to be four. But it is evident, that this

mitigated skepticism and hume , s natural theology

31

can never happen, while our faculties remain the same as at present. It will still be possible for us, at any time, to conceive the non-existence of what we formerly conceived to exist; nor can the mind ever lie under a necessity of supposing any object to remain always in being; in the same manner as we lie under a necessity of always conceiving twice two to be four. The words, therefore, necessary existence, have no meaning; or, which is the same thing, none that is consistent. (DNR 9.5–6)

Although this comes from Cleanthes rather than Hume’s spokesman Philo, it is reasonable in this case to attribute Cleanthes’s views to Hume himself. At this stage of the Dialogues Cleanthes and Philo are jointly attacking Demea’s ‘argument a priori’ (DNR 9.1), and Philo seems quite content to stand by and let Cleanthes make these preliminary points for the both of them.⁹ Moreover, the charge that there is no coherent concept answering to the words ‘necessary existence’ has impeccable Humean credentials both in the Treatise (T 1.3.7.3) and, most vividly, in the ﬁrst Enquiry: Whatever is may not be. No negation of a fact can involve a contradiction. The non-existence of any being, without exception, is as clear and distinct an idea as its existence. The proposition, which afﬁrms it not to be, however false, is no less conceivable and intelligible, than that which afﬁrms it to be. . . . [T]hat CÆSAR, or the angel GABRIEL, or any being never existed, may be a false proposition, but still is perfectly conceivable, and implies no contradiction. (EHU 12.28)

So here we have a second a priori knowable constraint on the intrinsic character of the deity. Just as we can know in advance that the original cause of all cannot be both immutable and mind-like, so we can know in advance that it cannot have the property of necessary existence. And as these two examples clearly show, we can say these things about the intrinsic character of the divinity, not because we have any particular knowledge of its species-speciﬁc character, but rather because these are completely general results, demonstrable by a priori conceptual argument, that will apply to any being whatsoever. Moreover, as our two examples also show, the fact that these conclusions are driven purely by conceptual reﬂection ⁹ Thus Cleanthes prefaces his attack on Demea’s a priori argument with the comment that ‘I shall not leave it to Philo, . . . (though I know that the starting objections is his chief delight), to point out the weakness of this metaphysical reasoning.’ (DNR 9.4)

32 mitigated skepticism and hume , s natural theology on the relations among our ideas does not mean that they will be obvious or uncontroversial, or that they pose no danger to traditional theological systems. Here then is one liminal form of natural theology that Hume does practice: an aprioristic natural theology of the via negativa stripe, maintaining that the ﬁrst cause simply cannot have certain combinations of properties.¹⁰ In addition to conclusions identifying conceptually guaranteed constraints on the intrinsic character of any given being (and hence, inter alia, on the intrinsic character of the deity), Hume is also prepared to admit certain conclusions identifying epistemically probable constraints on the intrinsic character of any given unknown being (and hence, inter alia, on the intrinsic character of the deity). Of course, no conceptually consistent combination of properties can be entirely ruled out in advance, and if we are to speculate about the character of an unknown original cause of all, we must always remember Hume’s maxim that ‘to consider the matter a priori, any thing may produce any thing’ (T 1.4.5.30; see also T 1.3.15.1, 1.4.5.32, EHU 12.29). But conceptual possibility is one thing and likelihood is another, and Hume does offer some tentative probabilistic judgments delimiting the likely intrinsic character of any unknown being, including, inter alia, the original cause of all. On some occasions, Hume appeals to the patterns and regularities observed in empirical phenomena in order to support extrapolative judgments concerning the probable character of any unknown being that is picked out in advance. Of course, we will need to be quite difﬁdent when projecting locally observed regularities into completely different spheres of existence. This is one of the central themes of Philo’s critique of Cleanthes’s ‘experimental theism’ (DNR 5.2), and one of the chief reasons why he generally counsels the limitation of our enquiries to the sphere of common life (DNR 2.22). But be that as it may, Hume does offer some provisional judgments (through Philo), ¹⁰ For a rich example of this aprioristic via negativa theorizing from more recent times, see Anthony Kenny, The God of the Philosophers (Oxford: Clarendon Press, 1979). Kenny argues, for instance, that there is a contradiction in the traditional conception of a god that is both omniscient and immutable (48, 121); that there is a contradiction in the notion of a god ‘who infallibly knows future free actions, and yet is not the author of sin’ (87, see also 121); and further that there may be an incoherence in the traditional notion of the deity as a form of disembodied intelligence (123–7).

mitigated skepticism and hume , s natural theology

33

based on analogical reasoning from empirically observed phenomena, asserting that certain combinations of properties are most unlikely to characterize the ﬁrst cause of all, apparently because these combinations of properties are unlikely to characterize any given unknown being. Thus Philo stresses to Cleanthes that if the ﬁrst cause of the universe is to be conceptualized as a mind or intelligence, we would expect it to share certain properties with all observed minds: In all instances which we have ever seen, ideas are copied from real objects, and are ectypal, not archetypal, to express myself in learned terms: You reverse this order, and give thought the precedence. In all instances which we have ever seen, thought has no inﬂuence upon matter, except where that matter is so conjoined with it as to have an equal reciprocal inﬂuence upon it. No animal can move immediately any thing but the members of its own body; and indeed, the equality of action and reaction seems to be an universal law of nature: but your theory implies a contradiction to this experience. (DNR 8.11, compare also DNR 6.5, 6.6)

Hume is not here claiming that he knows that a mind without these typical features is altogether impossible. (Indeed, this passage immediately leads into a strong statement of skepticism eschewing all theological speculation, the paragraph announcing the ‘complete triumph’ of the nomad-skeptic who holds that ‘no system ought ever to be embraced with regard to such subjects’ (DNR 8.12).) But in the current passage he does seem to be indicating that, prima facie, the deity is unlikely to be both some sort of mind, and yet, unlike all experienced minds, capable of acting without a body, or in deﬁance of ‘the equality of action and reaction,’ and so on. And again the underlying reasoning here is entirely general: given what we know from experience about the correlations between having a mind and being embodied, it is improbable (epistemically speaking) that any unknown being we might pick out will be both mind-like and yet disembodied. So perhaps Hume does not mean to entirely reject this experientially grounded form of analogical reasoning, but only to preclude the more conﬁdent and overreaching uses of it. The particular example of DNR 8.11 might be contested. Perhaps (it might be argued) this is simply one of those occasions where Philo is adopting the traditional natural theologian’s own methodology merely for the sake of the argument, and that he is allowing that analogical

34 mitigated skepticism and hume , s natural theology reasoning from experience can support conclusions about the original cause of all purely in order to subvert experimental theism on its own terms. However, while DNR 8.11 is in some ways an ambiguous case, there are other occasions where Philo seems to be plainly sincere in appealing to experience to bolster probabilistic conjectures about the deity’s intrinsic nature. For instance, in his peroration toward the end of the Dialogues lamenting the pernicious consequences of ‘popular religion’ and ‘vulgar superstition’ (DNR 12.11–23, 12.25–32), Philo announces that It is an absurdity to believe that the Deity has human passions, and one of the lowest of human passions, a restless appetite for applause. (DNR 12.31)

Philo surely intends this sincerely. But it is hardly a conceptual truth: he cannot mean to suggest that there is a conceptual absurdity or internal contradiction in the proposition that the deity has human passions.¹¹ Rather, Philo must mean that it is absurd to believe this proposition on account of its inordinate improbability. Here in part 12 of the Dialogues, Philo baldly asserts the ‘absurdity’ of this belief without offering any supporting argument. But he is presumably relying on the sort of considerations sketched out previously in part 3 (albeit by another character, Demea, the pious ‘mystic’):¹² All the sentiments of the human mind, gratitude, resentment, love, friendship, approbation, blame, pity, emulation, envy, have a plain reference to the state and situation of man, and are calculated for preserving the existence, and promoting the activity of such a being in such circumstances. It seems, therefore, unreasonable to transfer such sentiments to a supreme existence, or to suppose him actuated by them. (DNR 3.13)

The human passions are connected in all sorts of ways with the particularities of human circumstances and needs, and so it is highly unlikely that the deity will be moved by attitudes remotely like them. Hume had presented a similar line of argument in his own voice ¹¹ Obviously there is no prospect of a general proof of the conceptual impossibility of beings with human passions. And if such beings are conceptually possible, then for all a priori reﬂection on the relations among our ideas can tell us, the original cause of all might be such a being. ‘[F]or aught we can determine by the mere ideas, any thing may be the cause or effect of any thing’ (T 1.4.5.32). ¹² I defend this interpretive claim in Section 5.2.

mitigated skepticism and hume , s natural theology

35

several years earlier in the well-known letter to Francis Hutcheson of March 16, 1740: [N]othing but Experience can assure us, that the Sentiments [of different rational beings] are the same. What Experience have we with regard to superior Beings? How can we ascribe to them any Sentiments at all? They have implanted those Sentiments in us for the conduct of Life like our bodily Sensations, which they possess not themselves. (L i. 39)

Again, Hume takes the empirical fact that human sentimental psychology is adapted to the particular circumstances of human life as some sort of evidence that ‘superior Beings’ are unlikely to share this sentimental psychology with us. So he does seem to be allowing some provisional conclusions about the likely character of higher beings based in part on analogical argument from experience. Moreover, while the arguments set out in DNR 3.13 and the letter to Hutcheson are directed at showing the improbability of the proposition that we share our system of passions with ‘superior Beings’ or ‘a supreme existence,’ the underlying reasoning is, once again, entirely general. It does not presuppose any particular knowledge of the distinctive, species-speciﬁc character of this supreme existence (or of the distinctive, species-speciﬁc character of ‘superior Beings’), but could be applied to any unknown being that is picked out in advance. Hume seems to be assuming that, since our sentimental psychology is adapted to the ﬁne-grained particularities of human life, it is improbable that any given unknown being would share this system of passions with us. I return to this argument in Chapter 5, where I discuss its connection to moral atheism. If Hume is right that it is improbable that the deity has passions like ours, then it will follow (given his moral psychology) that it is improbable that the deity will be motivated to respect moral standards, and indeed improbable that the deity is a proper object of moral assessment. Hume makes the point himself, and in assessing this probabilistic case for strong moral atheism, I will offer a limited defense of his tolerance of probabilistic conclusions about the intrinsic character of unknown beings in general and the unknown ﬁrst cause of all in particular. Certain conceptually consistent combinations of properties are indeed highly unlikely to characterize any unknown being that is picked out in advance, and Hume can

36 mitigated skepticism and hume , s natural theology exploit this fact while continuing to maintain a robust skepticism toward species-speciﬁc speculation concerning the intrinsic character of the deity.

2.4.2. General Arguments Concerning Relational Properties Hume also permits certain sorts of conclusion concerning the relational properties of the original cause of all. Here we can usefully distinguish between two species of conclusion about the deity’s relational properties, each of which Hume allows in some form or other. Once again, there are conclusions that would apply to any given being (and hence, inter alia, to any original cause). But here in the case of relational properties there will also be conclusions that apply peculiarly to the original cause, and thus depend on some sort of distinctive, species-speciﬁc reasoning. First, the conclusions that rest on general reasoning. Since Hume allows that there are both conceptually necessary constraints and epistemically probable constraints on the intrinsic properties of any given being, he clearly must allow that there are some analogous sorts of constraint on the relational properties of any given being. For instance, if he grants that no being could be both immutable and intelligent, he must grant that no being could be both immutable and more intelligent than Aristotle. Likewise, if he grants that any unknown being that is picked out in advance is unlikely to have human passions, he must grant that any unknown being that is picked out in advance is unlikely to have passions more virtuous than Socrates. This way of understanding Hume’s reasoning neatly accommodates Philo’s famous ‘reversal’ at the end of the Dialogues. As is wellknown, while the overall discussion in the Dialogues is largely deﬁned by Philo’s systematic demolition of the argument from design, Philo appears to reverse himself in the closing part of the debate, and suggests that some version of the argument does go through, albeit one that establishes only the rather weak conclusion that ‘the cause or causes of order in the universe probably bear some remote analogy to human intelligence’ (DNR 12.33). This reversal can seem all the more puzzling since many of Philo’s preceding objections to the argument from design appeared quite devastating, and met with no real response. For instance, in part 7 he maintained that if the analogical principles driving the argument

mitigated skepticism and hume , s natural theology

37

were consistently applied, the origin of the universe ‘ought rather to be ascribed to generation or vegetation than to reason or design,’ for the universe ‘plainly resembles more an animal or a vegetable than it does a watch or a knitting loom’ (DNR 7.1, 7.3). Philo argued this point at length and his interlocutors could muster no real response. Thus it is all the more surprising to see him apparently changing tack at the end of the dialogue to accept a version of the design hypothesis. There is much disagreement over the correct interpretation of Philo’s apparent reversal, and no real prospect of resolving the controversy here: I would not want my overall account of Hume’s liminal natural theology considered hostage to this highly contentious issue. Nevertheless, it is worth noting that my proposal squares nicely with what is perhaps the most popular and in my view the most plausible interpretation of Philo’s reversal. On this reading Philo is sincere in granting that there is ‘probably . . . some remote analogy’ between the original cause of all and human intelligence: he is allowing that the deity most likely has the relational property of being somewhat similar to a human mind. But this concession is so weak that it tells us nothing distinctive about the divine attributes. After all, Philo argues earlier in the same section that resemblance and analogy are a matter of degree, and indeed that there is ‘a certain degree of analogy among all the operations of nature.’ Thus ‘the rotting of a turnip, the generation of an animal, and the structure of human thought [are] energies that probably bear some remote analogy to each other’ (DNR 12.7, my emphasis on the phrase echoed subsequently in 12.33). Given how weakly the claim can be taken, then, even an atheist must allow that it is ‘probable, that the principle which ﬁrst arranged, and still maintains order in this universe, bears . . . some remote inconceivable analogy to the other operations of nature, and among the rest, to . . . human mind and thought’ (DNR 12.7, my emphasis). So Philo is sincere in allowing that the original cause ‘probably [bears] some remote analogy’ to a human mind, but the admission is sly and potentially quite misleading. It is not a genuine reversal of his position, for the ‘remote analogy’ he permits may be so weak that it is utterly insigniﬁcant, and quite consistent with the ﬁrst cause bearing a much closer analogy to a principle of vegetation or generation. For our purposes the real point of interest here is that Philo’s ‘reversal’ (on the current interpretation) exploits the fact that a sufﬁciently weak analogy to a human mind

38 mitigated skepticism and hume , s natural theology could be said to hold of just about any being, for any given unknown being will ‘probably bear some remote analogy’ to a human intelligence, so long as we read ‘some remote analogy’ weakly enough. In admitting the likelihood of some remote analogy between the deity and a human mind, Philo is not then claiming any special insight into the distinctive character of the original cause of all—as he himself seems to acknowledge when he immediately goes on to stress that his admission cannot be used to support any further speciﬁc conclusions about the divine nature.¹³

2.4.3. Species-Speciﬁc Arguments Concerning Relational Properties Hume also admits certain species-speciﬁc conclusions concerning the deity’s relational properties. This is possible because he has stipulatively deﬁned the meaning of ‘the deity’ (and ‘God,’ ‘the divinity,’ and other correlate phrases) in relational terms. In Hume’s usage, ‘the deity’ just means ‘the being or principle, whatsoever its intrinsic nature, that produced the universe, or at least produced the order of the universe’—and this is a topic-neutral deﬁnite description that locks onto its referent purely by way of a relational property, picking it out simply in terms of the cause–effect relationship that it bears to the ordered universe.¹⁴ So we know one distinctive, species-speciﬁc fact about the relational properties of the deity (assuming that there is some such original being or principle) purely as a matter of deﬁnition. And with this initial stipulative fact in place, Hume is then in a position to exploit it as a premise in arguments that ascribe further relational properties to any such ﬁrst cause of all. Consider two lines of argument from this starting point, each of which can be found in Hume’s texts, and each of which aims to establish a (different) negative conclusion about the moral status of God. First, Hume is a determinist: he holds that all events in the ¹³ ‘If this proposition be not capable of extension, variation, or more particular explication: if it affords no inference that affects human life, or can be the source of any action or forbearance: And if the analogy, imperfect as it is, can be carried no further than to the human intelligence, and cannot be transferred, with any appearance of probability, to the other qualities of the mind; If this really be the case, what can the most inquisitive, contemplative, and religious man do more than give a plain, philosophical assent to the proposition, as often as it occurs; and believe that the arguments, on which it is established, exceed the objections, which lie against it?’ (DNR 12.33). ¹⁴ On Hume’s relational characterization of the meaning of ‘the deity’, see Section 1.2.

mitigated skepticism and hume , s natural theology

39

world, including all human actions, are entirely the result of prior causes whose operation conforms to exceptionless laws. It follows that if there is a deity, then this being or principle is the remote cause of every human action. In the ﬁrst Enquiry Hume sketches this result in the following terms (he puts these words in the mouth of a critic who is setting up an objection to Hume’s deterministic ‘doctrine of necessity,’ but Hume himself accepts this initial part of the argument): [I]f voluntary actions be subjected to the same laws of necessity with the operations of matter, there is a continued chain of necessary causes, preordained and pre-determined, reaching from the original cause of all, to every single volition of every human creature. . . . The ultimate Author of all our volitions is the Creator of the world, who ﬁrst bestowed motion on this immense machine, and placed all beings in that particular position, whence every subsequent event, by an inevitable necessity, must result. (EHU 8.32)

So God is the remote cause of all our actions. It follows, according to Hume’s assessment in his essays ‘Of Suicide’ and ‘Of the Immortality of the Soul,’ that the original being or principle cannot reasonably be aggrieved or offended by our actions, that we cannot owe this deity any moral duties, and that this ﬁrst cause of all cannot exert any moral claims on its creatures. While determinism is consistent with our making moral claims on one another, it is not (in Hume’s view) consistent with the ﬁrst cause of all making moral claims upon us. I discuss this argument from determinism against duties to God in Chapter 7. It is not an argument for any sort of moral atheism: it says nothing one way or the other about whether the deity is morally praiseworthy. But it will clearly inform the wider question of what overall moral status, if any, the original cause of all could possibly have. Another conclusion might also seem to follow from Hume’s view that the deity is the remote cause of every human action. Since this will make the deity causally responsible for all our vicious behavior, it might seem to entail weak moral atheism, the view that the ﬁrst cause of all is not morally praiseworthy. This at least is what is worrying the imagined critic in EHU 8.32, who frets that, given Hume’s determinism, ‘if [our human actions] have any turpitude, they must involve our Creator in the same guilt, while he is acknowledged to be

40 mitigated skepticism and hume , s natural theology their ultimate cause and author’ (EHU 8.32); and Philo is often taken to be endorsing the same basic argument in the Dialogues (speciﬁcally, in DNR 11.17). I examine Hume’s nuanced handling of this argument from determinism and evil in Chapter 7. My conclusion will be that Hume personally holds that this argument is not sound, but that he is certainly quite ready to exploit it for ad hominem purposes. Setting the issue of determinism to one side, we can turn to another, entirely different line of argument that Hume provides—one that does purport to establish moral atheism, indeed strong moral atheism. Without claiming to know anything about the intrinsic properties of the deity, or even whether such a being exists, we can ask whether human psychology imposes any particular constraints on the way in which our passions—our conative and affective attitudes, including our desires, aversions, and object-directed emotions of love, hatred, gratitude, envy, and the rest—might respond to any original cause or organizing principle responsible for the entire ordered universe. In the context of Hume’s sentimentalist moral theory—his projectivist view that moral properties are painted forth on the world by human sentimental responses—any such psychological limitations on our sentiments may have decisive consequences for the moral status of any ﬁrst cause of all. As a way into this second line of argument, we might begin by considering the introduction to the Treatise and the following manifesto for Hume’s ‘science of man,’ which is the only passage in this early work to address the possibility of natural theology in direct and explicit terms: ’Tis evident, that all the sciences have a relation, greater or less, to human nature; and that however wide any of them may seem to run from it, they still return back by one passage or another. Even Mathematics, Natural Philosophy, and Natural Religion, are in some measure dependent upon the science of man; since they lie under the cognizance of men, and are judg’d of by their powers and faculties. ’Tis impossible to tell what changes and improvements we might make in these sciences were we thoroughly acquainted with the extent and force of human understanding, and cou’d explain the nature of the ideas we employ, and of the operations we perform in our reasonings. (T Introduction 4)

So far this is simply Hume’s familiar claim that his ‘science of man,’ in charting the operations and limits of our mental faculties, will help to

mitigated skepticism and hume , s natural theology

41

establish the basis, legitimacy, and range of all possible human sciences, theology not excepted. It is this approach, of course, that underwrites his skeptical attitude toward transcendental metaphysics in general and theological speculation in particular.¹⁵ But the passage from the introduction continues, singling out natural theology (in particular, natural moral theology) and addressing its prospects in what, given Hume’s skeptical and irreligious tendencies, may seem a surprisingly optimistic vein. Thus in the continuation of the Treatise passage, Hume makes the following striking announcement: And these improvements [in the sciences] are the more to be hop’d for in natural religion, as it is not content with instructing us in the nature of superior powers, but carries its views farther, to their disposition toward us, and our duties toward them; and consequently we ourselves are not only the beings, that reason, but are also one of the objects, concerning which we reason. (T Introduction 4, my emphasis)

The science of man promises progress in that part of natural theology particularly addressing the ‘disposition’ of higher beings toward us, along with ‘our duties toward them.’ The reason is that an accurate conception of our own mental faculties gives us insight not only into the nature and scope of our reasoning powers in this sphere, but also into an important part of its subject matter: namely, ourselves. This points to a remarkable piece of philosophical jujitsu on Hume’s part. For Hume, facts about morality are, in the ﬁnal analysis, facts about human nature. According to his moral theory, there is a settled core of reactive attitudes ingrained in human psychology, approving benevolence, disapproving cowardice, and so on. The moral realm is then ultimately a product of these natural responses to character and behavior: the distinction between vice and virtue, for instance, is not merely mapped but constituted by the natural human propensity to experience sentiments of approval or disapproval toward certain character traits rather than others (T 2.1.7.5, 3.1.1.26, EHU 8.35, ¹⁵ As Hume would put it in the later Enquiry, ‘The only method of freeing learning, at once, from these abstruse questions [concerning the ‘airy sciences’], is to enquire seriously into the . . . human understanding, and show, from an exact analysis of its powers and capacity, that it is by no means ﬁtted for such remote and abstruse subjects’ (EHU 1.12).

42 mitigated skepticism and hume , s natural theology EPM Appendix 1.21, DP 2.14). Our sentiments thus establish the basic framework for all moral evaluation, and provide the only real grounding for an account of the geography of moral distinctions. ‘To suppose measures of approbation and blame, different from the human, confounds every thing. Whence do we learn, that there is such a thing as moral distinctions but from our own sentiments?’ (E 595). Although this sentimentalist theory grounds morality on our own natural human responses, it does not in and of itself present an obstacle to the thesis that non-humans might have moral attributes. An angel could have the sort of character that would trigger our moral sentiments: showing wisdom, courage, or compassion, say, or perhaps arrogance, cruelty, or cowardice. Such a being would accordingly have moral virtues or vices. We might make the point in terms of Hume’s own analogy between his account of moral properties and the modern theory of secondary qualities (T 3.1.1.26, E 166 note).¹⁶ According to the theory of secondary qualities, whether or not something is blue is a function of how it appears to the human visual system. But while visual experiences had by angels are quite beside the point, an angel may certainly have blue wings: wings that would produce the appropriate effects in human observers. Blueness and moral virtue are each deﬁned in terms of characteristic human responses; but this is quite consistent with non-humans being blue or morally virtuous. The point is readily extended to the deity, and in fact the leading early modern sentimentalists before Hume, Anthony Ashley Cooper (Lord Shaftesbury) and Francis Hutcheson, had each combined this species of moral theory with the view that God has a positive moral character. In Hutcheson’s Inquiry into the Original of Our Ideas of Beauty and Virtue (1725), for example, moral distinctions are constituted by the reactive attitudes of our sentimental human psychology.¹⁷ At the same time, since the deity possesses ‘a perfectly wise, uniform, impartial ¹⁶ Although Hume draws this analogy between his own treatment of moral properties and the account of secondary qualities presented by ‘the modern philosophy’ (T 3.1.1.26), he does not endorse this latter theory himself. On this point, see Simon Blackburn, ‘Hume on the Mezzanine Level,’ Hume Studies 19 (1993), 273–88. ¹⁷ Francis Hutcheson, An Inquiry into the Original of Our Ideas of Beauty and Virtue (1725), ed. Wolfgang Leidhold (Indianapolis: Liberty Fund, 2004). For commentary, see Kenneth P. Winkler, ‘Hutcheson and Hume on the Color of Virtue,’ Hume Studies 22 (1996), 3–22; P. J. E. Kail, ‘Hutcheson’s Moral Sense: Skepticism, Realism, and Secondary

mitigated skepticism and hume , s natural theology

43

Benevolence,’ he is ‘good, in a moral Sense.’¹⁸ God has a virtuous moral character: a character, that is, producing our own sentiments of approval.¹⁹ However, while there is no inevitable contradiction between a sentimentalist account of moral distinctions and the doctrine that God has moral attributes, in Hume’s hands the sentimentalist theory, together with further premises about the nature of our actual sentimental psychology, does threaten the moral status of the deity. As the theory of the passions in the Treatise and the Dissertation on the Passions would suggest, as Hume intimates in Essays, Moral and Political, and as he makes plain in private correspondence, he holds that this sort of ultimate cosmological posit is not the ‘natural object’ of any of our sentimental attitudes. The ﬁrst cause or organizing principle behind the ordered universe is not the natural object of any of Passion or Affection. He is no Object either of the Senses or Imagination, & very little of the Understanding, without which it is impossible to excite any Affection. . . . Please to observe, that I not only exclude the turbulent Passions, but the calm Affections. Neither of them can operate without the Assistance of the Senses, & Imagination, or at least a more compleat Knowledge of the Object than we have of the Deity. (NL 13)

As we shall see, this means that the natural mechanisms of the human mind will not produce any feelings or sentimental responses directed Qualities’, History of Philosophy Quarterly 18 (2001), 57–77; Elizabeth Radcliffe, ‘Hutcheson’s Perceptual and Moral Subjectivism,’ History of Philosophy Quarterly 3 (1986), 407–21. For a rival view, according to which Hutcheson’s moral sense discerns real mind-independent moral properties, see David Fate Norton, ‘Hutcheson’s Moral Realism,’ Journal of the History of Philosophy 23 (1985), 397–418. ¹⁸ Hutcheson, Inquiry, 181 (treatise 2, section 7.5). For discussion, see Michael Gill, The British Moralists on Human Nature and the Birth of Secular Ethics (Cambridge: Cambridge University Press, 2006), 186–7; James Moore, ‘Hume and Hutcheson,’ in M. A. Stewart and John P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994) 24–57: 37; Knud Haakonssen, ‘Moral Philosophy and Natural Law: From the Cambridge Platonists to the Scottish Enlightenment,’ Political Science 40 (1988), 97–110: 107. ¹⁹ Although sentimentalism is thus consistent with the proposition that God is morally praiseworthy, some eighteenth-century philosophers would raise other sorts of theological objection to the sentimentalist theory. See for instance John Balguy’s attack on Hutcheson in Balguy, The Foundation of Moral Goodness (1728–9), 2 vols (Facsimile: New York: Garland, 1976), i. 8–9, and George Berkeley’s attack on Shaftesbury in Berkeley, The Theory of Vision Vindicated and Explained (1733), sections 3 and 4.

44 mitigated skepticism and hume , s natural theology toward the deity—or at least, if there are any exceptions to this general psychological law, such cases will be statistically anomalous and perhaps even pathological. If pious believers seem to experience passions directed toward the ﬁrst cause of all, they are most likely deceiving themselves or mistaking the objects of their own emotions; at best they have abnormal psychologies that are no guide for the rest of us. But our moral sentiments are simply a subset of our passions, and so the deity is not a natural object of the moral sentiments either. Given Hume’s sentimentalist conception of what morality is, this entails strong moral atheism, the view that God is beyond all moral assessment, and cannot have either virtues or vices. Once again, we have a species-speciﬁc conclusion about the deity’s relational properties, this time concerning the relationship between any original cause and our human sentimental attitudes, and the consequent inability of any original cause to have certain sorts of properties projected upon it. I document and examine this argument from sentimentalism in Chapters 3 and 4, Chapter 3 focusing on Hume’s thesis that the deity is not a natural object of any human passion, and Chapter 4 on the way in which this thesis conspires with Hume’s sentimentalist moral theory to place this being or principle beyond the moral sphere.

2.5. Conclusion Hume advances arguments for several theses about the divine attributes. He does so through his spokesman Philo, through Philo’s sometime ally Cleanthes, and also in his own voice. These arguments each fall under the rubric of (what I have called) liminal natural theology, a form of reasoning about the divine attributes that avoids claims about the deity’s distinctive intrinsic character, and sticks either to claims that could equally be applied to any given being (or at least to any given unknown being), or to claims about the implications of its deﬁning relational property of being the cause of the ordered world. So Hume’s arguments are consistent at least with his rejection of the traditional program of core natural theology, the attempt to establish species-speciﬁc conclusions about the deity’s distinctive intrinsic properties.

mitigated skepticism and hume , s natural theology

45

Hume’s arguments might still seem to conﬂict with his strongest statements of the futility of cosmological and theological speculation, where he apparently disavows all such theorizing without exception or qualiﬁcation (such as EHU 12.25, DNR 8.12, and DNR 11.5). But now we have taken the measure of Hume’s actual practice across the range of relevant texts, we can see that those unqualiﬁed disavowals are best interpreted as moments of rhetorical grandstanding rather than strict philosophical care and precision. Hume does in fact practice a modest form of natural theology, and so we must read his disavowals of theological speculation as implicitly restricted in some way if he is to avoid transparent inconsistency. And a reading that treats these disavowals purely as a repudiation of core natural theology (leaving liminal natural theology untouched) is, I suggest, a perfectly natural and plausible way of interpreting Hume in his original eighteenth-century context. Read in the polemical context of the period, the Dialogues and the relevant sections of the Enquiry are clearly pitted against the ambitious programs in core natural theology expounded by the likes of Robert Boyle, Richard Bentley, Samuel Clarke, and Joseph Butler. Hume is placing himself at odds with this main tradition in early modern philosophical theology, a tradition that aims at reaching substantive conclusions about the distinctive intrinsic character of the ﬁrst cause of all. This restricted reading of Hume’s disavowals also ﬁts with the more detailed philosophical criticisms that he actually makes of natural theology, for these are one and all objections to its various core incarnations (such as the traditional attempt to prove that the ﬁrst cause of all is a necessary being, or to show that the ﬁrst cause of all bears a distinctive and religiously signiﬁcant similarity to a human mind).²⁰ Finally, at a deeper level, my proposed restriction mirrors the underlying structure of Hume’s general epistemology. For Hume, all our knowledge of matters of fact and existence rests on our meager human experience and the (fallibilistic) extrapolative judgments that we can responsibly make from that experience (EHU 4.1–13); no matter of fact or existence is a priori demonstrable (EHU 12.28–9, 12.32). This will obviously derail the cosmological ‘argument a priori’; and given the narrow range of our empirical evidence and the weakness of the analogy between ²⁰ There is one possible exception, which I address in Section 5.6.

46 mitigated skepticism and hume , s natural theology the observed world and a human artifact, it will also compromise any version of the teleological ‘argument a posteriori’ that purports to take us to a theologically signiﬁcant conclusion about the deity’s distinctive intrinsic character. These famous Humean criticisms of traditional natural theology are continuous with his fundamental epistemological position. However, no part of Hume’s underlying epistemology blocks the three forms of liminal natural theology that I have outlined in this chapter. This liminal natural theology trades either in general aprioristic reﬂection on the necessary constraints governing the character of any being that stands to the universe as some sort of original cause or organizing principle (whatever that being’s particular intrinsic character, and whether or not such a being actually exists), and our knowledge here turns purely on the conceptual relations among our ideas, rather than matters of fact or existence; or it trades in fallibilistic and highly tentative judgments about the likely character of any given unknown being, without claiming any particular insight into the distinctive character of the ﬁrst cause of all; or, alternatively, it trades in empirical knowledge of the limited range of our own human passions. The distinction between a (licit) liminal natural theology and an (illicit) core natural theology is not explicit in Hume’s texts. But no consistent interpretation of Hume’s philosophy of religion is possible that avoids appealing to some such distinction: a distinction that is sensitive both to the fact that Hume does advance some (highly unorthodox, negative, and irreligious) conclusions about the divine nature and to the fact that he ofﬁcially disavows speculation beyond the sphere of everyday experience and common life. My proposed distinction is a reﬂection of the various particular arguments that Hume actually allows (along with those he disallows), plausibly generalized in a way that preserves Hume’s deep sense of our permanent ignorance of the deity’s real essence and the futility of traditional forms of theological speculation. It also squares well with Hume’s general epistemological principles as formulated in the Enquiry and applied in the Dialogues—the principles, of course, that underwrite mitigated skepticism itself. Finally, even if the reader is not persuaded by my proposal that we read this distinction into Hume’s philosophy of religion, so long as it is agreed that Hume

mitigated skepticism and hume , s natural theology

47

does in fact endorse the relevant arguments about the divine attributes reviewed in Section 2.4 above (however his endorsement of these arguments is to be reconciled with mitigated skepticism), then my overall interpretation of Hume’s moral atheism can proceed all the same.

This page intentionally left blank

3 The Argument from Sentimentalism 1: Hume’s Critique of Religious Passions 3.1. Introduction In this chapter and the next I examine Hume’s case that the deity is beyond the range of our natural moral sentiments of approval and blame, and hence beyond the moral world altogether. I call this the argument from sentimentalism, since it turns on Hume’s familiar sentimentalist metaphysics of morals. But the argument also appeals to the underlying mechanics of Hume’s theory of the passions, which imposes limits on the reach of our moral sentiments and thus circumscribes the range of beings that those sentiments can stain or gild with moral properties. In initial overview, the argument from sentimentalism runs as follows. First, Hume claims that the deity (that is, the ﬁrst cause or organizing principle of the ordered world) is not the ‘natural object’ of any of our passions, including love, hatred, gratitude, envy, and the rest. None of these passions—none of our conative or affective attitudes, none of our desires, aversions, and intentional feelings or emotions—are naturally directed toward this sort of ultimate being. Second, according to Hume’s sentimentalist moral theory, moral standards are ﬁxed by the characteristic sentiments of approval and disapproval that humans naturally feel toward different sorts of character trait. Very roughly, if an unbiased human observer would naturally react to such-and-such a behavioral trait with characteristic sentiments of approval, then that trait is a virtue; if they would react

50 the argument from sentimentalism 1 with characteristic sentiments of disapproval, then that trait is a vice. But our sentiments of moral approval and disapproval are just one particular species of passion. So these sentiments of moral approbation and blame, like all our other passions, are not naturally directed toward the ﬁrst cause. It follows that this original being or principle stands outside the realm of moral assessment. The moral sphere is a closed world deﬁned by the limits of natural human sentimental response, and any ﬁrst cause is quite beyond it. To help make the basic strategy of this argument a little more intuitive, consider again the analogy between the sentimentalist account of moral properties and the modern theory of secondary qualities. Given this latter theory, whether or not something is colored is a question of whether, under such-and-such viewing conditions, it produces the appropriate response in the human visual system. Now, suppose that (for whatever reason) the human visual system is quite unable to respond to the deity: that we are constitutionall’y incapable of having a visual experience of the ﬁrst cause or organizing principle. So, given these two assumptions, it follows that the ﬁrst cause is outside the realm of color. It is not simply that we cannot know what color the deity is: rather, if our visual system is quite incapable of responding to the ﬁrst cause, then that being or principle has no color. Likewise in the argument from sentimentalism: since our sentiments of moral approval and disapproval do not take the ﬁrst cause as a natural object, that being or principle is neither virtuous nor vicious, nor morally assessable in any dimension. It is not simply that we cannot know what its moral characteristics are: rather, it has no moral characteristics. The primary goals of this chapter and the next are, ﬁrst, to examine the way in which Hume’s sentimentalism and his account of the limited range of the passions jointly entail strong moral atheism, and, second, to substantiate the attribution of this line of reasoning to Hume. Of course, Hume never openly endorses moral atheism (as we saw in Section 1.4). Nor does he expressly relate his sentimentalist theory of morals to his claim that the ﬁrst cause is beyond the natural reach of our passions: he never explicitly connects these two pillars of the argument from sentimentalism. But he does provide all the materials for the argument, and at least hints at its conclusion. Consider the following formalization:

the argument from sentimentalism 1

51

The argument from sentimentalism S1. The deity is not a natural object of any human passion. S2. Moral sentiments are a species of human passion. S3. If a being is not a natural object of the moral sentiments, then it cannot have moral attributes (either virtues or vices). Therefore S4. The deity cannot have moral attributes (either virtues or vices). Hume expressly endorses each of the three premises, and does so in writings composed within the space of a few years: the 1739–40 Treatise, the 1741–2 Essays, and private correspondence from 1743. (Of the later works, the second Enquiry (1751) reafﬁrms S2 and S3, while the ﬁrst Enquiry (1748) hints at the all-important premise S1.) And these three theses do jointly entail the conclusion S4: it is a fairly obvious consequence of Hume’s sentimentalist moral theory and his view that the original cause is beyond the natural reach of our passions that this being or principle cannot have either virtues or vices. Although he never openly endorses this scandalous conclusion, one would expect Hume to see that his doctrines lead here. There is evidence that he did so, or so at least I shall argue.¹ In the current chapter I focus on principle S1: the thesis that God is not the natural object of any human passion. This is perhaps ¹ If scholars have not previously identiﬁed or discussed the argument from sentimentalism, this is most likely for the following reason. Since Hume does not directly expound the argument in any one place, some detective work is required to connect together doctrines from different parts of his philosophy and follow out their joint implications. In particular, the argument turns crucially on the system of the passions developed in book 2 of the Treatise (and reafﬁrmed in the Dissertation on the Passions (1757)), which is rarely examined alongside his work on religion. Scholars of Hume’s philosophy of religion have concentrated their energies on the later works that address natural and revealed theology most explicitly, and have only recently given closer attention to the irreligious implications of the earlier Treatise. (On the irreligious agenda of much of the Treatise, see Paul Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion (New York: Oxford University Press, 2008.)) By the same token, scholars focusing on Hume’s theory of the passions in the Treatise have not connected this part of his work to his philosophy of religion in any sustained detail. (A noteworthy exception is Jane L. McIntyre, ‘Passion and Artiﬁce in Hume’s Account of Superstition,’ in D. Z. Phillips and Timothy Tesson (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 171–84.) So it is not perhaps so surprising that the argument from sentimentalism has been neglected or overlooked. Nevertheless, we should be able to see that the overall argument can plausibly be attributed to Hume, and that this is not a case of interpolation or excessively imaginative reconstruction.

52 the argument from sentimentalism 1 the least familiar of Hume’s premises, and yet it has far-reaching consequences both for the current question of God’s moral status and for Hume’s wider irreligious agenda. In Section 3.2 I examine the precise meaning of S1 and document Hume’s commitment to this momentous principle. I then explore Hume’s underlying reasons for endorsing S1, connecting the principle to Hume’s general account of the psychological mechanisms controlling all our passions (Section 3.3). In Section 3.4 I address an obvious challenge to S1, namely the fact that so many believers at least seem to experience passions that take God as their object. Hume’s response takes us into his error theory of religious emotion and his attempt to explain away putative cases of passions directed toward the ﬁrst cause of all. Finally, in an appendix (which can be skipped without cost to my main narrative) I relate Hume’s critique of religious sentiment to the wider background of early modern thought about the natural limits of the passions. For all the scandalously irreligious character of Hume’s thesis S1, it turns that out his case for this thesis overlaps to a surprising degree with a popular and perfectly mainstream contemporary understanding of the preconditions for genuine religious sentiment. With Hume’s case for S1 in hand, I then proceed in the following chapter to examine Hume’s more familiar premises S2 and S3, and the way in which they conspire with S1 to complete the argument from sentimentalism. The next chapter presents the evidence that Hume was aware that his theses S1, S2, and S3 jointly entail strong moral atheism. It also addresses the most pressing objections to the argument, weighs possible replies on Hume’s behalf, and takes stock of this sentimentalist case for divine amorality.

3.2. God and the Natural Objects of the Passions Our argument begins with proposition S1, the thesis that God is not the natural object of any human conative or affective attitude (or ‘passion’). As Hume puts it in a letter of June 30, 1743 to his intimate friend William Mure, the deity

the argument from sentimentalism 1

53

is not the natural object of any Passion or Affection. He is no Object either of the Senses or Imagination, & very little of the Understanding, without which it is impossible to excite any Affection. . . . Please to observe, that I not only exclude the turbulent Passions, but the calm Affections. Neither of them can operate without the Assistance of the Senses, & Imagination, or at least a more compleat Knowledge of the Object than we have of the Deity. (NL 13)

This thesis will need some interpretation, and to appreciate Hume’s reasons for endorsing it we will need to examine his broader theory of the passions. But if it appears a highly subversive doctrine with sweeping consequences for traditional piety and religious practice, this much is immediately conﬁrmed by Hume himself. Principle S1 presents (he says) an ‘Objection, both to Devotion & Prayer, & indeed to every thing we commonly call Religion, except perhaps the Practice of Morality, & the Assent of the Understanding to the Proposition that God exists’ (NL 12). Quite apart from its particular role in the argument from sentimentalism, S1 will threaten any religious practice that requires the direction of feelings or sentiments toward the ﬁrst cause of all. Something should be said about the status of Hume’s letter to Mure, which provides us with his only fully explicit and unambiguous statement of principle S1. Hume appears quite serious in advancing S1 in this letter, which can be regarded as something of a Rosetta stone in the interpretation of Hume on religious affect. As we shall see, the reasons Hume outlines in the letter for endorsing S1 are of a piece with his wider critique of natural and revealed theology: there is indeed a ‘remarkable consistency’ (as J. C. A. Gaskin remarks) between this early piece of correspondence and Hume’s later writings on religion.² Moreover, the letter’s framework of assumptions concerning the nature and operation of human sentiment also reﬂects the ofﬁcial account of the passions in the Treatise. In setting out his background reasons for endorsing S1, Hume’s letter thus helps us understand those published texts where he both hints at S1 and hints at these same background reasons. Moreover, the success of this piece of private correspondence in unlocking the published texts—in bringing out ² J. C. A. Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, NJ: Humanities Press International, 1988), 223.

54 the argument from sentimentalism 1 important subtleties and ramiﬁcations, and showing their systematic coherence and interdependence—will help to conﬁrm that it is indeed an accurate key: in opening up Hume’s other writings, the letter shows that it does indeed reﬂect his settled thinking on religious emotion.³ Let us return to the interpretation of thesis S1. How should we understand Hume’s claim that the deity is not the ‘natural object’ of any human passion? More generally, what is it for any being to count as a ‘natural object’ of a passion? What sort of relationship does Hume mean to pick out with this expression?⁴ Unfortunately, while Hume does use this phrase (along with certain other correlate terminology) in his published work on the passions as well as in the letter to Mure (T 2.1.3.2, 2.1.3.4, 2.1.5.3, EPM Appendix 1.21, DP 3.10), he never provides us with a direct and explicit deﬁnition. We can, however, gather a fairly clear understanding of his meaning from the context and occasions of his usage. Begin with what it is for a particular token of a passion—a particular episode of love, pride, gratitude, or fear, etc.—to take something as its ‘object.’ For Hume, any particular passion-token considered in isolation from the rest of our mental life is simply a raw feeling with no intrinsic or necessary intentional or object-directed character.⁵ However, as a matter of contingent but law-like fact, the ³ Hume wrote the letter with the intention that it be passed along to William Leechman (soon to be Professor of Divinity and eventually Principal at Glasgow University), though it was immediately addressed to William Mure, Hume’s close friend and Leechman’s ex-pupil. The purpose of the letter was to provide comments (which were clearly solicited either by Mure or Leechman) on Leechman’s recently published sermon On the Nature, Reasonableness, and Advantages of Prayer (Glasgow: 1743). In tone, the letter is candid and genial, if philosophically critical. But Leechman would not remain on good terms with Hume, and later opposed his candidacy for professorships at Edinburgh and Glasgow (perhaps due in part to the scandalous doctrines set out in this earlier letter). For details of Hume’s exchange with Leechman, see E. C. Mossner, The Life of David Hume, 2nd edn (Oxford: Clarendon Press, 1980), 148–9. On Leechman’s later opposition to Hume, see Mossner, 157; M. A. Stewart, ‘Two Species of Philosophy: The Historical Signiﬁcance of the First Enquiry,’ in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 67–95: 84. ⁴ I previously assayed a range of possible interpretations in Thomas Holden, ‘Hume on Religious Affect,’ Archiv für Geschichte der Philosophie 89 (2007), 283–306. What I say here expands and improves on that discussion. ⁵ Considered in isolation, any given passion ‘contains not any representative quality. . . . When I am angry, I am actually possest with the passion, and in that emotion have no more reference to any other object, than when I am thirsty, or more sick, or more than ﬁve feet high’ (T 2.3.3.5; compare also 3.1.1.9).

the argument from sentimentalism 1

55

passions always occur as part of a wider causal matrix of mental states, with any given passion being caused by a prior impression or idea, and the passion itself then causing another idea in turn. And this causal role—the passion’s place in a systematic and law-like causal network—does give the passion an extrinsic form of intentionality, with Hume characterizing whatever is represented by the idea that a given passion brings to mind as the object of that passion. Thus my reﬂection on a rival’s good fortune (the cause of the passion) brings about the raw feeling we call jealousy (the passion itself), which in turn brings to mind an idea of the rival (the object of this particular passion): I am jealous of him because of his good fortune. Likewise, the fact that the characteristic feelings we call pride and humility always bring to mind an idea of the self establishes that ‘the immediate object of pride and humility is self’; and so similarly, mutatis mutandis, ‘the object of love and hatred is some other person’ (T 2.2.1.2; compare also T 2.2.6.4, DP 2.22). When Hume writes of the objects of the passions, it is this extrinsic form of intentionality he has in mind: the fact that, in virtue of its causal role, an intrinsically non-representational passion serves as a constituent part of conative or affective attitude: a complex intentional or object-directed mental state comprehending both the passion and the idea it immediately produces.⁶ So we have an account of what it is for something to be an object of a particular passion-token. What is it then for something to be a natural object of a passion? This is a claim about the regular, law-like operation of types of passion, as determined by our common human nature. For Hume, it is a basic fact of human psychology that each of the various types of passion directs our thought toward, and thereby refers to, certain types of being rather than others. Our psychological hardwiring ties each species of passion to a given range of objects through universal or near-universal psychological laws, such that that passion-type dependably carries our thought to (and thereby refers to) ´ ⁶ For commentary on this point see P´all Ardal, Passion and Value in Hume’s Treatise (Edinburgh: Edinburgh University Press, 1966), ch. 2; Lilli Alanen, ‘The Powers and Mechanisms of the Passions,’ in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), 179–98: 185–6; Jane L. McIntyre, ‘Hume’s ‘‘New and Extraordinary’’ Account of the Passions,’ in Traiger (ed.), The Blackwell Guide to Hume’s Treatise, 199–215: 209–210.

56 the argument from sentimentalism 1 objects of just that sort. So, on the reading I propose, a being is a natural object of a given type of passion if and only if it falls in the class of objects that our human psychological hardwiring assigns to that passion. The closest Hume comes to directly conﬁrming just what he means by ‘natural object’ is in his well-known analysis of the particular passions of pride and humility: [Pride and humility] are determin’d to have self for their object . . . by a natural . . . property. No one can doubt but this property is natural, from the constancy and steadiness of its operations. ’Tis always self, which is the object of pride and humility; and whenever the passions look beyond, ’tis still with a view to ourselves; nor can any person or object otherwise have any inﬂuence upon us. (T 2.1.3.2)

This establishes at least a sufﬁcient condition for something to qualify as a natural object of a speciﬁc passion: if some passion-type (such as pride or humility) directs our thoughts toward a particular kind of being (such as the self) with a uniform ‘constancy and steadiness,’ then that being is a natural object of that passion. And in another suggestive passage, Hume seems to intimate that this same condition is necessary as well as sufﬁcient: 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 human mind, that these passions shou’d ever look beyond self. . . . Here at last the view always rests, when we are actuated by either of these passions; nor can we, in that situation of mind, ever lose sight of this object. (T 2.1.5.3)

If we read that ﬁrst sentence as an inference rather than a simple conjunction, then Hume is saying that if the self is a natural object of pride and humility—that is, it is ﬁxed as the object of those passions ‘by an original and natural instinct’—then those passions will invariably bring to mind an idea of the self. So on this reading Hume would be treating the regular, law-like direction of a passiontype toward some speciﬁc class of objects to be both a necessary and a sufﬁcient condition for objects of that class to be natural objects of that passion. Moreover, this passage also makes it clear that Hume holds that this sort of law-like regular connection between a

the argument from sentimentalism 1

57

passion and its natural objects must reﬂect ‘the primary constitution of the human mind’—that is, the psychological mechanisms hardwired in human nature. Hume’s theses about the natural objects of the passion are then claims made within descriptive psychology. They are hypotheses advanced within an experimental study of the laws of human thought and emotion, in keeping with his view that ‘in the production and conduct of the passions, there is a certain regular mechanism, which is susceptible of as accurate a disquisition, as the laws of motion, optics, hydrostatics, or any part of natural philosophy’ (DP 6.19).⁷ Hume also uses the phrase ‘proper object’ in close connection with ‘natural object,’ and it is clear from the context that a proper object of a given passion must also be a natural object of that passion: that a particular passion’s proper objects are a subset (perhaps a coincident subset) of its natural objects.⁸ With this in mind we can also look to passages about ‘proper objects’ for further cumulative evidence regarding the meaning of ‘natural object.’ Consider, for instance, the following pair of near-identical sentences from the Treatise and ﬁrst Enquiry, which demonstrates Hume’s willingness to simply substitute ‘proper object’ for ‘constant and universal object’: The only proper object of hatred or anger is a person or creature endowed with thought and consciousness. (EHU 8.29) The constant and universal object of hatred or anger is a person or creature endowed with thought and consciousness. (T 2.3.2.6) ⁷ Adam Smith uses the phrase ‘natural objects of the passions’ in the same descriptivepsychological sense in The Theory of the Moral Sentiments (1759), ed. Knud Haakonssen (Cambridge: Cambridge University Press, 2002), 317. Other early modern philosophers, including Archbishop William King and Bishop Joseph Butler, had used it in a different sense with an explicitly normative force (with ‘natural object’ meaning ‘appropriate or ﬁtting object’). Joseph Butler, Fifteen Sermons preached at the Rolls Chapel, 4th edn (London: 1749), 139 (sermon 8, ‘Upon Resentment’); William King, An Essay on the Origin of Evil (1731) (Reprint: New York: Garland, 1978), 209. ⁸ For instance: ‘let us suppose I am in company with a person, whom I formerly regarded without any sentiments either of friendship or enmity. Here I have the natural and ultimate object of all these four passions [i.e. love, hatred, pride, and humility]. Myself am the proper object of pride or humility; the other person of love or hatred’ (T 2.2.2.2). Unless all proper objects are also natural objects, this passage would make no sense. The letter to Mure also connects ‘proper & natural Object[s] of Affection’ (NL 13). (I cite the full passage in Section 3.3 below.)

58 the argument from sentimentalism 1 Or, in another illustrative case, Hume asserts that certain of our passions take intelligent beings rather than inanimate objects as their ‘proper objects’: There are a numerous set of passions and sentiments, of which thinking rational beings are, by the original constitution of nature, the only proper objects: And though the very same qualities be transferred to an insensible, inanimate being, they will not excite the same sentiments. (EPM 5.1 note)

Whatever normative ramiﬁcations these claims might have, it is clear that they are at least in part a matter of descriptive psychology. Thus certain types of passion, ‘by the original constitution of nature’ direct our thought constantly toward thinking beings; and only the qualities of intelligent agents can be expected to trigger certain sorts of sentimental response. Once again, these are hypotheses advanced within Hume’s experimental science of human nature. So claims about the natural objects of the passions are descriptive claims about the regular psychological laws connecting passion-types to kinds of objects. But what strength do these psychological laws have? Are they supposed to capture a universal and perfect regularity in human affective attitudes, with a given passion-type unfailingly bringing to mind only its natural objects? Or are they more like statistically reliable generalizations, reﬂecting merely a propensity of the mind to move from a certain sort of passion to that passion’s usual sort of object? Can the passions ever ‘spill over’ and range beyond their natural objects, or is this psychologically impossible? From the texts cited above, it is clear enough that pride and humility are tied to their natural object through a perfect and universal psychological law. (Thus ‘’tis absolutely impossible, from the primary constitution of the human mind, that these passions shou’d ever look beyond self’ (T 2.1.5.3).) However, Hume nowhere states that all the passions are like this, or that the natural objects of all the various passion-types are always and invariably their only actual objects. So we should at least leave open the interpretive possibility that certain passions are related to their natural objects by way of a general propensity rather than an unfailing and absolute law. Hume’s claims about the natural objects of the passions might then turn out to be similar to his thesis in the Treatise that our moral sentiments are ‘natural’ (in the sense ‘oppos’d to rare and unusual’) since they are found so

the argument from sentimentalism 1

59

universally ‘rooted in our constitution and temper’ (T 3.1.2.8). As Hume suggests, this thesis about the moral sentiments need not be threatened by a few statistically atypical counterexamples, particularly if such anomalous cases turn out to be explicable in terms of madness or disease (T 3.1.2.8).⁹ So perhaps Hume might similarly advance his claims about the natural objects of the various human passions while yet allowing the existence of a few exceptional (or freakish) cases that do not fall under the usually dependable psychological laws.¹⁰ With this understanding of Hume’s terminology in hand, we can now return to his claim that the deity ‘is not the natural object of any Passion or Affection’ (NL 13). Here Hume is placing the original cause or designer beyond the natural range of all of our passions: he is claiming that this primary being or principle stands outside the natural compass of each of the various types of conative and affective attitude. In fact the deity is by no means unique in this regard. For Hume, there are numerous kinds of being that are simply too remote ⁹ In this connection it is also useful to recall Hume’s treatment of the ‘artiﬁcial lives’ examined in ‘A Dialogue,’ appended to the second Enquiry. There Hume concedes that the occasional philosophical or religious zealot, such as Diogenes the Cynic or the otherworldly Pascal, might manage to suppress or adapt elements of their natural human psychology and thereby live out an artiﬁcial life ‘in a different element from the rest of mankind [where] the natural principles of their mind play not with the same regularity, as if left to themselves, free from the illusions of religious superstition or philosophical enthusiasm’ (EPM, ‘A Dialogue,’ paragraph 57). But Hume treats such cases simply as exceptions to generally reliable psychological laws, and does not take their existence to diminish the importance of providing an account of human nature that captures its usual or regular mechanisms. ¹⁰ As for the status of the psychological regularities connecting the passions to their proper objects: once again, Hume never explicitly states whether these regular conjunctions are universal or merely generally reliable. However, if—as seems likely—Hume intends the normative connotations usually associated with the word ‘proper’ (so that ‘proper objects’ are appropriate or ﬁtting objects) then presumably he would not hold that the psychological regularities connecting the passions to their proper objects are perfect and infallible, for this would entail the implausibly optimistic view that our passions never focus on inappropriate objects. Moreover, even if we bracket all normative considerations, simply from the examples that Hume cites in the passages quoted above it seems unlikely that the ‘proper’ objects of a given sort of passion are always the only actual objects of that passion-type. For instance, whether or not it is appropriate to do so, humans surely do sometimes feel anger toward inanimate objects. So on pain of a fairly obvious counterexample, Hume presumably does not mean to say that the proper objects of the passions are also their exclusive actual objects. But it is clear that he at least means to say that cases of a passion taking an improper object are statistically atypical, a divergence from the usual operation of the human mind, and, depending on the particularities of the case, perhaps even pathological in some way. (When Xerxes is consumed by rage toward the Bosporus for disrupting his invasion plans and orders that the waters be ﬂogged, his behavior is, to say the least, unusual.)

60 the argument from sentimentalism 1 from the emotional world of common life to be the natural object of any human sentiment. This sphere of emotionally inaccessible beings encompasses all transcendental or supersensible entities along with many of the distant posits of speculative cosmology. To bring this out, consider two important passages, each of which attacks a version of the high-minded proposal that we focus our sentiments and feelings away from parochial human concerns and toward ultimate cosmological realities. In the ﬁrst, from the Enquiry concerning Human Understanding, Hume attacks the Stoic doctrine that we can look beyond our local misfortunes to seek consolation in ‘an enlarged view’ encompassing ‘the whole system of nature.’¹¹ The Stoic sage can supposedly reﬂect on the perfection of the whole universe as a uniﬁed totality and thus regard every event within it as ‘an object of joy and exultation.’ But Hume regards all this as psychologically quite unreal: These enlarged views may, for a moment, please the imagination of a speculative man, who is placed in ease and security; but neither can they dwell with constancy on his mind, even though undisturbed by the emotions of pain or passion; much less can they maintain their ground, when attacked by such powerful antagonists. The affections take a narrower and more natural survey of their object; and by an œconomy, more suitable to the inﬁrmity of human minds, regard alone the beings around us, and are actuated by such events as appear good or ill to the private system. (EHU 8.34, my emphasis)

This passage does allow that a speculative man might (‘for a moment’) be able to strain his feelings beyond their usual range and experience sentiments directed toward the entire creation. But this is quite beyond most of us. Rather, our passions operate according to the tides and fortunes of ‘the private system’ and are much more localized in their range and sphere: they ‘take a narrower and more natural survey of ¹¹ Hume attributes the position to ‘some philosophers, and the ancient Stoics’ (EHU 8.34). Beauchamp suggests that ‘some philosophers’ may include the classicist Justus Lipsius (1547–1606). (Tom L. Beauchamp, ‘Editor’s Annotations’ to David Hume, An Enquiry concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000), 124–200: 165.) Other early modern candidates would include quietist philosophers such as Francois de Salignac de la Mothe F´enelon, Archbishop of Cambrai (1651–1715), and his disciple Andrew Michael Ramsay (‘the Chevalier Ramsay’) (1686–1743), and perhaps even Leibniz to the extent that he treats the doctrine of this world’s overall optimality as a topic for personal consolation.

the argument from sentimentalism 1

61

their object’ and ‘regard alone the beings around us.’ At least for most humans, the emotional world is a sublunary one, deﬁned by the concrete and immediate concerns of human life, not the abstract forms of speculative cosmology. The second passage of interest is drawn from the earlier essay ‘The Sceptic,’ published in Essays, Moral and Political. This essay is one of four companion pieces setting out the leading philosophical world views on happiness and the good life, and in each case Hume speaks in character as advocate for the system in question rather than in propria persona. However, ‘The Sceptic’ is at least generally close to Hume’s own views,¹² and the similarity between this essay’s skeptical treatment of religious emotion and the later rejection of Stoic ‘enlarged views’ is noteworthy. The argument found in ‘The Sceptic’ is a direct response to the commendation of devotional passions in the companion essay ‘The Platonist,’ subtitled ‘the man of contemplation, and philosophical devotion.’ In that prior essay, Hume had set out the world view of the theistic philosopher committed to adoration of the deity (E 155–8). But ‘The Sceptic,’ immediately following, challenges this ‘philosophical devotion’ quite explicitly. It is simply not psychologically possible to sustain the focus of any of our passions on as unfamiliar a being as the god of the philosophers: Philosophical devotion . . . like the enthusiasm of a poet, is the transitory effect of high spirits, great leisure, a ﬁne genius, and a habit of study or contemplation: But notwithstanding all these circumstances, an abstract, invisible object, like that which natural religion alone presents to us, cannot long actuate the mind, or be of any moment in life. To render the passion of continuance, we must ﬁnd some method of affecting the senses and imagination, and must embrace some historical, as well as philosophical account of the divinity. Popular superstitions and observances are even found to be of use in this particular. (E 167)

So here in ‘The Sceptic,’ the narrator asserts that there is a serious psychological obstacle facing any attempt to direct our passions toward the god of natural religion—although he does allow that historical accounts of the deity such as those provided by revealed religion and ‘[p]opular superstitions’ do a better job of stirring our senses and ¹² For instance, the narrator of this essay famously defends sentimentalism in ethics (E 166 note 3).

62 the argument from sentimentalism 1 imagination, and hence of focusing our emotions. (As we shall see in Section 3.4, this is less of a concession than it may at ﬁrst appear, for historical religions only succeed in engaging our passions insofar as they present us with sensible forms that cannot possibly be identiﬁed with the ultimate ﬁrst cause or designer.) Both the anti-Stoic and the anti-Platonist passages thus stress the difﬁculty of directing our feelings toward distant cosmological entities such as the universe at large or its original cause. Neither of these passages quite amounts to an explicit endorsement of S1, however, and each of them ofﬁcially allows that a rare ‘speculative man’ or contemplative of ‘ﬁne genius’ might succeed in lifting his sentiments toward such remote objects, at least for a moment. But we might wonder how serious Hume is in making this concession, or whether he is simply sugar-coating his main point that the natural mechanisms of human psychology in fact draw the passions to the more concrete and immediate objects of sense and imagination. Perhaps the Stoic sage and Platonic devotee, in their solipsistic withdrawal from the world of sense and everyday emotional commerce, are deceiving themselves or misinterpreting the objects of their own feelings. Or alternatively, supposing that they can indeed focus their passions on such ultimate cosmological realities, perhaps such individuals are sufﬁciently atypical that they do not contradict the thesis that the deity is not the natural object of any passion, so long as we understand that thesis to reﬂect the statistically regular or common workings of the human mind rather than a universal and absolute psychological law.¹³ While the anti-Stoic and anti-Platonist passages ofﬁcially allow the existence of people who can turn their sentiments toward ultimate cosmological realities, and treat such ﬁgures with a gentle, if stagy, reverence as contemplatives of ‘ﬁne genius’ and ‘enlarged views,’ in other places Hume adopts a rather more skeptical tone. In the letter to Mure (composed just months after the publication of the anti-Platonist passage in ‘The Sceptic’), he suggests that, in the actual course of human affairs, those who attempt to direct their passions away from the sensible world and toward adoration of a transcendental ﬁrst cause of all are simply deluding themselves, or at the very best have anomalous or freakish psychologies that are no model for the ¹³ I return to this issue in section 4.3.2.

the argument from sentimentalism 1

63

rest of us. Here it is instructive to contrast the deferential tone of the anti-Stoic and anti-Platonist passages with the dismissive remarks of Hume’s letter to Mure: A man . . . may have his Heart perfectly well disposd towards every proper & natural Object of Affection, . . . & yet from this Circumstance of the Invisibility and Incomprehensibility of the Deity may feel no Affection towards him. And indeed I am afraid that all Enthusiasts mightily deceive themselves. Hope & Fear perhaps agitate their Breast when they think of the Deity: Or they degrade him into a Resemblance of themselves, & by that means render him more comprehensible. Or they exult with Vanity in esteeming themselves his peculiar Favourites. Or at best they are actuated by a forc’d and strain’d Affection, which moves by Starts and Bounds, & with a very irregular disorderly Pace. (NL 13)

Here in Hume’s private correspondence, those who attempt to strain their feelings toward ultimate cosmological realities no longer appear as high-minded philosophical sages, but unbridled and self-deluding enthusiasts. Their claims to religious feeling are not treated as a serious threat to Hume’s thesis about the natural limits of our affective attitudes, but are to be pathologized and explained away. The Stuart volumes of the History of England present us with another illustration of Hume’s skeptical attitude toward those who claim to feel passions directed toward the ﬁrst cause of all, this time with reference to a concrete historical case. Here Hume provides the following account of the early seventeenth-century Scottish Kirk and its puritanical ideology of direct adoration of the divinity, unmediated by any sensible iconography, ritual, or Romish idolatry: A mode of worship was established, the most naked and simple imaginable; one that borrowed nothing from the senses; but reposed itself entirely on the contemplation of that divine Essence, which discovers itself to the understanding only. This species of devotion, so worthy of the supreme Being, but so little suitable to human frailty, was observed to occasion great disturbances in the breast, and in many respects to confound all rational principles of conduct and behavior. The mind, straining for these extraordinary raptures, reaching them by short glances, sinking again under its own weakness, rejecting all exterior aid of pomp and ceremony, was so occupied in this inward life, that it ﬂed from every intercourse with society, and from every chearful amusement, which could soften or humanize the character. . . . In order to mellow these humours, James [the Sixth of Scotland

64 the argument from sentimentalism 1 and First of England] endeavoured to infuse a small tincture of ceremony into the national worship, and to introduce such rites as might, in some degree, occupy the mind, and please the senses, without departing too far from that simplicity, by which the reformation was distinguished. (HE v. 68–9)

Hume’s incredulity comes through clearly, as does his projective identiﬁcation with James’s exasperation in dealing with these fanatics. At least for most humans, the attempt to ﬁx our passions on this sort of transcendental object (as opposed to the sensible forms of icons and ceremony) leads only to the breakdown of controlled emotion into self-deceiving raptures, the hopeless projection of one’s own inner sentiments, and perhaps also, in the withdrawal from the moderating inﬂuence of the world of everyday human life, intolerance, and unsociability. Thus while the History’s account of Scottish iconoclasm and the letter to Mure are each strictly consistent with the possibility of some rare contemplative experiencing passions that are directed toward the cosmological empyrean, they suggest that Hume is typically quite skeptical of claims to such ‘extraordinary raptures.’ It is in fact quite possible to read him as implying, with no more than the usual level of irony he brings to such theologically sensitive topics, that no one can in fact genuinely direct their passions toward the original cause or designer.¹⁴ But even if Hume does not go so far as to completely rule out the psychological possibility of passions that are directed toward the ﬁrst cause, it is clear that he regards such attitudes as at best highly atypical and quite out of line with the usual workings of the human mind. (This weaker reading of S1 is all that the overall argument from sentimentalism will require.) To summarize. Hume does endorse principle S1, which can now be understood as the thesis that, given our natural human psychological hardwiring, our conative and affective attitudes either cannot be ¹⁴ Hume’s real response to those who attempt to direct their passions away from their characteristic natural objects would then coincide with the following remarks, placed in the mouth of his character ‘The Epicurean’ in opposition to the rival visions of high-minded spiritual uplift sponsored by ‘The Platonist’ and ‘The Stoic’: ‘To what purpose should I pretend to regulate, reﬁne or invigorate any of those springs and principles, which nature has implanted in me? . . . When by my will alone I can stop the blood, as it runs with impetuosity along its canals, then may I hope to change the course of my sentiments and passions. In vain should I strain my faculties, and endeavour to receive pleasure from an object, which is not ﬁtted by nature to affect my organs with delight’ (E 139–40).

the argument from sentimentalism 1

65

directed toward the ﬁrst cause or organizing principle of the ordered universe—or at least, if there are any exceptions to this general psychological law, such cases are out of line with the regular mechanisms governing the passions, statistically unusual, and perhaps even pathological. This constraint on the natural objects of our passions is stated explicitly in the letter to Mure, but it is also suggested by the anti-Stoic passage from the ﬁrst Enquiry and the anti-Platonist passage from ‘The Sceptic.’ Further, as we shall see in the following section, it is also conﬁrmed when we examine Hume’s account of the psychological machinery controlling the passions, and thus his underlying reasons for endorsing S1.

3.3. Sense, Imagination, and the Natural Objects of the Passions Why exactly does Hume endorse S1? Why should we think that the ﬁrst cause is not the natural object of any human passion? As our passage from ‘The Sceptic’ indicates, the fundamental problem is that to guide any of our passions, ‘we must ﬁnd some method of affecting the senses and imagination’ (E 167). The letter to Mure presents the full argument, drawing a useful analogy with the case of a remote ancestor who (like the deity) is not adequately represented by either of these faculties, and hence cannot be an object of our passions: It must be acknowledg’d that Nature has given us a strong Passion of Admiration for whatever is excellent, & of Love & Gratitude for whatever is beneﬁcial, & that the Deity possesses these Attributes in the highest Perfection & yet I assert he is not the natural Object of any Passion or Affection. He is no Object either of the Senses or Imagination, & very little of the Understanding, without which it is impossible to excite any Affection. A remote Ancestor, who has left us Estates & Honours, acquir’d with Virtue, is a great Benefactor, & yet ’tis impossible to bear him any Affection, because unknown to us; tho in general we know him to be a Man or human Creature, which brings him vastly nearer our Comprehension than an invisible inﬁnite Spirit. A man, therefore, may have his Heart perfectly well disposd towards every proper & natural Object of Affection, Friends, Benefactors, Countrey, Children &c, & yet from this Circumstance of the Invisibility and Incomprehensibility of the Deity may feel no Affection towards him. . . . Please to observe, that I

66 the argument from sentimentalism 1 not only exclude the turbulent Passions, but the calm Affections. Neither of them can operate without the Assistance of the Senses, & Imagination, or at least a more compleat Knowledge of the Object than we have of the Deity. (NL 13)

No matter how much a descendant might have beneﬁted from a distant forebear, so long as that descendant lacks sufﬁcient direction from his senses, imagination, or understanding, he will not in fact be able to experience feelings of gratitude toward his ancestor. The objects of the passions are limited to those beings that are adequately painted forth by these other representational faculties, and our conative and affective attitudes are thus far parasitic on the deliveries of sense, imagination, and the understanding. Of course, the descendant in Hume’s scenario could conceivably come to learn more about his remote ancestor, and presumably might then experience feelings of gratitude toward him.¹⁵ But Hume indicates that our cognitive situation vis-à-vis the deity is in this respect much worse than that of the descendant’s vis-à-vis his remote ancestor. God is not merely currently unseen and unknown, but ‘Invisib[le] and Incomprehensib[le]’: the modal predicates seem to place the deity permanently beyond adequate representation by our human faculties. So, Hume concludes, the deity is not the natural object of any human passion. While ‘The Sceptic’ framed the issue simply in terms of sense and imagination, here in the letter to Mure Hume states the argument for S1 in terms of three faculties: the senses, the imagination, and the understanding. But this is really a distinction without a difference. We should recall that Hume self-consciously uses ‘the imagination’ both in a narrow sense that excludes the understanding, and in a broader sense covering all the ways in which we bring non-memory ideas to mind, including the operations of the understanding.¹⁶ (With this latter usage, the understanding just is ‘the general and more establish’d properties ¹⁵ For a complication, see note 18. ¹⁶ ‘To prevent all ambiguity, I must observe, that where I oppose the imagination to the memory, I mean in general the faculty that presents our fainter ideas. In all other places, and particularly when it is opposed to the understanding, I understand the same faculty, excluding only our demonstrative and probable reasonings’ (T 2.2.7.6 note; see Norton and Norton’s editorial comments on this paragraph at p. 653, and compare T 1.3.9.19 note, T 1.4.7.7). For discussion, see Don Garrett, Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997), 11–40, esp. 28–9.

the argument from sentimentalism 1

67

of the imagination’ (T 1.4.7.7): the settled and regular associative mechanisms by which we produce our non-memory ideas, rather than the more haphazard associative operations that constitute the ‘fancy,’ i.e. imagination in the narrower sense.) With this in mind, we can read Hume as insisting that to be the object of any passion, a being must be represented in sufﬁcient richness and detail either (i) by the senses or (ii) by the imagination, here intended in the broader sense encompassing both the fancy and the understanding. (From hereon, I always use ‘imagination’ in this wider sense.) And since all the ideas of the imagination are themselves ultimately derived from sensory and reﬂective experience, it follows that we will need some sort of imagistic representations in order to focus our passions—imagistic, that is, in a broad sense covering all the impressions of outer sense (including auditory, gustatory, olfactory, tactile, and visual impressions of sensation) and inner reﬂection (including love, anger, and other impressions of reﬂection), plus the ideas we can construct from copies of these impressions. For Hume the understanding no less than the rest of the imagination trades exclusively in image-like ideas that copy the impressions of outer and inner sense: his stripped-down, severely empiricist model of mind handles all cognitive representation in terms of mental imagery that is ultimately traceable to experience, ruling out the sort of non-imagistic ‘intellectual’ ideas posited by Descartes and other rationalist philosophers.¹⁷ This is the crucial point for our purposes: Hume requires sufﬁciently rich imagistic representations via either the senses or the imagination in order for the passions to engage. Thus it is no accident when the Treatise insists that ‘Our love and hatred are always directed to some sensible being external to us’ (T 2.2.1.2, my emphasis). And this is why both ‘The Sceptic’ and the letter to Mure emphasize the ‘invisible’ character of the ﬁrst cause or designer in placing this being beyond the natural reach of our passions (E 167, NL 13; see also NHR 5.2). Hume is telling us that we lack the sort of rich mental imagery of the original cause required to stir and focus our conative and affective attitudes toward it. ¹⁷ Don Garrett, Cognition and Commitment, 14–17. On Descartes’s non-imagistic intellectual ideas, see Gary S. Hatﬁeld, ‘The Cognitive Faculties,’ in Daniel Garber and Michael Ayers (eds), The Cambridge History of Seventeenth-Century Philosophy, 2 vols (Cambridge: Cambridge University Press, 1998), ii. 953–1002: 970.

68 the argument from sentimentalism 1 This general limitation on all our passions drives the case for S1 in both the letter to Mure and ‘The Sceptic.’ But it also emerges from Hume’s theory of the passions, the subﬁeld of his science of human nature that maps the regular laws controlling the ‘production and conduct of the passions’ (DP 6.19). Thus in book 2 of the Treatise and again (without substantive change) in the Dissertation on the Passions Hume delineates the relationship between the ﬂux of imagery in our minds on the one hand and the onset and direction of our sentimental attitudes on the other, offering up a system of hypotheses concerning the ways in which our ideas can succeed in engaging and holding our emotional attention. Here he identiﬁes at least two psychological mechanisms that help to explain his exclusion of the deity from the natural scope of the passions, and would seem to ground his case for S1 as sketched out in the letter to Mure and hinted at in ‘The Sceptic.’ First, Hume claims that our passions are particularly responsive to our more vivid or lively mental imagery, including our richer and more vibrant ideas and (more lively still) the bold, dense imagery of immediate sensory and reﬂective impressions. Fainter ideas will not engage our emotional responses so readily, and below a certain threshold will not detain us at all. (We might consider, for instance, the weak emotional purchase of half-forgotten memories, or of idle musings that barely register at the periphery of consciousness.) It follows that in our emotional life we are constantly drawn to and preoccupied with those objects that are represented by our more vivacious mental imagery—including, ﬁrst and foremost, the immediate impressions of current sensory and reﬂective experience, but also the various enlivened ideas that (as Hume notoriously has it) constitute the salient beliefs crowding in on our consciousness through an association of ideas (T 1.3.7.5–8, 1.3.8.2). Treatise 2.3.6—‘Of the inﬂuence of the imagination on the passions’—and Dissertation on the Passions 6.9–18 are devoted to this doctrine, with Hume maintaining that the imagination and affections have a close union together, and . . . nothing, which affects the former, can be entirely indifferent to the latter. Wherever our ideas of good and evil acquire a new vivacity, the passions become more violent; and keep pace with the imagination in all its variations. (T 2.3.6.1; see also DP 6.13)

the argument from sentimentalism 1

69

He argues for this harmonic correspondence between the ﬂux of imagery in our minds and the course of our passions by way of the following sort of observations: Any satisfaction, which we lately enjoyed, and of which the memory is fresh and recent, operates on the will with more violence, than another of which the traces are decayed and almost obliterated. (DP 6.14; compare T 1.3.6.5) Nothing is more capable of infusing any passion into the mind, than eloquence, by which objects are represented in their strongest and most lively colours. We may of ourselves acknowledge, that such an object is valuable, and such another odious; but till an orator excites the imagination, and gives force to these ideas, they may have but a feeble inﬂuence either on the will or the affections. (T 2.3.6.7; compare DP 6.16) [B]elief is nothing but a lively idea related to a present impression. This vivacity is a requisite circumstance to the exciting all our passions, the calm as well as the violent; nor has a mere ﬁction of the imagination any considerable inﬂuence upon either of them. ’Tis too weak to take any hold of the mind, or be attended with emotion. (T 2.3.6.10; see also T 1.3.10.3–4)

Our conative and affective attitudes are then dependent on vivacious mental imagery—‘a requisite circumstance to the exciting all our passions’—and absent an immediate impression or a sufﬁciently lively idea our sentiments simply will not engage. Further, just as the letter to Mure excludes both ‘the turbulent Passions, [and] the calm Affections,’ so here in the Treatise ‘all our passions, the calm as well as the violent’ require stimulus and direction by sufﬁciently vibrant mental imagery. So we have one reason why our passions are not naturally directed toward the deity. Although Hume does not deny that we have some sort of cognitive representation of the original cause or designer by way of the understanding, he insists that this representation is at best a very limited one. (Thus the deity ‘is no Object either of the Senses or Imagination [i.e. the fancy], & very little of the Understanding,’ and we would require ‘a more compleat Knowledge of the Object than we have of the Deity’ in order for it to excite any of our passions (NL 13, my emphases).) And this sort of limited conception—like the limited conception the descendant has of his remote ancestor—clearly lacks the sort of rich cognitive content that Hume associates with liveliness

70 the argument from sentimentalism 1 and vivacity.¹⁸ So although we may have some sort of conception of a being or principle that stands to the world as an original cause or designer, that idea may be too faint to engage our passions. A second factor controlling the onset and direction of our passions concerns the speciﬁcity or generality of our thoughts, and hence (in Hume’s cognitive economy) the stability of the mental imagery at work in our minds. According to Hume, the more speciﬁc our ideas are in their representative content, the more readily they will attract and hold our emotional attention. But as our ideas become more general and abstract, the less they engage our affective responses. Although all our ideas are imagistic and hence perfectly speciﬁc and determinate in their own qualitative character (T 1.1.7.5–6), an idea can nevertheless ‘be particular in [its] nature, but general in [its] representation’ when it is ‘annex’d to a general term; that is, to a term, which from a constant conjunction has a relation to many other particular ideas, and readily recals them in the imagination’ (T 1.1.7.10). But this model of how abstract thought functions makes the particular imagery at work readily substitutable, as we ‘readily recal[l]’ to mind any particular idea that falls under the general term in place of the particular idea we started with: we ‘keep ourselves in a readiness to survey any of them, as we may be prompted by a present design or necessity,’ the custom associated with the general term ‘produc[ing] any other individual [idea], for which we may have ¹⁸ There may also be a further reason why our idea of the deity is faint rather than vivacious. According to Hume, the more distant we take an object to be from us in space and time, the fainter our ideas of that object will become, since we must in our thoughts both trace out its distance from us, passing ‘thro’ all the intermediate space betwixt ourselves and the object’, and also ‘renew our progress every moment, being every moment recall’d to the consideration of ourselves and our present situation’ (T 2.3.7.2). The result is that objects that are regarded as remote will less readily engage our passions. ‘[C]onsider two kinds of objects, the contiguous and remote, of which the former, by means of their relation to ourselves, approach an impression in force and vivacity; the latter, by reason of the interruption in our manner of conceiving them, appear in a weaker and more imperfect light. This is their effect on the imagination. If my reasoning be just, they must have a proportionable effect on the will and passions’ (T 2.3.7.2–3, my emphasis; see also T 2.3.7.3, 2.3.8.1). So even if the descendant did know a good deal more about his ancestor, still, if the ancestor lived sufﬁciently long ago, then the descendant (however well-informed) might still be unable to feel gratitude toward him due to this remoteness in time. Similarly, if our thoughts of the original cause of all have to trace back to this primary being or principle through the entire history of the universe, our idea of this ﬁrst cause may grow fainter and its ability to engage our emotions decrease.

the argument from sentimentalism 1

71

occasion’ (T 1.1.7.7). Thus when someone employs the general term ‘triangle,’ determinate mental imagery of a particular triangle will be brought to mind—imagery, perhaps, of an equilateral triangle of such and such a speciﬁc color and size. But as one begins to reason with this idea and the conversation proceeds, the original imagery may easily be switched out, according to need or whim, for alternative imagery of other particular triangles (T 1.1.7.8). This instability of imagery in abstract thought makes it ill-adapted to hold our emotional attention: ’tis certain, that the more general and universal any of our ideas are, the less inﬂuence they have upon the imagination. A general idea, tho’ it be nothing but a particular one consider’d in a certain view, is commonly more obscure; and that because no particular idea, by which we represent a general one, is ever ﬁx’d or determinate, but may easily be changed for other particular ones, which will serve equally in the representation. (T 2.3.6.2) When a quality becomes very general, and is common to a great many individuals, it leads not the mind directly to any one of them; but by presenting at once too great a choice, does thereby prevent the imagination from ﬁxing on any single object. (T 1.1.5.3),

So the more abstracted and general our thoughts are, the less ﬁxity and stability in our imagistic representations, and the less readily they will engage us. While our more abstract thoughts (like all thoughts) do involve mental imagery, the images here may then be too mercurial to draw our conative and affective attitudes. That would explain Hume’s view that theoretical reason ceases to engage our passions when it trades in highly abstract thoughts too far removed from the immediate world of sense and experience: The affections are not susceptible of any impression from the reﬁnements of reason or imagination; and it is always found, that a vigorous exertion of the latter faculties, necessarily, from the narrow capacity of the human mind, destroys all activity in the former. (EPM appendix 2.7) The straining of the imagination always hinders the regular ﬂowing of the passions and sentiments. . . . As the emotions of the soul prevent any subtile reasoning and reﬂection, so these latter actions of the mind are equally prejudicial to the former. (T 1.4.1.11; see also EHU 1.3)

So we can see why Hume holds that our passions tend to focus on the concrete and immediate rather than the abstract and remote.

72 the argument from sentimentalism 1 The attractive force of our more vivacious and our more speciﬁc ideas, along with the attractive force of our reliably vivacious and speciﬁc impressions, will ineluctably draw our emotional attention back toward the world of immediate and familiar experience. Indeed, Hume is clear that this natural tendency toward parochialism in our sentimental attitudes is quite essential to human survival. It prevents our passions from being feverishly distracted by every passing thought of good and evil;¹⁹ it enables us to steer toward achievable goals;²⁰ and it even saves us from the ‘philosophical melancholy and delirium’ and paralyzing total skepticism that Hume famously wrestles with in the conclusion of Treatise book 1 (T 1.4.7.9).²¹ Hume’s association of abstract thought with instability of imagery, and instability of imagery with an inability to focus and engage our passions, all squares well with the example of the remote ancestor and the argument of Hume’s letter to Mure. The remote ancestor, recall, is ‘unknown to us; tho’ in general we know him to be a Man or human Creature’ (NL 13). From this position of ignorance, the descendant can only conceive of his distant benefactor by way of a highly abstract relative idea, the cognitive correlate of the relational description ‘the human creature, whosoever it might be, that bequeathed these ¹⁹ ‘[D]id every idea inﬂuence our actions, our condition wou’d not be much mended. For such is the unsteadiness and activity of thought, that the images of every thing, especially of goods and evils, are always wandering in the mind; and were it mov’d by every idle conception of this kind, it wou’d never enjoy a moment’s peace and tranquility’ (T 1.3.10.2; see also T 1.3.10.3). ²⁰ ‘It is wisely ordained by nature, that private connexions should commonly prevail over universal views and considerations; otherwise our affections and actions would be dissipated and lost, for want of a proper limited object’ (EPM 5.42 note). ²¹ Consider the psychological mechanism by which nature saves us from a ‘total scepticism’ and enables us to return to the natural beliefs of common life: ‘that singular and seemingly trivial property of the fancy, by which we enter with difﬁculty into remote views of things, and are not able to accompany them with so sensible an impression, as we do those, which are more easy and natural’ (T 1.4.7.7, my emphasis). Highly abstract reﬂections do not engage either our passions or our cognitive attention as readily as concrete and immediate perceptions, and thus a simple game of backgammon can cure us of the paralysis of total skepticism (T 1.4.7.9; compare also T 1.4.1.9–10). Similarly, in the ﬁrst Enquiry, ‘Pyrrhonism or the excessive principles of scepticism’ ‘may ﬂourish and triumph in the schools; where it is, indeed, difﬁcult, if not impossible, to refute them. But as soon as they leave the shade, and by the presence of the real objects, which actuate our passions and sentiments, are put in opposition to the more powerful principles of our nature, they vanish like smoke, and leave the most determined sceptic in the same condition as other mortals’ (EHU 12.21).

the argument from sentimentalism 1

73

estates’.²² (Although this idea refers to a particular individual, it remains highly abstract in its descriptive content: any thoughts the descendant has of his ancestor could draw on a wide-open play of almost indeﬁnitely substitutable imagery.) And similarly, Hume holds that we can only conceive of the deity by way of a highly abstract relative idea, the cognitive correlate of the relational description ‘the being or principle, whatsoever its intrinsic nature, that stands to the universe as some sort of ﬁrst cause or designer’.²³ Here again, we could run through a vast array of possible imagery, a succession of endlessly substitutable particular ideas that are simply too evanescent to engage our conative and affective attitudes. So again, while Hume only explicitly afﬁrms principle S1 in the letter to Mure (and merely hints at it in published texts like the anti-Platonist and anti-Stoic passages), his ofﬁcial theory of the passions would seem to support it. We need concrete sensory impressions or sufﬁciently rich (i.e. vivid and determinate) imagistic ideas in order to excite our conative and affective attitudes and direct them toward a particular object.²⁴ ²² On relative ideas, see Section 1.2. ²³ Recall that Hume ﬁxes the reference of ‘God’ (and its synonyms) in terms of the following highly general, relational speciﬁcation: ‘the original cause of this universe (whatever it be) we call God’ (DNR 2.3). (For discussion, see Section 1.2.) Recall also that he suggests (through Philo) that ‘the whole of natural theology’ ultimately reduces to a ‘somewhat ambiguous, at least undeﬁned, proposition’ countenancing a remote analogy between human intelligence and ‘the cause or causes of order in the universe’ (original emphasis). Again, the idea of the divinity here seems highly abstract, and indeed its emotional impotence is borne out in its inability to excite any action or conduct. In Hume’s judgment, natural theology may (at best) license a weak probabilistic inference to a remote analogy between the original cause and human intelligence—but crucially ‘this proposition affords no inference that affects human life, or can be the source of any action or forbearance’ (DNR 12.33, see also EHU 11.27). So reﬂection on the content of this proposition cannot produce any human behavior or activity, presumably because it cannot excite any of our passions. (On this last point, see also Jane L. MacIntyre, ‘Passion and Artiﬁce,’ 183–4.) ²⁴ For a recent discussion of the relationship between mental imagery and the operation of the emotions, see Martha Nussbaum, Upheavals of Thought: The Intelligence of Emotions (Cambridge: Cambridge University Press, 2001). Nussbaum maintains that our emotional attitudes are at least often accompanied by mental imagery: ‘[T]he experience of emotion usually contains . . . rich and dense perceptions of the object, which are highly concrete and replete with detail’; ‘the emotions typically have a connection to imagination, and to the concrete picturing of events in imagination, that differentiates them from other, more abstract judgmental states’; ‘human emotions are shaped by the fact that we are perceiving creatures: they derive their rich texture from those sensory abilities’ (65; see also 66–7). Hume would of course endorse these claims, and perhaps we can see him as taking a step further, asserting (perhaps less plausibly) that introspection reveals that this sort of concrete

74 the argument from sentimentalism 1

3.4. Religious Passions and Self-Deceit According to principle S1, the ﬁrst cause or organizing principle is not the natural object of any human passion: most humans (if not all) are unable to experience passions that genuinely refer to the deity. But this claim raises the obvious question of what we are to make of those many believers who at least take themselves to experience sentiments and feelings directed toward the ﬁrst cause of all. Indeed, the fact that so many people appear to experience passions directed toward the divine may seem to render principle S1 quite incredible. Perhaps Hume has a background theory that sufﬁciently rich imagistic representations are required to engage our passions. But given that so many pious believers do experience sentiments directed toward the deity, then (contra Hume) either humans often in fact have adequate mental imagery to guide such attitudes, or that background theory is false. How does Hume propose to deal with this clear and obvious objection to his claim that the deity lies beyond the natural range of human sentiment and feeling? Hume engages this sort of challenge both in the letter to Mure and in his published work, including the early Essays, the later Natural History of Religion, and, in a more oblique fashion, the History of England. His basic claim is that in most (if not all) cases of religious sentiment, the believer’s passions are not really directed toward the transcendental ﬁrst cause of all, and if the believer thinks otherwise, then he must be misinterpreting the objects of his own feelings. In the letter to Mure, Hume diplomatically restricts this error theory of religious feeling to the ill-regulated fervor of ‘enthusiasts,’ and indeed this sort of diagnosis of the extravagant devotion of ecstatics and charismatics was common enough in his day.²⁵ But his published picturing is not merely the usual attendant of emotion, but a constituent part of any conative or affective attitude whatsoever. ²⁵ Henry More, Enthusiasmus Triumphatus, in A Collection of Several Philosophical Writings of Dr. Henry More, 2nd edn (London, 1662); John Locke, An Essay concerning Human Understanding, ed. Peter H. Nidditch (Oxford: Clarendon Press, 1975), book 4, chapter 19. For commentary, see Frederick Beiser, The Sovereignty of Reason: The Defense of Rationality in the Early English Enlightenment (Princeton: Princeton University Press, 1996), 184–219; R. A. Knox, Enthusiasm: A Chapter in the History of Religion (Oxford: Clarendon Press, 1950).

the argument from sentimentalism 1

75

works suggest that this sort of misreading of one’s own sentiments is much more widespread, and the logic of Hume’s position dictates that all, or nearly all, of those who take themselves to experience passions directed toward the divinity must be guilty of some such confusion. In Hume’s writings, there are two basic ways of explaining away passions that appear to be directed toward the ﬁrst cause. First, he deploys a psychological theory of displacement, according to which many instances of religious sentiment are in fact cases of misinterpreted feelings about matters of mundane everyday life. And, second, he argues that many of the passions we mistakenly take to be focused on the original cause of all are in fact directed toward more readily visualized and comprehensible substitutes, either ﬁctitious or real: anthropomorphic demigods, ﬁnite angels, and saints—or even statues, paintings, and the sensible forms of ritual and ceremony. Further, each of these particular diagnoses is embedded within Hume’s broader account of ‘religious hypocrisy’: a tendency he ﬁnds among committed believers to engage in wishful thinking and self-deception regarding the character of their own beliefs and passions. I take each of these points in turn. When Hume writes in the letter to Mure that ‘I am afraid that all Enthusiasts mightily deceive themselves. Hope & Fear perhaps agitate their Breast when they think of the Deity’ (NL 13), he is drawing on a theory of displacement sketched two years earlier in the essay ‘Of Superstition and Enthusiasm’ in the ﬁrst volume of Essays, Moral and Political (1741). In that essay two common personality types, characterized by an excess of fear and self-doubt on the one side, and pride and self-assertion on the other, lead to two sorts of self-deceiving ‘false religion’ (E 73) and the transfer of feelings about relatively mundane objects to imaginary supernatural powers.²⁶ The mind of man is subject to certain unaccountable terrors and apprehensions, proceeding either from the unhappy situation of private or public affairs, from ill health, from a gloomy and melancholy disposition, or from the concurrence of all these circumstances. In such a state of mind, inﬁnite evils are dreaded from unknown agents; and where real objects of terror are ²⁶ This essay distinguishes fearful ‘superstition’ from prideful ‘enthusiasm,’ whereas in the letter to Mure Hume is apparently using ‘enthusiasm’ in a broader sense, covering both the anxious and the assertive forms of false religion.

76 the argument from sentimentalism 1 wanting, the soul, active to its own prejudice, and fostering its predominant inclination, ﬁnds imaginary ones, to whose power and malevolence it sets no limits. (E 73–4) But the mind of man is also subject to an unaccountable elevation and presumption, arising from prosperous success, from luxuriant health, from strong spirits, or from a bold and conﬁdent disposition. In such a state of mind, the imagination swells with great, but confused conceptions, to which no sublunary beauties or enjoyments can correspond. Every thing mortal and perishable vanishes as unworthy of attention. And a full range is given to the fancy in the invisible regions 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. Hence arise raptures, transports, and surprising ﬂights of fancy; and conﬁdence and presumption still encreasing, these raptures, being altogether unaccountable, and seeming quite beyond the reach of our ordinary faculties, are attributed to the immediate inspiration of that Divine Being, who is the object of devotion. In a little time, the inspired person comes to regard himself as a distinguished favourite of the Divinity; and when this frenzy once takes place, which is the summit of enthusiasm, every whimsy is consecrated. (E 74)

So mundanely generated hopes or fears (according to individual temperament and circumstance, aided and abetted, no doubt, by the sympathetically registered temper of the local culture and religious traditions) are transferred to a fanciful spiritual cosmology. Feelings of angst and depression metastasize into feelings about an order of hostile spirits or the indignation of a celestial patriarch; feelings of pride and self-importance become a trust in tutelary spirits and a sanctifying sense of heavenly patronage (E 74). These are the ‘spectres of false divinity’ projected by the restless mind and its wheel of inner passions, a mundus imaginalis thrown up by the psyche’s deeper emotional currents (NHR 14.8; compare also DNR 12.29).²⁷ While this confabulated realm of spirits does not of course cause our conative and affective attitudes, nothing in ‘Of Superstition and Enthusiasm’ rules out the ²⁷ There are some similarities here with Hobbes’s psychopathology of much religious emotion, as when he suggests that a nagging ‘perpetual fear [of ‘death, poverty, or other calamity’], always accompanying mankind in the ignorance of causes (as it were in the dark), must needs have for object something’ and thereby leads to the ‘feigning of as many Gods, as there be men that feign them.’ Thomas Hobbes, Leviathan, ed. Edwin Curley (Hackett: Indianapolis, 1994), 64 (chapter 12, paragraph 6).

the argument from sentimentalism 1

77

possibility that our passions are really about such a domain. But the diagnosis of Hume’s letter implies that we may well be mistaken on this second point: ‘Enthusiasts’ take their hopes and fears to be focused on an otherworldly spiritual order, but herein they ‘mightily deceive themselves.’ The suggestion is that they misinterpret the ultimate objects of their own sentiments. They think that they fear hostile spirits or the divinity, perhaps, but at bottom the fear that grips them is really anxiety over something more mundane, like abandonment, failure, or declining health. Or perhaps they take themselves to be experiencing sentiments of adoration or gratitude toward the original cause of all, while all the time they are really moved by veiled feelings of self-satisfaction, delight in nature, or perhaps communal, familial, or sexual love.²⁸ Hume’s second explanation of religious feeling is suggested by his comment that some may ‘degrade [God] into a Resemblance of themselves, & by that means render him more comprehensible’ (NL 13). Given the difﬁculty of directing our passions toward a being or principle that deﬁes adequate sensory and imaginative representation, our sentiments tend to ﬁx on real or ﬁctitious beings that are more readily visualized and, given the anthropomorphizing tendencies of the imagination, often rendered in human or quasi-human form. In other texts Hume also suggests further sorts of sensible or imaginable objects that ²⁸ In a letter of October 15, 1754 to the literary scholar and critic Joseph Spence, Hume suggests that ‘mystics’ are often experiencing forms of sexual passion when they speak of adoration of the divine, but can be persuaded through an association of ideas that their feelings are genuinely religious: ‘The mystics certainly have associations by which their discourse, which seems jargon to us, becomes intelligible to themselves. I believe they commonly substitute the feelings of a common amour, in the place of their heavenly sympathies: and if they be not belied, the type is very apt to engross their hearts, and exclude the thing typiﬁed’ (L i. 201). The force of sublimated sexual passion also seems to be at work in the case of Joan of Arc, who is presented in an extended narrative in Hume’s History of England as a particularly vivid example of ‘that visionary and enthusiastic spirit’ whose religious affections rise and fall with terrestrial fortunes (HE ii. 404, see also 397–9, 409). ‘[T]he peculiar character of Charles [the ‘young prince’ of France], so strongly inclined to friendship and the tender passions, naturally rendered him the hero of that sex, whose generous minds know no bounds in their affections.’ Faced with the distress of besieged Orleans, Joan was thus ‘inﬂamed by the general sentiment, was seized with a wild desire of bringing relief to her sovereign in his present distresses. Her unexperienced [i.e. sexually naïve?] mind, working day and night on this favourite object, mistook the impulses of passions for heavenly inspirations: and she fancied, that she saw visions, and heard voices, exhorting her to re-establish the throne of France, and to expel the foreign invaders’ (HE ii. 398–9, my emphases).

78 the argument from sentimentalism 1 the sentiments of devotees might in fact be focused on, including ‘the ceremonial & ornamental Parts of . . . Worship’ and ‘pictures, postures, vestments, buildings’: all ‘sensible, exterior observances, which might occupy [the mind] during religious exercises.’²⁹ A worshipper’s sentiments might then be tracking physical or mental representations of saints, anthropomorphic demigods, or the protagonists of sacred history—or simply the readily available sensible phenomena of statuary, iconography, ceremony, sermon, and song, all of which might be invested with resonant emotional signiﬁcance through the association of ritual and worship with the highs and lows of personal and communal life, and the rhythm of a sacred calendar. Religious passions will often be focused on this sort of ritualized imagery and its associated play of ideas, rather than tracking the ultimate cause or organizing principle itself. Of course, one could always produce a physical or a mental picture of (say) an enthroned patriarch and call it a representation of the ﬁrst cause of all. But Hume will insist that insofar as this sort of imagery fails to give greater accuracy and speciﬁcity to the highly abstract cosmological thought by which we actually refer to this ultimate being or principle (i.e. the ‘relative’ or relational idea of something that stands to the universe as an ultimate originating cause or organizing principle), it cannot succeed in directing our emotions toward it—and further that, given the mechanisms governing human psychology, the emotions of believers inevitably end up focusing on the concrete imagery (the sovereign patriarch, the pillar of ﬁre, the Madonna and child) at the expense of the more abstract and elusive thought that all this is supposed to signify the primary cause of all.³⁰ ²⁹ See the suppressed preface to the second volume of the History of England (cited below in note 30) and HE v. 459–60 (quoted below in note 31). ³⁰ Hume’s letter to Mure underlines the point: ‘all wise Men have excluded the Use of Images & Pictures in Prayer; tho they certainly enliven Devotion; because tis found by Experience, that with the vulgar these visible Representations draw too much towards them, & become the only Objects of Devotion’ (NL 14). There is another general statement of the pressures forcing us toward idolatry in a combative preface that Hume drafted in 1756 for the second volume of the History but ultimately suppressed: ‘The Idea of an Inﬁnite Mind, the Author of the Universe seems at ﬁrst Sight to require a Worship absolutely pure, simple, unadorned; without Rites, Institutions, Ceremonies; even without Temples, Priests, or verbal Prayer or Supplication; Yet has this Species of Devotion been found to degenerate into the most dangerous Fanaticism. When we have recourse to the aid of the Senses & Imagination, in order to adapt our Religion, in some degree to human Inﬁrmity; it is very difﬁcult, & almost impossible, to prevent altogether the Intrusion of Superstition, or keep

the argument from sentimentalism 1

79

To help ﬂesh out this interpretation, we might consider the Natural History’s account of what Hume styles ‘the gross polytheism of the vulgar’ (NHR 5.1), including cults of saints and icons, as well as more explicitly polytheistic religions: however strong men’s propensity to believe invisible, intelligent power in nature, their propensity is equally strong to rest their attention on sensible, visible objects; and in order to reconcile these opposite inclinations, they are led to unite the invisible power with some visible object. (NHR 5.2) as an invisible spiritual intelligence is an object too reﬁned for vulgar apprehension, men naturally afﬁx it to some sensible representation; such as either the more conspicuous parts of nature, or the statues, images, and pictures, which a more reﬁned age forms of its divinities. (NHR 5.9)

Similarly, both in the Treatise and the ﬁrst Enquiry Hume stresses the role of sensible imagery in determining what are often taken to be attitudes toward God—and here he is quite ready to appeal to his eighteenth-century Protestant audience’s conception of Roman Catholicism as a tissue of idolatry and priestcraft: The devotees of that strange superstition usually plead in excuse of the mummeries, with which they are upbraided, that they feel the effect of those external motions, and postures, and actions, in enlivening their devotion, and quickening their fervour, which otherwise wou’d decay away, if directed entirely to distant and immaterial objects. We shadow out the objects of our faith, say they, in sensible types and images, and render them more present to us by the immediate presence of these types, than ’tis possible for us to do, merely by an intellectual view and contemplation. (T 1.3.8.4; virtually the same passage occurs at EHU 5.16; compare also T 3.2.4.2 and HE iv. 14)

However, notwithstanding his tactical focus on Catholicism in the Treatise, in Hume’s view all popular religions, including the strictest and most self-conscious forms of monotheism, are prone to the same tendency toward idolatry. Although the major monotheistic traditions have attempted to elevate their divinity to an incomprehensible perfection beyond imaginative representation, here pious language simply outpaces the possibilities of human thought. The conceptions Men from laying too great Stress on the ceremonial & ornamental Parts of their Worship.’ (Quoted in Mossner, The Life of David Hume, 307.) Compare also Hobbes, Leviathan, 441–2 (chapter 45, paragraph 10).

80 the argument from sentimentalism 1 of believers cannot keep up with their own exaggerated panegyric as they ‘invent new strains of adulation’ and ‘pompous epithets of praise,’ ultimately ‘beget[ting] the attributes of unity and inﬁnity, simplicity and spirituality’ (NHR 6.5, 8.2; compare also NHR 7.1, 13.2, and DNR 11.1). Thus we may observe, that the assent of the vulgar is, in this case, merely verbal, and that they are incapable of conceiving those sublime qualities, which they seemingly attribute to the Deity. Their real idea of him, notwithstanding their pompous language, is still as poor and frivolous as ever. (NHR 7.1)

The result is that popular monotheism marries a ‘merely verbal’ assent to the existence of a scarcely conceivable ultimate being together with the inevitable and inexorable psychological pressure focusing the actual thoughts of believers (and their attendant emotions) on sensible or imaginable forms. Even the strictest laws of Moses and Mohammed could not prevent their followers’ conceptions from drifting back toward the sensible (NHR 8.2). Popular monotheism thus experiences a ‘ﬂux and reﬂux’ between the attempt to approximate a pure ideal of abstract theism on the one hand and the concrete reality of backsliding idolatry on the other: The feeble apprehensions of men cannot be satisﬁed with conceiving of their deity as a pure spirit and perfect intelligence; and yet their natural terrors keep them from imputing to him the least shadow of limitation and imperfection. They ﬂuctuate between these opposite sentiments. The same inﬁrmity still drags them downwards, from an omnipotent and spiritual deity, to a limited and corporeal one, and from a corporeal and limited deity to a statue or visual representation. The same endeavour at elevation still pushes them upwards, from the statue or material image to the invisible power to an inﬁnitely perfect deity, the creator and sovereign of the universe. (NHR 8.2)

This is also the sting in the tail of the anti-Platonist passage in ‘The Sceptic.’ Recall that in that essay Hume—or at least his character ‘The Sceptic’—argued that the ultimate ﬁrst cause or organizing principle of natural religion is ‘an abstract, invisible object [that] cannot long actuate the mind, or be of any moment in life’ even for the most contemplative philosopher. The corollary was that in order for religious sentiments to take hold ‘we must ﬁnd some method of affecting the senses and imagination, and must embrace some historical as well as philosophical account of the divinity’ (E 167). The passions

the argument from sentimentalism 1

81

of believers cannot track the remote cosmological posit presented by natural theology, and must focus instead on saints, prophets, demigods, and spiritual kings—immanent, visualizable beings that act within the created universe—or even simply on icons, music, song, ritual, and ceremony. ‘Popular superstitions and observances are even found to be of use in this particular’ (E 167). Hume’s History of England provides us with a speciﬁc historical example of monotheism’s inescapable drift toward an idolatry of sensible forms. As we saw in Section 3.2 above, in Hume’s account the fanatical iconoclasts in control of the early seventeenth-century Scottish Kirk attempted to establish a puritanical form of worship ‘little suitable to human frailty’: ‘one that borrowed nothing from the senses; but reposed itself entirely on the contemplation of that divine Essence, which discovers itself to the understanding only.’ According to Hume, the result of this ‘most naked’ form of worship is in practice nothing but a solipsistic and desocializing turmoil of inner passions (HE v. 68).³¹ But the ﬁnal irony in Hume’s darkly humorous narrative of Scottish iconoclasm is that, as with other attempts to banish graven images, idolatry inevitably creeps in through the back door. [M]usic was grating to the prejudiced ears of the Scottish clergy; sculpture and painting appeared instruments of idolatry; the surplice was a rag of popery; and every motion or gesture, prescribed by the liturgy, was a step toward that spiritual Babylon, so much the object of their horror and aversion. Every thing was deemed impious, but their own mystical comments on ³¹ For this reason, Hume is quite ready to condone the use of iconography and ritual in popular religion, if only to calm and stabilize the religious spirit of the masses. Consider, for example his (perhaps rather surprising) defense of Archbishop William Laud, the ferocious persecutor of the Puritans under Charles I and great champion of imagery and ceremonial practice within Anglicanism: ‘Whatever ridicule, to a philosophical mind, may be thrown on pious ceremonies, it must be confessed, that, during a very religious age, no institutions can be more advantageous to the rude multitude, and tend more to mollify that ﬁerce and gloomy spirit of devotion, to which they are subject. Even the English church, though it had retained a share of popish ceremonies, may justly be thought too naked and unadorned, and still to approach too near the abstract and spiritual religion of the puritans. Laud and his associates, by reviving a few primitive institutions of this nature, corrected the error of the ﬁrst reformers, and presented to the affrightened and astonished mind, some sensible, exterior observances, which might occupy it during its religious exercises, and abate the violence of its disappointed efforts. The thought, no longer bent on that divine and mysterious essence, so superior to the narrow capacities of mankind, was able, by means of the new model of devotion, to relax itself in the contemplation of pictures, postures, vestments, buildings’ (HE v. 459–60; compare also v. 558).

82 the argument from sentimentalism 1 the Scriptures, which they idolized, and whose eastern prophetic style they employed in every common occurrence. (HE v. 69)

With music, image, ceremony, and ritual banished, the hapless Puritans inevitably latch onto the last remaining sensible forms, falling into an open bibliolatry and consecrating their own mystical and exotic form of preaching. For most of us at least, the human need to ﬁx emotion on concrete sensible objects is simply irresistible.³² So Hume explains away apparent cases of passions directed toward the divine both with a theory of displacement, and with the suggestion that many of the attitudes that we naively take to be directed toward the ﬁrst cause of all in fact turn out to focus on more comprehensible and readily imaginable substitutes. (No doubt both sorts of error could be at work in any given episode of religious emotion.) In addition to these two more speciﬁc diagnoses, Hume also suggests that believers tend to hide the true character of their feelings from themselves through a species of wishful thinking or self-deceit that in his judgment permeates religious belief and practice. Here he draws on his familiar charge that pious world views are mired in self-deception: a ‘religious hypocrisy,’ which ‘being generally unknown to the person himself, though more dangerous . . . implies less falsehood than any other species of insincerity’ (HE vi. 142).³³ Hume’s psychopathology of religious sentiment is in the end quite similar to his unsparing treatment of religious belief: We may observe, that, notwithstanding the dogmatical, imperious style of superstition, the conviction of the religionists, in all ages, is more affected than real, and scarcely ever approaches, in any degree, to that of solid belief and persuasion, which governs us in the common affairs of life. Men dare not avow, even to their own hearts, the doubts which they entertain on such subjects: They make a merit of implicit faith; and disguise to themselves their real inﬁdelity, by the strongest asseverations and most positive bigotry. But nature is too hard for all their endeavours, and suffers not the obscure, glimmering light, afforded in those shadowy regions, to equal the strong ³² On Hume’s account of the psychological pressures moving religion toward idolatry, see Donald T. Siebert, ‘Hume on Idolatry and Incarnation,’ Journal of the History of Ideas 45 (1984), 379–96. ³³ On Hume’s category of ‘religious hypocrisy’ see Jennifer Herdt, Religion and Faction in Hume’s Moral Philosophy (Cambridge: Cambridge University Press, 1997), chapter 5, esp. 197–206.

the argument from sentimentalism 1

83

impressions, made by common sense and by experience. The usual course of men’s conduct belies their words, and shows, that their assent in these matters is some unaccountable operation of the mind between disbelief and conviction, but approaching much nearer to the former than to the latter. (NHR 12.15; compare also T 1.3.9.13, EHU 10.29, EPM 3.38 note, NHR 15.7)

Just as people manage to persuade themselves that they have genuine religious beliefs despite the fact that their conviction is ‘more affected than real,’ so (Hume holds) people also persuade themselves that they experience emotional attitudes toward a transcendental ﬁrst cause of all, when their feelings are in fact about quite different objects.³⁴ So while the heart may have its reasons of which reason knows nothing, those reasons are not always transparent to the believer himself, and we should be wary of theologies that assume otherwise.

3.5. Conclusion In this chapter we took our ﬁrst look at the argument from sentimentalism, and examined its crucial ﬁrst premise: principle S1, the thesis that the ﬁrst cause or designer is not a natural object of any human passion. ³⁴ This helps to explain Hume’s inﬂammatory charge that religious devotion can lead to a corrupting practice of dissimulation. ‘Many religious exercises are entered into, where the heart, at the time, feels cold and languid: A habit of dissimulation is by degrees contracted: And fraud and falsehood become the predominant principle’ (DNR 12.17; see also the notorious note on priests in ‘Of National Characters’, E 199–200). According to Hume, since religious emotion and religious belief often rest on a form of self-deception, these attitudes are also relatively unstable and prone to the shock of sudden self-awareness. Religious zeal then becomes a form of pre-emptive defense against criticism and the discomfort of self-doubt: ‘the theological animosity, so ﬁerce and violent, far from being an argument of men’s conviction in their opposite sects, is a certain proof, that they have never reached any serious persuasion with regard to these remote and sublime subjects. Even those who are most impatient of contradiction in other controversies, are mild and moderate in comparison of polemical divines; and wherever a man’s knowledge and experience give him a perfect assurance in his own opinion, he regards with contempt, rather than anger, the opposition and mistakes of others. But while men zealously maintain what they neither clearly comprehend, nor entirely believe, they are shaken in their imagined faith, by the opposite persuasion; and vent on their antagonists that impatience, which is the natural result of so disagreeable a state of the understanding’ (HE iii. 431–2). (Hume puts this analysis in the mouth of Elizabethan advocates of religious toleration, but there can be little doubt that he endorses these sentiments himself.)

84 the argument from sentimentalism 1 Hume’s commitment to this principle is clear enough, as is its connection to his underlying account of the mechanics of the passions, and in particular the role of sense and imagination in controlling the objects of our conative and affective attitudes. Hume does argue against religion of the heart, just as he argues against religion of the understanding. As Hume indicates in the letter to Mure, principle S1 threatens an entire range of traditional religious commitments and practices. It challenges the view that we ought to feel emotions such as love, reverence, gratitude, and contrition toward the ﬁrst cause of all. It challenges any practice of devotion and prayer, insofar as this is intended to involve the expression of affective attitudes toward the deity.³⁵ It will also compromise any aesthetic appreciation of God, insofar as aesthetic appreciation involves experiencing feelings and sentiments toward the object of evaluation. Overall, principle S1 endangers the whole conative and affective dimension of religious practice, threatening to leave us with a purely intellectual religion devoid of all emotional import.³⁶ For our purposes, however, the particular threat of interest presented by S1 is its role in the argument from sentimentalism: the way in which S1 interacts with Hume’s sentimentalist theory of moral properties to challenge the moral status of God. In the next chapter, I complete the exposition of this overall argument for moral atheism and consider its merits.

Appendix Sense, Imagination, and Religious Passions: The Early Modern Background Although thesis S1 would have shocked the Christian culture of Hume’s eighteenth-century Europe, much of the background framework of assumptions he draws on in order to arrive at the principle ³⁵ This puts a new light on Hume’s approving quotation from Seneca in the closing paragraphs of the Dialogues: ‘To know God, says Seneca, is to worship him. All other worship is indeed absurd, superstitious, and even impious’ (DNR 12.32). ³⁶ If there are any affective states that are not directed toward God but still deserve to be called religious—perhaps including non-intentional feelings of peace or beatitude reached through chanting or meditation, for instance—then these will fall outside the scope of Hume’s critique. See Ninian Smart, The Concept of Worship (Edinburgh: Macmillan, 1972), 24f., for an analysis of the difference between meditative states and the intentional attitudes involved in worship. Compare also Peter C. Appleby, ‘On Religious Attitudes,’ Religious Studies 6 (1970), 359–68: 364.

the argument from sentimentalism 1

85

was in fact settled doctrine among early modern moralists and theologians. In this appendix (which can be passed over without cost to my main interpretive argument) I connect Hume’s own account to this wider background of seventeenth- and eighteenth-century thought about the preconditions for genuine religious passions. Many early modern thinkers agreed that we need sufﬁciently rich representations of an object by way of our cognitive faculties if our passions are to be directed toward that object—and this condition, moreover, was explicitly applied to religious passions. Thus when addressing the speciﬁc topic of religious emotion, the seventeenthcentury mystical poet and divine Thomas Traherne reports that it is ‘a maxim in the schools’ that ‘there is no love of a thing unknown’: that according to the teaching of the universities, some knowledge of the deity is required for genuine religious sentiment.³⁷ At the same time, the mainstream position among these same scholastics was the empiricist view, endorsed by Aristotle and St Thomas, that all our cognitive representations, including those of the original cause, are ultimately derived from the representations afforded by our natural sensory faculties, and thus are fundamentally imagistic in character.³⁸ In combination with one another, these two traditional scholastic doctrines entail the principle, much insisted on by Hume, that our passions can only be directed toward the deity if we have adequate imagistic representations of that original being.³⁹ This sort of principle also seems to have been taken up in the wider educated culture outside the universities. For instance, the main ³⁷ Thomas Traherne, Centuries of Meditations (London: P.J. & A.E. Dobell, 1950), 3 (century 1, section 2) (original emphasis). Rather in the spirit of Hume’s character Demea, Traherne himself rejects the maxim, allowing a form of mystical adoration of the deity unmediated by cognitive representations or any discursive knowledge of this ultimate cause of all. But this would have smacked too much of irrationalism or enthusiasm for most early modern philosophers. ³⁸ Aristotle, De Anima 432a7, in The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941), 595. Thomas Aquinas, Summa Contra Gentiles, 4 vols (New York: Benziger, 1924), iii., 108–111 (book 3, chapter 47); Summa Theologica, 3 vols (New York: Benziger, 1947–8), i. 58–9 (part 1, question 12, articles 11–12). ³⁹ For a useful account of theories of the relationship between mental imagery and the production of the passions in the tradition of rhetoric from Cicero and Quintilian through to Renaissance thinkers like Ben Jonson and Francis Bacon, see Quentin Skinner, Reason and Rhetoric in the Philosophy of Hobbes (Cambridge: Cambridge University Press, 1996), ch. 5.

86 the argument from sentimentalism 1 theme of John Donne’s poem Aire and Angels, a century before Hume, is that romantic and religious love both need a sensible object in order to engage, and thus that when angels appear to humans they must hide their blinding empyreal brightness and take on a sensible form of woven air: For, nor in nothing, nor in things Extreme, and scatt’ring bright, can love inhere; Then as an Angell, face, and wings Of aire, not pure as it, yet pure doth weare So thy love may be my love’s spheare

Donne here distils a precise body of scholastic doctrine concerning the manifestation of angels,⁴⁰ and presumably he is also borrowing from scholastic accounts of the conditions for love to ‘inhere.’ In any case, the poem suggests that the wider literary culture of the seventeenth century was open to the suggestion that love, at least, needs to be directed by sensible forms or mental imagery. Among seventeenth-century philosophers, it is the strict empiricist Thomas Hobbes who comes closest to Hume, both in adopting the view that adequate mental imagery of an object is a necessary precondition for our passions to engage that object, and in using this principle to cast doubt on the claim that the natural mechanisms of our mind can direct our passions toward God (where, as with Hume, ‘by the name God is meant the cause of the world’⁴¹). Consider the following brisk remarks from Humane Nature (1650), where Hobbes announces that our inability to form an adequate imagistic idea (or ‘conception’) of the deity precludes our feeling love toward this ultimate being: [C]oncerning man’s affections to Godward, they are not the same always that are described in the chapter concerning passions [i.e. Humane Nature chapter 7]. For there, to love is to be delighted with the image or conception of the thing loved; but God is unconceivable; to love God therefore, in the Scripture, is to obey his commandments, and to love one another.⁴² ⁴⁰ For commentary, see E. M. W. Tillyard, The Elizabethan World Picture (London: Chatto and Windus, 1943), 39. ⁴¹ Hobbes, On the Citizen (1642), ed. Richard Tuck and Michael Silverthorne (Cambridge: Cambridge University Press, 1998), ch. 15, section 14. ⁴² Thomas Hobbes, Human Nature (1650), in Hobbes, Human Nature and De Corpore Politico, ed. J. C. A. Gaskin (New York: Oxford University Press, 1994), chapter 11, section 11.

the argument from sentimentalism 1

87

In Hobbes’s ingenious (and possibly mischievous) reinterpretation of scripture, human ‘love’ of God does not so much show itself in obedience to his commands and love for other humans as it is exhaustively constituted by such behavior, for to love God in any literal, straightforward sense is quite impossible. And, like Hume, Hobbes also provides an error theory to account for the fact that people often report experiencing passions that they take to be directed toward the ﬁrst cause of all.⁴³ There is at least one important difference between the two empiricists, however. For Hobbes, the original cause of all must be regarded as an inﬁnite being, and it is this inﬁnitude that prevents us from forming an imagistic conception of it sufﬁcient to support feelings of love.⁴⁴ But Hume does not commit to any view about the ﬁnitude or inﬁnitude of the deity, and in his analysis our inability to form a sufﬁciently rich imagistic idea of God is rather the result of the highly abstract and attenuated character of the thought that this transcendental being or principle (whatever its intrinsic nature) stands in some sort of cause–effect relationship with the entire ordered universe.⁴⁵ There were of course seventeenth-century thinkers who took a radically different view both to Hobbes and to the empiricist tradition within scholasticism, and many would have taken Hobbes’s terse demonstration of the impossibility of genuine human love of God rather as a reductio of its major premise. Thus various philosophers of an anti-empiricist bent insisted that, while our passions are no doubt often directed by sense and imagination (perhaps indeed regrettably often), nevertheless we can also have non-imagistic, ‘intellectual’ ideas that might serve to guide and focus our passions beyond the sphere of the sensible and mundane. The Cambridge Platonist John Smith presents a good example here, both in his complaint that mental imagery often seeps into and distorts religious emotion, and in his insistence that there remains, all the same, the possibility of ⁴³ See note 27 above. ⁴⁴ Thomas Hobbes, Leviathan, 15, 239, 444 (chapter 3, paragraph 12; chapter 31, paragraph 20; chapter 45, paragraph 15); see also Human Nature chapter 11, section 2. Compare also Hobbes’s Third Set of Objections to Descartes, collected in Ren´e Descartes, The Philosophical Writings of Descartes, tr. and ed. John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, 3 vols (Cambridge: Cambridge University Press, 1984–91), ii. 121–37: 127. ⁴⁵ As we saw above in Section 3.3.

88 the argument from sentimentalism 1 genuine religious emotion guided by higher faculties than sense and imagination: There are such things in our Christian religion, when a carnal, unhallowed mind takes the chair, and gets the expounding of them, that may seem very delicious to the ﬂeshy appetites of men; some doctrines and notions of free grace and justiﬁcation, the magniﬁcent titles of sons of God and heirs of Heaven, ever ﬂowing streams of joy and pleasure that blessed souls shall swim in to all eternity, a glorious Paradise in the world to come, always springing up with well-scented and fragrant beauties, a new Jerusalem paved with gold, and bespangled with stars, comprehending in its vast circuit such numberless varieties, that a busy curiosity may spend itself about to all eternity. I doubt not but that sometimes the most ﬂeshy and earthy men, that ﬂy in their ambition to the pomp of this world, may be so ravished with the conceits of such things as these, that they may seem to be made partakers of the powers of the world to come. . . . [A]s the motions of our sense and fancy and passions, while our souls are in this mortal condition, sunk down deeply into the body, are many times more vigorous, and make stronger impressions upon us, than those of the higher powers of the soul, which are more subtle, and remote from these mixed animal perceptions: that devotion which is there seated, may seem to have more energy in it, than that which gently, and with a more delicate kind of touch, spreads itself upon the understanding . . . [But] True religion is no piece of artiﬁce; it is no boiling up of the imaginative powers, nor the glowing heats of passion; though these are too often mistaken for it, when in our jugglings in religion we cast a mist before our own eyes.⁴⁶

Similar remarks can be found in anti-empiricist philosophers like Pascal and Malebranche, where the key claim that non-imagistic spiritual perceptions or pure rational intellect can direct the passions toward non-sensible objects is again balanced with an admission that the representations of sense and imagination usually govern our conative and affective attitudes.⁴⁷ Eighteenth-century writers also addressed the preconditions for genuine religious passions and the dependence of such sentiments on ⁴⁶ John Smith, Select Discourses (London: 1660), 370–2. ⁴⁷ ‘[W]hat [the soul] perceives by the senses affects and stirs it greatly. What it knows through the imagination affects it much less. But what the understanding presents to it, i.e., what it perceives by itself, or independently of the senses and imagination, hardly stirs it at all,’ Nicolas Malebranche, The Search After Truth (1674–5), tr. and ed. Thomas M. Lennon and Paul J. Olscamp (Cambridge: Cambridge University Press, 1997), 79–81. (Hume was of course familiar with this inﬂuential work.) Compare also Malebranche, Treatise on Nature and Grace (1680), tr. and ed. Patrick Riley (Oxford: Clarendon Press, 1992), 181–4 (dialogue 3, section 18–24); Blaise Pascal, Pens´ees (1670), tr. A. J. Krailsheimer (London: Penguin, 1995), 9–10 (section 44 in the Lafuma numbering, 90 in the Sellier numbering).

the argument from sentimentalism 1

89

the representations of our cognitive faculties. Thus Francis Hutcheson states ﬁrmly in his Essay on the Nature and Conduct of the Passions and Affections (1728) that ‘there can be no Affection present to the Mind, toward any Object, while the Idea of it is not present’ and accordingly that No Object which is entirely unknown, or of which we have no Idea, can raise Affection in the best Temper; consequently want of affection to an unknown Object evidences no evil. This would be the Case of those who never heard even the Report of a DEITY, if ever there were any such.⁴⁸

The nonconformist minister Isaac Watts provides another pertinent example from the generation before Hume, writing in his popular and much-reprinted devotional guidebook Discourses of the Love of God and its Inﬂuence on all of the Passions (1729) that Where the understanding has but a poor and scanty furniture of the things of God, the pious affections will have the fewer springs to raise them: And if our ideas of divine things are obscure and confused, our passions are in great danger of running wildly astray, and being led away by every delusion. Seek therefore, not only a large and plenteous acquaintance with the things of God, but endeavour, as far as possible, to get clear and distinct conceptions of them, that the pious passions may have solid ground whence to take their rise.⁴⁹

Hume fully agrees with the diagnosis of the ﬁrst sentence, although of course he has doubts about the viability of Watts’s proposed cure. We simply cannot hope for ‘a large and plenteous acquaintance with the things of God,’ and certainly not ‘clear and distinct conceptions’ of such matters. For Hume, the consequence is that we do in fact lack Watts’s ‘springs to raise’ passions focused on the divinity. Perhaps the most instructive comparison with Hume on imagery, the natural limits of the passions, and the possibility of religious sentiment in his contemporary eighteenth-century context is presented by the American philosopher and theologian Jonathan Edwards in ⁴⁸ Francis Hutcheson, An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense (1728), ed. Aaron Garrett (Indianapolis: Liberty Fund, 2002), 191, 195. ⁴⁹ Isaac Watts, Discourses of the love of God, and its inﬂuence on all the passions (1729) (Philadelphia: Woodward, 1799), 190–1 (discourse 7, paragraph 3). For commentary on Watts (and other moderate dissenters) on ‘affectionate religion,’ see Isabel Rivers, Reason, Grace and Moral Sentiment: a study of the language of religion and ethics in England, 1660–1780 (Cambridge: Cambridge University Press, 1991–2000), i., ch. 4.

90 the argument from sentimentalism 1 his ﬁrst major work, A Treatise Concerning the Religious Affections.⁵⁰ Published in 1746, just a few years after Hume’s Treatise, Essays, and letter to Mure, Edwards’s Calvinist tract is of course the polar opposite of Hume’s work in its religious orientation and spiritual attitude. But the parallels between Hume’s and Edwards’s accounts are nevertheless remarkable, and help to situate Hume’s critique of religious passions in the context of eighteenth-century thought about pious emotion. (The American and the Scot apparently developed their respective accounts quite independently of one another. Reading Edwards does not disclose Hume’s immediate inﬂuences so much as bring out a wider background of thought that informs each of their systems.⁵¹) Edwards’s Treatise is an assiduous (indeed, rather unremitting) study of the character of ‘Truly Gracious and Holy Affections,’ and an attempt to distinguish such authentic religious emotion from false and delusory enthusiasm.⁵² For Edwards as for Hume, the passions (or ‘affections’) all have an object-directed or intentional character: they always refer us to some being or other.⁵³ Furthermore, our affections can only be directed toward an object insofar as we have representations of that object by way of our cognitive faculties: ‘Such is the nature of man, that it is impossible that his mind should be affected, unless it be by something that he apprehends, or that his mind conceives of.’ For ‘natural man’—that is, man in his natural state unassisted by supernatural inspiration—this means that we need representations by way of the imagination: ‘that power ⁵⁰ Jonathan Edwards, A Treatise Concerning the Religious Affections (1746), ed. John E. Smith, in The Works of Jonathan Edwards, ed. Perry Miller, 23 vols (New Haven: Yale University Press, 1957–), ii. 84–461. John Wesley, the founder of Methodism, produced a popular abridgement of Religious Affections in 1773—although he warned in a preface that the original work contains ‘deadly poison’ as well as ‘much wholesome food.’ See James Harris, Of Liberty and Necessity (Oxford: Clarendon Press, 2005), 130 n. 51; Richard B. Steele, ‘Gracious Affection’ and ‘True Virtue’ According to Jonathan Edwards and John Wesley (Metuchen, N. J.: The Scarecrow Press, 1994). ⁵¹ On the question of Hume’s and Edwards’s inﬂuence on one another (or the lack thereof ), see Jasper Reid, ‘The Metaphysics of Jonathan Edwards and David Hume,’ Hume Studies 32 (2006), 53–-82: 55–7. ⁵² Jonathan Edwards, A Treatise Concerning the Religious Affections, 191. ⁵³ As we saw, for Hume this intentionality is an extrinsic and contingent but law-like feature of the passions: it is a consequence of any given passion’s role in a causal matrix of mental states, including the intrinsically representational idea that that passion brings to mind. (See Section 3.2 above.)

the argument from sentimentalism 1

91

of the mind, whereby it can have a conception, or idea of things of an external or outward nature (that is, of such sort of things as are the objects of the outward senses), when those things are not present, and be not perceived by the senses.’⁵⁴ Bracketing the fact that Edwards’s imagination corresponds purely to outer sense, whereas Hume’s covers representations reﬂecting both outward and inward experience—sensation and reﬂection—Edwards and Hume are otherwise in full agreement.⁵⁵ Given the natural mechanisms of the mind, humans can only direct their passions toward objects that are represented by the senses or the imagination.⁵⁶ And they agree on one more crucial point: given the natural workings of human affective psychology, these imagistic representations are unable to support passions directed toward the original cause of all. And as with Hume’s error theory of pious emotion, this casts doubt on claims to experience passions directed toward the deity when those sentiments are in fact excited by the imagination: [I]mpressions which some have made on their imagination, or the imaginary ideas which they have of God, or Christ, or heaven, or anything appertaining to religion, have nothing in them spiritual, or of the nature of true grace. Though such things may attend what is spiritual, and be mixed with it, yet in themselves they have nothing that is spiritual, nor are they any part of gracious experience. Natural men may have conceptions of many things about spiritual affections; but there is something in them which is as it were the nucleus, or kernel of them, that they have no more conceptions of, than one born blind has of colors.

Edwards develops the point at some length, largely in order to drive home his own attack on the enthusiast’s claims to experience genuine religious emotion: I have insisted largely on this matter, because it is of great importance and use, evidently to discover and demonstrate the delusions of Satan, in many kinds of ⁵⁴ Ibid., 271, 210–11. On Edwards’s account of the imagination, see San Hyun Lee, The Philosophical Theology of Jonathan Edwards (Princeton: Princeton University Press, 1988), 115–69. ⁵⁵ Here Hume is more closely in line with the Lockean tradition. Locke, Essay, 104–6 (book 2, chapter 1 sections 2–3). Compare also George Berkeley, A Treatise of the Principles of Human Knowledge, ed. Jonathan Dancy (Oxford: Oxford University Press, 1998), 103 (part 1, section 1). ⁵⁶ Edwards, A Treatise Concerning the Religious Affections, 210–11.

92 the argument from sentimentalism 1 false religious affections, which multitudes are deluded by, and probably have been in all ages of the Christian church.⁵⁷

Explanatory appeals to the Devil aside, the ‘American Augustine’⁵⁸ does eventually part company with Hume, for Edwards holds that not all humans are ‘natural men.’ Some of us may be may be assisted by divine grace and lifted above our own natural powers. In this way, God provides the Calvinist saints with spiritual and decidedly nonimagistic representations that enable genuine religious emotion: ‘gracious affections’ that are truly directed toward the ﬁrst cause of all. [I]t is exceeding apparent that [imagistic] ideas have nothing in them which is spiritual and divine, in the sense wherein it has been demonstrated that all gracious affections are spiritual and divine. These external ideas are in no wise of such a sort, that they are entirely, and in their whole nature diverse from all that men have by nature, perfectly different from, and vastly above any sensation which ’tis possible a man should have by any natural sense or principle, so that in order to have them, a man must have a new spiritual and divine sense given to him. [I]n those gracious exercises and affections which are wrought in the minds of the saints, through the saving inﬂuences of the Spirit of God, there is a new inward perception or sensation of their minds, entirely different in its nature and kind, from anything that ever their minds were the subjects of before they were sanctiﬁed.⁵⁹

Through extraordinary divine intervention, the Calvinist saint is thus endowed with spiritual ideas entirely different from the deliveries of natural sensation and reﬂection, ideas that permit his passions to range beyond the sphere of sensible and imaginable beings.⁶⁰ He becomes, in effect, the sort of visionary, quasi-divine being presented in the anthropology of Christian Platonists like Malebranche and John Smith. So for Edwards a narrow elect can readily experience sentiments directed toward the ﬁrst cause of all. But for Hume there is no such dispensation of supernatural sensory powers. Our natural ⁵⁷ Edwards, A Treatise Concerning the Religious Affections, 210, 205, 210. ⁵⁸ Avihu Zakai, Jonathan Edwards’ Philosophy of History (Princeton, NJ: Princeton University Press, 2003), 14–16. ⁵⁹ Edwards, A Treatise Concerning the Religious Affections, 213, 205. ⁶⁰ Norman Feiring, Jonathan Edwards’s Moral Thought in its British Context (Chapel Hill, NC: The University of North Carolina Press, 1981), 123–9; Thomas Dixon, From Passions to Emotions (Cambridge: Cambridge University Press, 2003), 79–80.

the argument from sentimentalism 1

93

faculties of sense, reﬂection, and imagination provide the only ideas we have, and in the normal operation of a natural human mind the passions simply will not react to this sort of ultimate cosmological being.⁶¹ ⁶¹ The case of Edwards and Hume on religious affect interestingly parallels the case of Edwards and Hume on personal identity, free will, and causation. Thus according to Jasper Reid, ‘although the two men developed their theories in isolation from one another, their minds were nevertheless following almost identical paths on several of the most central issues in metaphysics (including the natures of body and mind, personal identity, causation, and free will). . . . [However] wherever Hume came to rest in a skeptical position, Edwards would initially approach the very same position, but would then pull back at the last minute and bring in God to ﬁll the gaps, yielding a Christian system of philosophy with a distinctly Humean ﬂavour’ (Jasper Reid, ‘The Metaphysics of Jonathan Edwards and David Hume,’ 53).

This page intentionally left blank

4 The Argument from Sentimentalism 2: Religious Passions and the Deity’s Moral Status 4.1. Introduction In combination with his sentimentalist account of moral distinctions, Hume’s thesis that the ﬁrst cause or designer is not a natural object of any human passion places this being outside the domain of morally assessable agents. Recall the basic structure of the argument: The argument from sentimentalism S1. The deity is not a natural object of any human passion. S2. Moral sentiments are a species of human passion. S3. If a being is not a natural object of the moral sentiments, then it cannot have moral attributes (either virtues or vices). Therefore S4. The deity cannot have moral attributes (either virtues or vices). In the previous chapter I documented Hume’s commitment to principle S1 and explored his underlying reasons for locating the original cause or designer beyond the natural reach of our passions. I now complete the exposition by examining premises S2 and S3, and the way in which they conspire with S1 to remove the deity from the moral domain.

96 the argument from sentimentalism 2 The chapter proceeds as follows. In Section 4.2, I document Hume’s theses S2 and S3 and the role they play in the overall argument from sentimentalism. I also argue that Hume did connect his sentimentalist theory of morals to his claim that the deity lies outside the natural compass of the human passions, if only implicitly. Although Hume never openly draws the scandalous conclusion S4, he does seem to hint at the point in his private correspondence, implying that his principles lead to just such a form of moral atheism. I then examine possible objections to the argument from sentimentalism, survey its overall strengths and weaknesses (Section 4.3), and take stock in a conclusion (Section 4.4).

4.2. Sentimentalism and the Natural Objects of the Passions The basic outline of Hume’s sentimentalist account of morals is well-known. According to Hume, the world of moral standards is ultimately constituted by characteristic sentiments of approbation and disapprobation that humans naturally feel in response to particular sorts of actions, motives, or character traits. To say that cowardice is a vice is just to say that humans naturally feel characteristic sentiments of disapproval toward this sort of behavioral trait—or at least would feel such sentiments in appropriately reﬂective, calm, and disinterested circumstances. The result is a projectivist picture of ethics, where moral standards are not ﬁxed by some sort of eternal and immutable mind-independent normative order, but rather simply reﬂect the reactive attitudes of approval and blame encoded in human nature. The most probable hypothesis, which has been advanc’d to explain the distinction betwixt vice and virtue, and the origin of moral rights and obligations, is, that from a primary constitution of nature certain characters and passions, upon the very view and contemplation, produce a pain, and others, in like manner excite a pleasure. The uneasiness and satisfaction are not only inseparable from vice and virtue, but constitute their very nature and essence. To approve of a character is to feel an original delight upon its appearance. To disapprove of it is to be sensible of an uneasiness. (T 2.1.7.5; virtually the same text reappears at DP 2.14)

the argument from sentimentalism 2

97

[W]hen you pronounce any action or character to be vicious, you mean nothing, but that from the constitution of your nature you have a feeling or sentiment of blame from the contemplation of it. Vice and virtue, therefore, may be compar’d to sounds, colours, heat, and cold, which, according to modern philosophy, are not qualities in objects, but perceptions in the mind. (T 3.1.1.26; see also T 3.1.2.3, EPM Appendix 1.21)

Moral attributes thus depend on human sentiment: virtues are those traits that naturally produce certain characteristic sentiments of admiration in human beings, and vices are those traits that naturally excite certain characteristic sentiments of disapproval or blame. And these moral sentiments are simply one species of human passion. Hume is in fact quite speciﬁc: a virtue or a vice is a character trait that, given an appropriately reﬂective and disinterested ‘general survey,’ ‘gives rise to our approbation or blame, which is nothing but a fainter and more imperceptible love or hatred’ (T 3.3.5.1; compare also T 2.2.1.7, 3.3.1.3, EHU 8.35). (Love and hatred are, for Hume, broad categories encompassing a wide variety of positive and negative affective responses, including, for instance, sentiments of esteem and respect as well as romantic or familial affection.) So a virtue is a character trait that naturally triggers a particular sort of human affective response, which Hume identiﬁes with a particular form of love.¹ All this underlines Hume’s commitment to principle S2: the moral sentiments are a species of human passion. T 3.3.5.1, quoted above, indicates that the moral sentiments are a ‘fainter and more imperceptible’ form of passion. Hume stresses this in several places, categorizing feelings of moral approval and blame as ´ ¹ P´all Ardal, Passion and Value in Hume’s Treatise (Edinburgh: Edinburgh University Press, 1966), 109–47; Christine Korsgaard, ‘The general point of view: love and moral approval in Hume’s ethics,’ Hume Studies 25 (1999), 3–41; Paul Russell, Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility (Oxford: Oxford University Press, 1995), 61–3. Annette Baier challenges this picture somewhat, suggesting that Hume at least ought not treat the moral sentiments as species of love and hate, since they are directed toward character traits rather than persons. But Baier agrees on the central point of current interest: Hume’s moral sentiments are certainly some sort of passion; Annette Baier, A Progress of Sentiments: Reﬂections on Hume’s Treatise (Cambridge, MA: Harvard University Press, 1991), 134. For a useful summary of the issues, see Jacqueline Taylor, ‘Virtue and the Evaluation of Character,’ in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), 276–95: 283.

98 the argument from sentimentalism 2 ‘calm’ passions, that is, passions that are usually felt faintly and ‘are more known by their effects than by the immediate feeling or sensation’ (T 2.3.3.8).² The Treatise in particular repeatedly emphasizes the calm nature of the moral sentiments, largely in order to help explain the moral rationalist’s tendency to interpret moral judgments (which are in fact ultimately dependent on human sentiment) as the deliverances of pure reason (T 3.1.2.1, 2.3.8.13; compare also EPM 6.15). As Hume puts it in a well-known passage announcing his sentimentalist theory of morals, ‘Morality . . . is more properly felt, than judg’d of; tho’ this feeling or sentiment is commonly so soft and gentle, that we are apt to confound it with an idea’ (T 3.1.2.1). The point is important for our purposes, since Hume goes out of his way in the letter to Mure to emphasize that principle S1 places the deity beyond the reach of our calmer passions as well as the more violent. (‘Please to observe, that I not only exclude the turbulent Passions, but the calm Affections’ (NL 13).) This is unlikely to be a gratuitous remark on Hume’s part. Anyone familiar with the basic principles of the Treatise would know that the moral sentiments are Hume’s leading example of a calm species of passion—and presumably Hume could assume this of his intimate friend and conﬁdante Mure. So in stressing that S1 applies to the calm as well as the strong passions, Hume is likely underlining the implications of S1 for the moral sentiments. Without spelling it out in so many words, he may be pointing to the fact that S1, in conjunction with some of the most conspicuous doctrines of the Treatise, entails that the deity is not a natural object of the moral sentiments. He is, at the least, emphasizing that S1 implies that the deity is not a natural object of any of the calm sentiments, which in his recent Treatise prominently includes responses of moral approval and disapproval. If (as S1 has it) the ﬁrst cause is not the natural object of any human passion, then given Hume’s familiar thesis S2, it follows that this being or principle is not a natural object of the moral sentiments. Just as the mechanisms of human psychology, in their usual or natural operation, will not generate any feelings of gratitude ² In some places Hume uses ‘passion’ to refer narrowly to the more violent form of conative or affective attitude (T 2.1.1.3). But usually he uses ‘passion’ as the generic term covering all conative and affective attitudes, calm as well as violent (T 2.3.8.13, 2.3.6.10, EPM 6.15).

the argument from sentimentalism 2

99

or contrition directed toward the deity, so they will not generate any feelings of moral approval or blame directed toward this original cause.³ We are now ready to connect theses S1 and S2 together with Hume’s ﬁnal premise S3, and tie off the argument from sentimentalism. Given Hume’s sentimentalist moral theory, if a being is not a natural object of the moral sentiments, then it is beyond the sphere of moral assessment. The reason that stones and snowstorms do not have moral virtues or vices is that human affective psychology does not naturally react to stones and snowstorms, however helpful or destructive they are, with the appropriate species of moralized approval or disapproval. There might of course be anomalous cases: perhaps a mad or hopelessly confused person could conceivably experience moral sentiments toward such things.⁴ But still, stones and snowstorms will not excite moral approbation or blame in normal human observers, and so these sorts of things are not natural objects of the moral sentiments. The qualiﬁcation ‘natural’ is then doing real work in Hume’s famous declaration that the human mind, in ‘gilding or staining all natural objects with the colours, borrowed from internal sentiment, raises, in a manner, a new creation’ (EPM Appendix 1.21, my emphasis). Hume is restricting the scope of his projectivist theory, and excluding those beings that are not the natural objects of our various conative and affective attitudes from the ‘new creation’ of projected properties like virtue and vice. And as we can now see, all this applies to the deity. For Hume, if there is a ﬁrst cause or designer of the ordered universe, this being or principle is not the natural object of any human passion, and hence is not a natural object of the moral sentiments. It follows that this ultimate being or principle is beyond moral assessment, and cannot have virtues or vices. Hume’s ³ Hume’s claim in the letter to Mure that the calm passions as much as the violent do not take the deity as a natural object squares with the system of the passions presented in the Treatise. There Hume requires sufﬁciently vivacious mental imagery as ‘a requisite circumstance to the exciting all our passions,’ explicitly including ‘the calm as well as the violent’ (T 2.3.6.10)—and as we have seen, the idea of an ultimate ﬁrst cause or organizing principle responsible for the ordered universe is simply too faint to play this role (Section 3.3). ⁴ See Section 3.2 on the interpretation of ‘natural object’ that allows that the passions might in rare circumstances range beyond their natural objects.

100 the argument from sentimentalism 2 sentimentalism along with his account of the natural objects of the passions entails strong moral atheism.

4.3. Objections and Replies In the remainder of this chapter, I examine three possible objections to the argument from sentimentalism. In my view, the attribution of this argument to Hume is justiﬁed in the light of the various texts highlighted in this and the previous chapter. At the very least, we can say that Hume endorses each of the premises and hints at their connection to one another even if he does not explicitly draw their joint conclusion. With this in mind, I treat the following objections as possible criticisms of an argument of Hume’s. But of course, if any of these objections prove sufﬁciently damaging, they might then suggest that Hume could not really have endorsed the argument from sentimentalism after all. Rather than embarrassing Hume himself, they might tell against my own interpretation of Hume as committed to that argument.

4.3.1. First Objection: The Argument from Sentimentalism Fails to Tell us Anything about the Intrinsic Character of the First Cause Our ﬁrst objection stresses that, insofar as the argument from sentimentalism focuses on human passions and their failure to naturally range out toward the ﬁrst cause of all, it actually says nothing at all about the intrinsic character of that ultimate being or principle. But if the argument says nothing about the intrinsic nature of the divinity, then it cannot rule out that being having traits that look for all the world like virtues and vices. For instance, nothing in the argument from sentimentalism rules out the possibility that the ﬁrst cause or designer is a mind-like being that cares for us, that watches over us, that rewards our good behavior and punishes our evil actions, and so on. The point is that the argument from sentimentalism says more about us than it does about the ﬁrst cause, and thus seems ill-suited to deliver a robust conclusion about the nature of God. The objection is correct so far as it goes, but its limitations also need to be registered. The argument from sentimentalism manifestly

the argument from sentimentalism 2

101

concerns the way in which human sentimental psychology would relate to any original cause of all, rather than addressing the intrinsic nature of this primary being or principle. This is of course no accident. Since Hume doubts our ability to reach any species-speciﬁc knowledge of the intrinsic character of the original cause, any Humean argument for moral atheism must avoid such overreaching commitments. So the argument cannot rule out a ﬁrst cause that watches over us, aims to reward our virtuous behavior, or what you will. But if one takes the projectivist character of Hume’s sentimentalist theory seriously—if one admits the sentimentalist claim that the world of morality, like the world of color, is a reﬂection of the human mind, deﬁned and delimited by our natural human attitudes—Hume’s argument can still show that the deity is quite outside the domain of moral assessment, lacking all possibility of virtue and vice. This is the key point that the current objector is perhaps in danger of underestimating. Consider an analogy: malaria inﬂicts terrible harm on humans, but we do not regard it as morally vicious. Apple trees provide us with food, but we do not regard them as morally virtuous. Even if there were a disease that happened to ‘reward’ virtuous behavior and ‘punish’ vicious behavior (as cirrhosis of the liver ‘punishes’ excessive drinking), we would not regard this disease as a moral agent. The reason, of course, is that only thinking, rational beings are regarded as moral agents, as beings capable of virtue and vice. But the deeper reason for this fact, according to Hume’s sentimentalist theory, is that, as a matter of contingent but deeply ingrained human psychology, our sentiments of moral approbation and disapprobation only respond to thinking, rational beings (EPM 5.1 note). It is just a contingent fact of human nature that we do not stain or gild trees or diseases with moral properties, and that is all there is to it. (To demand that there always be a rational basis for moral distinctions, and to insist that such distinctions cannot simply bottom out in brute facts about human psychological attitudes, is of course simply to abandon sentimentalism for moral rationalism.) And according to the argument from sentimentalism, there is a further limitation on the natural objects of our moral sentiments. Whatever its intrinsic character might be, any being or principle that ﬁts the relational description ‘the ﬁrst cause or organizing principle of the ordered universe’ is not a natural object of the moral sentiments (or indeed of any human passion). So if we accept the

102 the argument from sentimentalism 2 sentimentalist account of virtue and vice provided by Hume’s moral theory, then it is no more appropriate to speculate about the deity’s moral character than it is to ask about the moral attributes of a tree or a disease. It may again be helpful to compare Hume’s treatment of moral properties with the case of secondary qualities. Given the modern theory of secondary qualities, if the human visual system cannot respond to the ﬁrst cause of all, then that being has no color. This is not simply a fact about our ignorance of the deity’s particular color, nor is it consistent with that being having an intrinsic color independently of the way it relates to our human visual system. Only if one rejects the modern theory of secondary qualities for an account that treats colors as intrinsic and mind-independent features of things would it make sense to ask what color the deity is, notwithstanding our own inability to experience it. Of course, philosophers with realist intuitions about moral properties may balk at the suggestion that a being can be said to lack moral attributes quite independently of its own intrinsic nature. If one endorses realist intuitions that treat the existence of moral properties as a metaphysical fact that obtains independently of human attitudes, then the failure of human sentimental psychology to react to the ﬁrst cause will have no impact on its moral status. So the current objection certainly underlines the fact that the Hume’s overall argument trades on his sentimentalist conception of moral attributes as relational and mind-dependent properties (as premise S3 already makes quite plain). But so long as we work within the framework of Hume’s sentimentalist theory of morals, the argument can proceed.⁵

4.3.2. Second Objection: Some Rare Humans Might be Able Experience Passions Directed Toward the Deity Assuming that Hume succeeds in explaining away the apparent cases of passions directed toward the deity among most believers,⁶ there still remains a possible objection drawing on the case of the ‘speculative ⁵ For recent defenses of the sentimentalist theory of morals (in a non-historical mode, but with due deference to Hume), see Simon Blackburn, Ruling Passions (Oxford: Oxford University Press, 1998), and Jesse Prinz, The Emotional Construction of Morals (Oxford: Oxford University Press, 2007). Blackburn develops the theory in an expressivist direction; Prinz in a subjectivist and relativistic direction. ⁶ See Section 3.4.

the argument from sentimentalism 2

103

man’ or Stoic sage admitted in EHU 8.34 and the Platonist devotee of ‘ﬁne genius’ allowed in ‘The Sceptic’ (E 167). Recall that it is possible to interpret Hume’s thesis S1 as asserting, not that passions directed toward the deity are psychologically impossible, but merely that such passions will be extremely rare, and quite out of line with the common or statistically regular workings of the human mind.⁷ (The former, stronger reading of S1 is also consistent with Hume’s usage of ‘natural object,’ but nothing in the text positively mandates it.) And in the case of the Stoic sage and Platonist devotee, Hume seems to be conceding that a few powerful intellects might be able to strain their feelings toward such ultimate cosmological realities, if only for brief episodes. So even if the rest of us are unable to feel moral sentiments toward the ﬁrst cause of all, there remains the possibility that these more enlightened contemplatives are able to do so. And perhaps their ability to feel moral sentiments toward this original being or principle is sufﬁcient to gild or stain it with moral attributes. One possible response to this objection would be to deny that Hume is serious in admitting that these sages can genuinely experience passions directed toward the ﬁrst cause. After all, the main point of the anti-Stoic and anti-Platonist passages is to stress that the normal mechanisms of human psychology restrict our emotional attention to the mundane ‘private system’ (EHU 8.34) and will not direct our passions toward ultimate cosmological realities such as the universe at large or its original cause. This seems to be the real message of these texts, and it is quite possible to read Hume’s apparent concession to the ‘speculative man’ either as so much deadpan irony, or as a ﬁg leaf of diplomatic insincerity, permitting Hume to plead that his critique of religious feeling is not quite so scandalous as it certainly appears.⁸ Given this reading, we might then adopt the stronger interpretation of thesis S1 and take Hume’s real view to be that no human could ever experience passions directed toward the original cause of all. I ⁷ See Section 3.2. ⁸ Hume regularly employs this trope of ironic or insincere deference in his writings on religion, as, for example, when he remarks that ‘A correct Judgment,’ will ‘[avoid] all distant and high enquiries, [and conﬁne] itself to common life,’ and thereby ‘leav[e] the more sublime topics to . . . the arts of priests and politicians’ (EHU 12.25). See also E 598, EHU 11.16, DNR 12.33, and, more generally, my discussion of Hume’s use of faux-pious language in Section 1.4.

104 the argument from sentimentalism 2 have already suggested that there is a fair amount of circumstantial evidence supporting this more uncompromising interpretation.⁹ But since this reading is not inescapably compelled by the text, it might seem like special pleading to insist on it here. In any case, a second sort of response is also available. To see this, consider that a sage whose passions do indeed range out to the deity would have to differ from the rest of us in one of two crucial respects. Either (i) he must have some sort of highly unusual sentimental psychology, such that his passions can respond to the original cause of all even while he knows no more about this ultimate being or principle than the rest of us. (For most humans, the passions cannot ‘operate without the Assistance of the Senses, & Imagination, or at least a more compleat Knowledge of the Object than we have of the Deity’ (NL 13); the passions of this ﬁrst sort of sage would differ in exactly this respect.) Or, alternatively, (ii) if the sage has the same basic sort of sentimental psychology as other humans, then he must have signiﬁcantly more knowledge of the original cause than other humans: knowledge that enables him to experience sentiments directed toward this remote being or principle in a way that the rest of us, in our ignorance, cannot. Take the sage whose sentimental psychology is quite different from that of a normal human. Hume could simply admit the possibility of a few rare individuals like this (at least for the sake of the argument), while insisting that their ability to experience sentiments of moral approval or blame toward the deity is insufﬁcient to confer moral attributes upon it. For Hume, moral standards are not deﬁned by just anyone’s projective sentiments. Instead, like the aesthetic standards of beauty and deformity, they rest on ‘the common sentiments of human nature,’ systematized in the form of an ideal observer theory. Thus in Hume’s aesthetic theory, standards are ﬁxed by ‘general experience concerning what has been universally found to please in all countries and ages,’ corrected and organized with reference to idealized stabilizing conditions providing an observant, practiced, and open-minded audience (E 232, 231; see also EPM 1.9). And similarly, in Hume’s moral theory, the standards of virtue and vice are constituted by the natural moral sentiments encoded in our common human ⁹ See Section 3.2.

the argument from sentimentalism 2

105

nature, again corrected with reference to the stabilizing conditions of a calm and disinterested general point of view (T 3.3.1.15–21, 3.3.3.2; EPM 5.41–2).¹⁰ Correct moral judgments must therefore reﬂect our shared natural sentiments: The notion of morals implies some sentiment common to all mankind, which recommends the same object to general approbation, and makes every man, or most men, agree in the same opinion or decision concerning it. (EPM 9.5) When a man denominates another his enemy, his rival, his antagonist, his adversary, he is understood to speak the language of self-love, 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 choose a point of view, common to him with others: He must move some universal principle of the human frame, and touch a string, to which all mankind have an accord and symphony. (EPM 9.6)

In the moral as in the aesthetic case, there remains room for the correction of popular attitudes, since the moral or aesthetic expert may surpass the obtuse or inexperienced in attentiveness to the object of assessment, in delicacy of discrimination, in freedom from prejudice, and in reasoning ability. However, the judgment of the genuine expert, even when wide of the popular view, still ultimately resolves to an accurate assessment of the object of evaluation’s ability to excite the appropriate sentiments in our common human nature. Expert judges are not above our common human sentiments, but simply better predictors of them. Moral standards, like aesthetic standards, thus ultimately reﬂect the sentimental attitudes hardwired into human nature, and our moral judgments, like our aesthetic judgments, must track these shared natural attitudes if they are to have any authority. The result is that the idiosyncratic sentimental attitudes of the ﬁrst sort of sage cannot draw the deity into the sphere of morally assessable beings. They will have no impact on the ‘universal standard’ of virtue and vice, which is constituted by the ‘natural principles of the human ¹⁰ For useful surveys of these parts of Hume’s system, see Don Garrett, Cognition and Commitment in Hume’s Philosophy (New York: Oxford University Press, 1997), 193–5; Jacqueline Taylor, ‘Virtue and the Evaluation of Character,’ 284–7.

106 the argument from sentimentalism 2 mind,’ not by the atypical responses of an exceptional, superhuman, or unhinged sentimental psychology (EPM ‘A Dialogue’ 56, 57). Hume drives this point home in the closing paragraphs of ‘A Dialogue,’ the short work appended to the second Enquiry, when he discusses the case of ‘artiﬁcial lives’—that is, lives lived out in apparent deﬁance of our natural human sentiments, perhaps in an attempt to conform to the unnatural doctrines of this or that radical philosophical sect or religious ideology. Of particular relevance for our purposes is Hume’s assessment of Pascal—surely the very epitome of a sagelike speculative man of ﬁne genius—and also, by implication, that of St Dominic and St Loyola: cultivated contemplatives who would presumably rank among the stronger claimants to genuine religious emotion (EPM ‘A Dialogue’ 54–7). In Hume’s account, Pascal did indeed adopt at least some sentimental attitudes at odds with those built into human nature—in his case, an obsessive preoccupation with the next life, a self-abasing ﬁxation on his own ‘numberless wants and inﬁrmities,’ and the cultivation of a severe ‘contempt and hatred of himself.’ He shows us, by example, that certain individuals are indeed able to forge an artiﬁcial life by straining some of their passions in unusual directions. However, while Pascal may have ‘met with general admiration in [his age], and [has] been proposed as [a model] of imitation,’ Hume is repelled by his conduct and character. In his behavior and sentimental attitudes—which either warped or suppressed ‘the natural principles of the mind’ and ‘the maxims of common reason’—the Frenchman simply set aside his own humanity. Pascal’s unusual sentimental attitudes carry no authority for Hume but are simply freakish, a particularly striking example of the dehumanizing power of ‘the illusions of religious superstition or philosophical enthusiasm’ (EPM ‘A Dialogue’ 55, 57). At odds with our natural sentiments, they are merely bizarre and inappropriate—and, Hume would add, as likely as not to degrade and corrupt Pascal’s own moral character.¹¹ ¹¹ Hume’s provocative assault on ‘monkish’ character traits earlier in the same Enquiry (EPM 9.3) could easily have been written with Pascal’s ‘abasement,’ ‘hatred of himself,’ and ‘extreme contempt of this life, in comparison with the future’ in mind (EPM ‘A Dialogue’ 55). As Hume notoriously puts it, ‘A gloomy, hair-brained enthusiast, after his death, may have a place in the calendar; but will scarcely ever be admitted, when alive, into intimacy and society, except by those who are as delirious and dismal as himself’ (EPM 9.3).

the argument from sentimentalism 2

107

But what about our second sort of sage, whose ability to experience sentiments directed toward the divine is not the result an idiosyncratic system of passions, but rather the result of greater knowledge or insight into the divine nature? I discuss this case under the rubric of the third (and ﬁnal) objection below.

4.3.3. Third Objection: Is Our Inability to Experience Passions Toward the Deity Merely the Result of our Own Ignorance? The ﬁnal objection runs as follows. According to Hume, we lack sufﬁciently rich imagistic representations of the original cause to engage our passions, and as a result the ﬁrst cause is not a natural object of the moral sentiments. (Again, this is because the passions cannot ‘operate without the Assistance of the Senses, & Imagination, or at least a more compleat Knowledge of the Object than we have of the Deity’ (NL 13).) But surely we need to ask whether this lack of imagery is simply a matter of our current state of ignorance, or whether it rests on more fundamental considerations. Could we at least in principle acquire richer imagistic representations of the ﬁrst cause, ﬁlling our current meager and highly general conception of an unknown ultimate cause of all? Or is there some deeper reason why human sense and imagination (including the understanding—i.e. ‘the general and more establish’d properties of the imagination’ (T 1.4.7.7)) can never adequately capture the original cause and thus direct our sentimental responses toward it? To ﬁx ideas, treat our basic human faculties as a given. Suppose that our natural faculties of sense and imagination remain the same (they might acquire new representations, but without uprooting their basic structure and constitution), and that our sentimental psychology also remains unchanging (so that we continue to feel approval and disapproval toward such-and-such particular character traits, and that inanimate beings, for instance, remain beyond the sphere of our moralized responses). While these natural human faculties incorporate all sorts of contingent features, they are nevertheless bedrock for Hume in providing an account of the source and nature of moral distinctions. (‘To suppose measures of approbation and blame, different from the human, confounds every thing. Whence do we learn, that there is such a thing as moral distinctions but from our own sentiments?’ (E 595).) Given these

108 the argument from sentimentalism 2 ﬁxed parameters of our human faculties, then, is it possible that our knowledge of the original cause could increase, ﬁlling out our conception of this being or principle with the sort of richer imagistic representations that might naturally engage our passions? Suppose that the answer is ‘yes.’ On this view, the deity could prove to be within the moral sphere after all. The ﬁrst cause might be beyond moral assessment only in the superﬁcial and uninteresting sense that a man trapped alone on a desert island is beyond moral assessment: none of us is in a position to know about his character traits, and so none of us will in fact feel sentiments of approval or blame toward him.¹² But obviously this desert islander does have moral attributes, and Hume can readily account for this by providing a subjunctive analysis: were a normal human to know about his behavior and character then they would (under the stable conditions of a calm and disinterested point of view) feel such-and-such moral sentiments toward him (T 3.3.1.15, EPM 9.6, 9.8). Much the same point could be made in the terms of Hume’s own analogy of the remote ancestor. Despite the fact that the remote ancestor is a ‘great Benefactor,’ so long as this is all we know about him, we remain unable ‘to bear him any Affection’ (NL 13). But of course we could learn more about this ancestor, enabling us to feel gratitude toward him. And this fact indicates that the ancestor is indeed worthy of gratitude, notwithstanding our current inability to experience such sentiments. Of course, Hume doubts that humans will ever learn more about the original cause. But the point remains that, insofar as our inability to feel moral sentiments toward God may simply stem from our state of ignorance, we cannot appeal to that inability to support moral atheism, but only moral agnosticism. Suppose then that the answer is ‘no.’ The problem with this response is that it can seem arbitrary and unmotivated. If Hume’s claim is that there is some deeper reason why we are in principle unable to have the sort of rich imagistic representations of the divinity that would make that being a natural object of the passions, then what is that reason? Here Hume would be claiming, not merely that we in ¹² In strictness, we must also stipulate that this desert islander is for some reason unable to survey his own behavior and experience sentiments of praise or blame toward himself. This makes the case rather artiﬁcial, but let this pass.

the argument from sentimentalism 2

109

fact lack sufﬁciently rich imagery of the deity, but that such imagery is humanly impossible. But it is not immediately obvious what could justify this strong claim. Such an assertion may seem simply dogmatic, and perhaps even to implicate the sort of overreaching speculation about the deity’s distinctive intrinsic character that Hume supposedly disallows. This dilemma presents a serious challenge to the argument from sentimentalism. It threatens to expose an equivocation in the sense of ‘natural object’ between S1 and S3. In S3, Hume needs to say that, if normal humans (bracketing cases of madness, disease, ‘artiﬁcial lives,’ and the like) are constitutionally incapable of experiencing moral sentiments toward a certain being, then that being cannot have moral attributes—as with the case of inanimate objects, for instance. But then the overall argument will only go through if we similarly read S1 as asserting that humans are constitutionally incapable of experiencing passions directed toward the ﬁrst cause—and not simply that humans are currently incapable of experiencing such passions given our current (but in-principle remediable) ignorance of that being or principle. But is Hume entitled to this stronger claim? To make good the argument from sentimentalism, Hume thus needs to provide reasons for thinking that humans could never acquire the sort of rich mental imagery required to render the ﬁrst cause of all a natural object of the passions. He needs to show that the poverty of our imagistic representations tracing the nature of this ultimate being or principle is not simply a matter of our current ignorance, but rather stems from permanent features of the human condition. And in fact this does seem to be the way that Hume is thinking of the issue. In asserting that the deity is not the ‘natural’ or the ‘proper’ object of any human passion, he is employing the same language he uses elsewhere to address the permanent and constitutional limits of our various particular passions. For instance, in asserting that the self is the only ‘natural’ object of the speciﬁc passions of pride and humility, and that inanimate beings are not the ‘proper objects’ of any of our moral sentiments, Hume surely means to be understood as making claims about permanent and non-negotiable facts of human sentimental psychology, not simply the limits of our passions relative to some current state of ignorance (T 2.1.3.2, EPM 5.1 note). So in using this same terminology in asserting that that God ‘is not

110 the argument from sentimentalism 2 the natural object of any Passion or Affection’ and that a man ‘may have his Heart perfectly well disposd towards every proper and natural Object of Affection . . . & yet . . . feel no Affection towards [the deity]’ (NL 13), Hume presumably means to locate this being permanently beyond the reach of our sentimental attitudes. This would also explain why Hume is so conﬁdent in assuming that the unusual sentimental responses of a Pascal or a Loyola must stem from an abnormal sentimental psychology, rather than from a perfectly normal sentimental psychology that happens to be accompanied by a rare visionary insight into the nature of the divine. Hume’s implicit premise is that no such insight is possible even for a genius like Pascal, and so it can only be abnormalities in the Frenchman’s sentimental psychology that explain his idiosyncratic attitudes and depressing contemptus mundi. How might Hume try to argue that the deity is beyond the permanent horizon of our sentiments, and not merely beyond the reach of our sentiments given our current (and in principle remediable) position of ignorance? We are already quite far beyond Hume’s own explicit arguments, and so all we can do here is to ask about his best options. Still, we might consider the following three sorts of argument, each of which seems broadly Humean in spirit. One possible approach would be to stress that our responses to the world, including both our formation of cognitive representations and our subsequent sentimental responses, are always from the perspective of one situated within the unfolding causal order of the universe. With this in mind, Hume might hold that our faculties can only successfully refer to objects related to us via the causal sequences that are internal to the developing universe, and thus cannot reach out to an object exterior to the whole creation.¹³ He does after all stress that all our knowledge of matters of fact—including all our knowledge of particular beings—is founded either on the immediate reports of sense and memory, or on causal inference (based on extrapolations of experienced regularities) out from this initial starting ¹³ For versions of this sort of argument in the more recent literature, see Kai Nielsen, An Introduction to the Philosophy of Religion (London: Macmillan, 1982), 169–70, and Anthony Flew, God and Philosophy (London: Hutchinson and Co., 1966), 30–2. For a response, see Paul Helm, Eternal God (Oxford: Clarendon Press, 1988), 195–217.

the argument from sentimentalism 2

111

place (EHU 4.3–4). Our knowledge of particular beings thus seems limited to objects placed within the causal sequences that are internal to the ordered universe. (Here Hume presages the famous Kantian thesis that our cognitive representations are restricted to the world of ‘conditioned’ beings, and cannot range out to an ‘unconditioned’ entity such as an uncaused cause.) So if the original cause is outside the ordered totality of causes that deﬁnes the universe, Hume might then hold that there is no possibility that our situated faculties could ever represent this sort of entity—or at least, no possibility that our humans minds could ever form any speciﬁc representation of it beyond the thinnest and most general conception of an unknown ultimate originating being or principle. Another possible tactic would be to argue that the ﬁrst cause or designer of the ordered universe, being external to the universe itself, is not picturable or imaginable in the manner necessary for our passions to engage. As we have seen, Hume holds that our passions can only be directed toward sensible or imaginable beings. But it is difﬁcult to see how we could think of the original cause as something sensible or imaginable while still doing justice to its transcendental status as something beyond the universe itself. Sensible beings seem eo ipso to be within the universe, and imaginable beings are surely pictured as such. (When Milton renders God in notoriously concrete imagistic detail, ‘ben[ding] down his eye’ from his ‘High Thron’d’ seat to the left of his shining son, and when he places the divine triumvirate in a richly described ‘continent of spacious Heav’n, adorn’d / With Plant, Fruit, Flow’r Ambrosial, Gems and Gold,’ he seems to be locating the deity within the overall ordered universe rather that outside it.¹⁴) Here we might contrast the case of the desert islander with Hume’s original cause, for while the former is not currently sensed or imagined, he is of course sensible or imaginable. But Hume repeatedly describes the ﬁrst cause, not merely as unseen or unknown, ¹⁴ John Milton, Paradise Lost (1667), ed. Christopher Ricks (New York: New American Library, 1968) book 3, line 58; book 6, lines 474–5. Milton wrestles with the problem of directly picturing God in paradoxical passages like the following angelic address to the divinity: ‘Fountain of Light, thyself invisible / Amidst the glorious brightness where thou sitt’st / Thron’d inaccessible, but when thou shad’st / The full blaze of thy beams, and through a cloud / Drawn round about thee like a radiant Shrine, / Dark with excessive bright thy skirts appear, / Yet dazzle Heav’n, that brightest Seraphim / Approach not, but with both wings veil their eyes’ (Paradise Lost, book 3, lines 377–82).

112 the argument from sentimentalism 2 but in stronger terms as ‘invisible’ and ‘incomprehensible’ (NL 13, E 167, NHR 5.2)—suggesting, perhaps, that we are permanently incapable of picturing this sort of empyreal being. So again, perhaps Hume thought it impossible for us ever to have the sort of vivid imagistic representations of the original cause that we would need to engage our passions. A third and ﬁnal possible tactic is as follows. In order to show that we are constitutionally incapable of acquiring adequate representations of the ﬁrst cause or designer, perhaps Hume might appeal, not so much to this being’s external situation, as to its unique status. Given that we can have no immediate experience of the original cause, the only way that we could ever learn anything more about it is by inference from the character of the world that it produced. But as is well-known, Hume holds that we can only make such cause–effect inferences ‘[w]hen two species of objects have always been observed to be conjoined together’ (DNR 2.24). We need to have had experience of regular patterns of cause–effect pairs in order to know anything about the typical causes of any given type of effect. Thus ‘were an effect presented, which was entirely singular, and could not be comprehended under any known species, I do not see, that we could form any conjecture or inference at all concerning its cause’ (EHU 11.30, see also DNR 2.24). Hume is usually read as implying that the great dissimilarity between the universe and all other objects of experience makes it impossible for us to make any responsible judgment about the likely character of the universe’s cause whatsoever. We need not go this far, but it is clear at least that Hume holds that the great dissimilarity between the universe and all other objects of experience ensures that any argument from the character of the universe to the attributes of its cause can only provide (at best) a very weak degree of evidential support, permitting only highly tentative, attenuated, and weakly probable conclusions (at best).¹⁵ Given that the universe is so radically unlike all other empirical phenomena, our speculations about the intrinsic character of its cause must then remain largely guesswork. So we cannot hope to learn enough to ﬁll out our anemic, minimalistic picture of an essentially unknown ﬁrst cause of all with the sort of informative detail that we would need to focus and engage our passions—and ¹⁵ See my discussion of this ‘unique cause objection’ in Section 5.6.

the argument from sentimentalism 2

113

this is a constitutional incapacity, a permanent fact about the human epistemic predicament, not merely a current failing.

4.4. Conclusion Hume is committed to the argument from sentimentalism. His sentimentalist theory of moral properties together with his account of the underlying mechanisms controlling our passions jointly place the deity beyond the sphere of virtue and vice. Hume is clear that we need rich mental imagery to direct our passions, and that such imagery is not available in the case of the original cause. He is also clear that if a being is beyond the natural reach of our passions, then it is beyond the natural reach of our moral sentiments and hence cannot have moral attributes. These theses combine to remove the original cause from the moral domain. Hume was most likely aware that his views led to moral atheism, and he seems to hint at the point when he goes out of his way to emphasize that the deity is not a natural object of our calm affections, which paradigmatically include moral approval and disapproval. While he never explicitly endorses the argument from sentimentalism, the letter to Mure comes quite close. Most importantly, the premises driving the argument are all core Humean doctrines that appear both in early works like the Treatise and Essays, and in the later period of the Enquiries, Natural History, and Dissertation on the Passions. It is of course important to note that, despite his general theory that mental imagery is necessary to direct our passions (in the Treatise and Dissertation), and his hints at the particular obstacle this presents to the direction of sentiments toward ultimate cosmological entities (in the anti-Platonist passage in ‘The Sceptic’ and the anti-Stoic passage in the ﬁrst Enquiry), Hume’s published works never explicitly connect this issue to the question of God’s moral status. In particular, Hume never presses the point in his published works on religion. Perhaps this may reﬂect some hesitation on Hume’s part about the merits of the argument. On the other hand, it may simply reﬂect the prudential obstacles to publishing such a scandalous piece of reasoning, particularly one that is not so readily disguised by the sort of faux-ﬁdeist indirection Hume uses to such effect in his publications

114 the argument from sentimentalism 2 on religion. Or perhaps, in his later works on religion, Hume may have simply wanted to avoid basing his arguments on his controversial sentimentalist moral theory and system of the passions, thoroughly naturalistic doctrines that his wider audience would have been unlikely to accept.

5 The Argument from Motivation 5.1. Introduction In the last two chapters I emphasized Hume’s view that the deity lies beyond the natural reach of our human passions and hence beyond the gilded and stained world of virtue and vice. I now turn from questions about our attitudes toward the deity to questions about the intrinsic character of this original cause or organizing principle. Regardless of our own inability to feel any moral approval or disapproval toward it, does the ﬁrst cause of all itself show any respect for moral standards? Is the deity benevolent, merciful, generous, and the rest? Or is it indifferent to all such concerns? In Hume’s view it is practically certain that the deity (if indeed there is such an original being or principle) is not moved by moral concerns. He argues that it is overwhelmingly unlikely that the deity has a sentimental psychology anything like the human, and that without this sort of approximately anthropomorphic sentimental psychology the deity will not be motivated to moral behavior. This initial stage of the argument then presents us with a probabilistic case for weak moral atheism, the view that the deity is not morally praiseworthy. But the same basic line of reasoning goes further, for it will equally show that in all likelihood the deity lacks the sort of sentimental psychology required to ground humanly comprehensible character traits, traits that might serve as appropriate objects for moral evaluation. This second stage of the argument then presents a probabilistic case for strong moral atheism, the view that the deity is not morally assessable.

116 the argument from motivation This argument from motivation aims to take us, in part, from empirical premises about our own human sentimental psychology to a conclusion about the intrinsic nature of the deity. This might seem an unlikely strategy for Hume to adopt, particularly given his usual archness toward empirically-motivated or ‘experimental’ argument concerning the deity’s intrinsic character. Of course, I have already argued that Hume does allow some conclusions about the divine nature, and that he does practice a limited form of natural theology (Sections 2.3–2.5). But it might seem that there is a particular problem in attributing the current argument to Hume, given that it both boldly advances an account of the deity’s intrinsic nature, and that it does so partly on the back of empirical premises. Such an argument might seem to fall foul of Hume’s own skeptical arguments against traditional (‘core’) natural theology in general and his objections to ‘experimental theism’ in particular (see, for instance, EHU 11.30, DNR 2.4, 2.24). So it might seem that Hume cannot really mean to insist that the deity has no concern for moral standards as opposed to maintaining a strict skepticism or agnosticism on the question, and still less that he would do so by way of argument from experience. At least, it might seem that he cannot indulge in such negative dogmatism without ignoring his own skeptical critique of natural theology and cosmological speculation. In the current chapter I document and examine the argument from motivation, Hume’s probabilistic case for the deity’s moral unintelligibility. Hume does argue this way, maintaining that it is overwhelmingly unlikely that the ﬁrst cause of all is a morally comprehensible being (Section 5.2). Moreover, he can argue this way while simultaneously maintaining a robust skepticism regarding all speculation about the deity’s distinctive or species-speciﬁc intrinsic character. Finally, he ought to argue this way: given Hume’s account of moral motivation and his sentimentalist theory of the nature and basis of moral distinctions, he is right to suggest that the original cause is in all likelihood morally unintelligible (Sections 5.3 and 5.4). There will also be a more general point about Hume’s attitude to natural theology. In examining the epistemological principles controlling Hume’s argument, we will get a clearer view of the speciﬁc strictures he imposes on theological and cosmological speculation, sharpening our account of the distinction between the core natural theology that he rejects and the liminal natural theology that he allows (Sections 5.5 and 5.6).

the argument from motivation

117

5.2. The Argument from Motivation: The Texts Hume presents a version of this argument in both his early private correspondence (in the letter to Hutcheson of March 16, 1740) and his later published work (in the 1757 Natural History, and, more obliquely, in the posthumous Dialogues (1779)).¹ The argument takes off from observations about the way in which human sentimental psychology—our system of passions or sentiments, our constitution of conative and affective attitudes and dispositions, including our various desires, aversions, feelings, and emotions—has a particular utility in the speciﬁc circumstances of human life: All the sentiments of the human mind, gratitude, resentment, love, friendship, approbation, blame, pity, emulation, envy, have a plain reference to the state and situation of man, and are calculated for preserving the existence, and promoting the activity of such a being in such circumstances. It seems, therefore, unreasonable to transfer such sentiments to a supreme existence, or to suppose him actuated by them; and the phenomena, besides, of the universe will not support us in such a theory. (DNR 3.13)

This is the character Demea speaking, and of course Demea (the pious ‘mystic’) often represents views quite antithetical to Hume’s own. However, there is good reason to think that Hume’s spokesman Philo is happy to let Demea make these points on his behalf. First, at this point of the Dialogues Demea is in a tactical alliance with Philo: the two of them are insisting in general that (as Philo puts it) the deity ‘is inﬁnitely superior to our limited view and comprehension’ (DNR 2.3), and in particular that (as Demea puts it) it is the ‘grossest and most narrow partiality’ to represent this being as ‘similar to a human mind’ ¹ The argument from motivation is brieﬂy touched on by the following commentators: Norman Kemp Smith, ‘Introduction’ to David Hume, Dialogues concerning Natural Religion, 2nd (edn), ed. Norman Kemp Smith (Indianapolis: Bobbs-Merrill, 1947), 1–75: 31–3; Bernard Williams, ‘Hume on Religion’, in D. F. Pears (ed.), David Hume: A Symposium (New York: St Martin’s Press, 1966), 77–88: 86–7; anthologized in Bernard Williams, The Sense of the Past: Essays in the History of Philosophy, ed. Myles Burnyeat (Princeton: Princeton University Press, 2006), 267–76: 272; Dorothy Coleman, ‘Introduction’ to David Hume, Dialogues concerning Natural Religion, ed. Dorothy Coleman (New York: Cambridge University Press, 2007), xi–xl: xxxi.

118 the argument from motivation (DNR 3.14). Second, no character refutes Demea on the present point. (The best that the constructive natural theologian Cleanthes can do is to point out that in insisting on the ‘mysterious, incomprehensible nature of the Deity,’ Demea is in danger of becoming a religious skeptic without quite realizing it. But this hardly embarrasses the observation about the human sentiments itself (DNR 4.1).) Third, in his ﬁnal remarks toward the end of the Dialogues, Philo underlines Demea’s basic point, bluntly declaring that ‘It is an absurdity to believe that the Deity has human passions’ (DNR 12.31). Philo gives no additional argument for this view, so here Hume is presumably alluding back to Demea’s observations in DNR 3.13. Taking these points together, we can say that in the Dialogues Hume appears to be suggesting, through Demea and Philo, that human sentimental psychology appears adapted to the particular state and condition of human life, and that as a result it is ‘unreasonable . . . to suppose’ that the divinity shares anything like this system of passions with us. In all philosophical precision, the language Demea uses to state this conclusion does not rule out a strictly agnostic attitude toward the question of whether the deity has sentiments like ours. (Maybe it is ‘unreasonable . . . to suppose’ but also unreasonable to deny that the deity has such sentiments: perhaps suspension of belief is what is called for.) But this way of putting the conclusion simply reﬂects a certain dry understatement on Demea’s part, for his argument surely suggests something stronger than strict agnosticism: it suggests that the ﬁrst cause or organizing principle in all likelihood lacks sentiments like the human. There is conﬁrmation that this is Demea’s own understanding of the matter, for just two sentences later he proceeds on the assumption that the argument actually entitles him to this stronger conclusion. Thus, having argued for the additional subsidiary thesis that the deity could not have sensory ideas like the human (since all such ideas ‘are confessedly false and illusive’), he then draws the two points together in the following terms: ‘as the ideas of internal sentiment, added to those of the senses, compose the whole furniture of human understanding, we may conclude that none of the materials of thought are in any respect similar in the human and in the divine intelligence’ (DNR 3.13). So Demea’s position (which is apparently ratiﬁed by Philo) is not that we are totally unable to judge

the argument from motivation

119

whether or not the deity has sentiments like ours, but rather that we can judge that in all likelihood it does not. Suppose then that this is Hume’s view: in all likelihood, the deity lacks sentiments remotely like the human. This ultimate being or principle either has no conative or affective attitudes at all, or if it does have some sort of sentimental psychology, its particular system of passions is quite different to our own. Given Hume’s view of the role of the passions in driving action and his theory that moral distinctions reﬂect human sentimental attitudes, the implications of this picture for the deity’s moral character are sweeping. For Hume, the virtues and vices just are character traits that produce characteristic sentiments of moral approbation and disapprobation in disinterested human observers. And the particular traits that in fact produce our approval are (he maintains) those that are useful for our individual or social wellbeing, while the traits that produce our disapproval are those that are detrimental to human ﬂourishing (T 3.3.1.30, EPM 9.1, 9.12). There are then at least two ways in which moral conduct is immediately grounded in human sentimental psychology. First, a respect for moral standards results from our characteristic sentiments as we survey the behavior of ourselves and others: sentiments of approval that draw us toward moral behavior and sentiments of disapproval that make us recoil from wrongdoing. As Hume puts it in the second Enquiry, ‘Extinguish all the warm feelings and prepossessions in favour of virtue, and all disgust or aversion to vice: Render men totally indifferent towards these distinctions; and morality is no longer a practical study, nor has any tendency to regulate our lives and actions’ (EPM 1.8; see also EPM Appendix 1.3). Without a psychological system of conative and affective attitudes attracting us to moral conduct and repelling us from vice, morality would be motivationally inert. And here it is worth stressing that Demea’s list of the sentiments inextricably bound up with human life explicitly includes ‘resentment, love, . . . approbation, [and] blame’—an array of passions that would seem to comprehend any feelings of moral approval or disapproval, and thus any sentiments that might motivate speciﬁcally moral conduct (DNR 3.13). Second, the virtues and vices are deﬁned in terms of human wellbeing, and human wellbeing itself reﬂects the nature of our sentimental psychology, our particular (moral and nonmoral) human likes and dislikes, our particular human pains

120 the argument from motivation and pleasures. In sum: morality, for Hume, rises out of our sentimental responses to character and behavior, and it concerns the tendencies of that character and behavior for the ﬂourishing of beings with passions like ours. It is an anthropological phenomenon root and branch, and quite unintelligible outside of this human context. So if we grant that the original cause lacks a sentimental psychology like ours and is quite unmoved by the sentiments governing human life, it will be unmoved by any of the concerns of morality. Two other passages conﬁrm this picture, and in these further texts Hume makes the connection to divine amorality quite explicit. The ﬁrst is from Hume’s well-known letter to Francis Hutcheson of March 16, 1740 soliciting comments on the ﬁrst two volumes of the Treatise: I wish from my Heart, I coud avoid concluding that, that since Morality, according to your Opinion, as well as mine, is determin’d merely by Sentiment, it regards only human Nature & human Life. . . . If Morality were determind by Reason, that is the same to all rational Beings: But nothing but Experience can assure us, that the Sentiments are the same. What Experience have we with regard to superior Beings? How can we ascribe to them any Sentiments at all? They have implanted those Sentiments in us for the conduct of Life like our bodily Sensations, which they possess not themselves. (L i. 39)

Although Hume stresses our ignorance of the character of ‘superior Beings,’ the conclusion he draws is not the pure skeptic’s strict moral agnosticism but rather a provisional form of moral atheism. Appealing to the fact that our sentiments are ‘implanted’ in us ‘for the conduct of Life like our bodily Sensations,’ he expresses a clear presumption in favor of the view that such beings lack sentiments like ours, and hence that morality ‘regards only human Nature & human Life.’ The case is similar in our ﬁnal text, a passage from the Natural History of Religion (1757). As in the early letter to Hutcheson and the later Dialogues, Hume draws attention to the utility of our moral attitudes in the particular context of human life. And, once again, although Hume registers our general ignorance of the divine nature, he goes on to endorse, not a pure skepticism or strict agnosticism regarding the deity’s concern or lack of concern with morality, but rather a presumptive moral atheism:

the argument from motivation

121

Nothing can preserve untainted the genuine principles of morals in our judgment of human conduct, but the absolute necessity of these principles to the existence of society. If common conception can indulge princes in a system of ethics, somewhat different from that which should regulate private persons; how much more those superior beings, whose attributes, views, and nature are so totally unknown to us? ‘Sunt superis sua jura.’ The gods have maxims of justice peculiar to themselves. (NHR 13.7)

In Hume’s translation, the aphorism from Ovid refers us particularly to the question of divine justice.² But Hume’s own point seems to be a more general one about all of morality: thus all ‘the genuine principles of morals’ (the phrase is quite unqualiﬁed) have their place and purpose in the speciﬁc circumstances of human life. They are adapted to the particular necessities of our state and situation, and so (Hume seems to be suggesting) the gods will have no interest in any of them.³ Taking this passage together with the letter to Hutcheson ² Hume is quoting Ovid, Metamorphoses, book 9, line 499. ³ Hume’s particular reference to justice (and his analogy with the different codes governing princes) may suggest a more speciﬁc argument regarding this particular virtue. For Hume, justice is the paradigmatic ‘artiﬁcial’ virtue: a virtue that (unlike benevolence and other ‘natural’ virtues) presupposes a conventional system of rules adopted for mutual beneﬁt in certain speciﬁc circumstances. In the case of justice, these circumstances include a competition over scarce resources and the ability of the competitors to make one another ‘feel the effects of their resentment’ (EPM 3.18; see also T 3.2.1–2, EPM 3.1–21, EPM appendix 3.3). (If princes do not share these circumstances with their subjects, they are thus far removed from the obligations of justice—though not from the obligations of benevolence and the other natural virtues.) So given that the deity is neither in competition with us over scarce resources nor prone to our resentment, it follows that this being cannot have obligations of justice (in this narrow, Humean sense) toward us, nor we toward it. The deity is not in the right sort of community with humans, and so cannot have property rights we are bound to respect (or vice versa), or be bound to us by contract or promise (or vice versa). If successful, this argument would rule out the popular conception of a divine creator who has property rights over his creation, and the Mosaic notion of a god who makes binding covenants with humans. More generally, it threatens to place the deity beyond the more deontic or rule-governed part of morality: the ‘artiﬁcial’ sphere of deﬁnite obligations, perfect duties, sharply deﬁned rights, prerogatives, and responsibilities. Hume is certainly committed to this argument, which is implicit in his overall account of justice as an artiﬁcial virtue that operates only in a society of equals competing over scarce resources: we can add it to his overall arsenal of arguments placing the deity outside our moral world. However, in our extract from the Natural History, Hume appears to be making a more general point about morality tout court, not simply about justice. His claim that ‘Nothing can preserve untainted the genuine principles of morals in our judgment of human conduct, but the absolute necessity of these principles to the existence of society’ (NHR 13.7) seems to be a general thesis about all of morality’s utility for human life, not

122 the argument from motivation and with Demea’s and Philo’s comments in the Dialogues, we can say that Hume is committed to at least the ﬁrst stage of the argument from motivation, the inference to weak moral atheism: given that morality emerges from a sentimental psychology that is peculiarly adapted to human ﬂourishing, it is in all likelihood an exclusively human preoccupation, of no concern to the ﬁrst cause of all.⁴ But we can, I think, go further. Hume’s reasoning also implies that this original being or principle is in all likelihood morally unintelligible: it points, that is, to strong moral atheism. Just as the deity most likely lacks the sort of sentimental psychology that would produce any concern with morality, so it most likely lacks the sort of sentimental psychology that would produce any humanly comprehensible behavior. In all likelihood, the ﬁrst cause of all is so alien a being—so radically different to human beings with their own peculiar system of passions—that if it has any conscious or willed behavior at all, it will be quite unintelligible in human terms.⁵ While Hume never explicitly draws this stronger conclusion, I think that we can see him hinting at it both in his letter to Hutcheson and in his suggestion in the Natural History that the gods are in all likelihood incomprehensible beings, creatures just about the role of the particular virtue of justice. Justice might be a particularly clear example, having its purpose and place only in certain clearly speciﬁable circumstances; but much the same could also be said of the other parts of morality, including those natural virtues that do not presuppose any convention or artiﬁce. So while this passage may also point to an argument narrowly concerning justice, it also seems to suggest the more general argument found in the letter to Hutcheson and in the Dialogues: all our moral attitudes are adapted to the particular circumstances of human life, and so (in all likelihood) they will not be shared by the deity. (On the ramiﬁcations of Hume’s account of justice for religious systems of ethics, see David Fate Norton, ‘Hume, Atheism, and the Autonomy of Morals,’ in Anthony Flew, Donald Livingston, George I. Mavrodes, and David Fate Norton (eds), Hume’s Philosophy of Religion (Winston-Salem, NC: Book Service Associates, 1986), 97–144: 128–31. For a useful discussion of natural vs. conventionalist theories of justice and their relevance for moral theology, see also A. J. Simmons, The Lockean Theory of Rights (Princeton: Princeton University Press, 1992), 30, 269–70).) ⁴ Hume would have encountered a similar argument in Cicero’s De Natura Deorum. See Marcus Tullius Cicero, The Nature of the Gods, tr. Horace C. P. McGregor (London: Penguin, 1972), 207–8. On Hume’s familiarity with De Natura Deorum, see Chapter 1, note 28. ⁵ The point here is especially forceful within Hume’s own moral theory, for he holds that we cannot directly assess actions for their moral worth, but must be able to interpret those actions, placing them in a matrix of motives and circumstances, and thus read through patterns of overt conduct to discern a stable and recognizable system of underlying character traits that can serve as the real locus of moral evaluation (T 3.2.1.2, 3.3.1.4, EHU 8.29).

the argument from motivation

123

with ‘maxims of justice peculiar to themselves’ (NHR 13.17). In all likelihood (Hume seems to be suggesting) the ﬁrst cause of all lacks the sort of anthropomorphic sentimental psychology that would render it a morally comprehensible agent, a possible object of either positive or negative moral evaluation. In its basic thrust, the argument from motivation might seem intuitive enough. But Hume’s texts here are very abrupt, and there is some need to draw out the implicit principles that carry him to the conclusion that the deity is not likely to have sentiments like ours. The canons of reason controlling this argument are not purely demonstrative, nor are they inductive in any obvious and straightforward sense. Our challenge is then to understand the argument’s underlying logic, and in particular to see how Hume can arrive at this thesis concerning the deity’s intrinsic character without overstepping the cautious parameters of liminal natural theology. For instance, how is Hume’s empirical point about the utility of human sentimental psychology in the speciﬁc circumstances of human life supposed to inform a conclusion about ultimate transcendental realities? Is this a posteriori observation supposed to provide some sort of evidential support for the conclusion about the intrinsic character of the ﬁrst cause of all? Can this sort of observation be applied in other domains of natural theology? And in particular: how can Hume permit such an argument consistent with his usual skepticism toward theological speculation?

5.3. The Argument from Motivation: Logic and Rationale Here is one way to understand Hume’s reasoning. Perhaps the empirical fact that human sentimental psychology is so well suited to the needs and pressures of human life, along with the attendant empirical fact that this particular sentimental psychology is (in our experience) found in just those beings facing just these needs and pressures, together support an argument by negative induction (or disanalogy) to the effect that a being in entirely different circumstances is most unlikely to have a sentimental psychology like ours. And since the deity—the original cause or organizing principle responsible for the

124 the argument from motivation ordered world—is not likely to be in circumstances relevantly similar to the human, we can conclude that the deity is most unlikely to have this sort of sentimental psychology. Call this the argument from environmental need. Although this sort of argument does plausibly constitute part of Hume’s reasoning in the overall argument from motivation, as it stands it is quite incomplete. The problem is that it requires the premise that the deity is not likely to be in circumstances like the human—that the deity is not likely to be facing the sort of environmental needs and pressures to which human sentiments are so well adapted. Hume no doubt accepts this premise (and no doubt he is right to do so), but we must ask what entitles him to it. After all, he cannot simply rule out a ﬁrst cause that faces environmental pressures like ours on conceptual grounds (for ‘to consider the matter a priori, any thing may produce any thing’ (T 1.4.5.30; see also T 1.3.15.1, 1.4.5.32, EHU 12.29)), and at one point in the Dialogues Hume has Philo explicitly raise the possibility that the deity is a social and collaborative being (or group of beings) very much like the human, and even the possibility that the deity is mortal, procreates, is male or female, and has a human ﬁgure (DNR 5.8–11). Philo does not of course think that any of these possibilities are particularly likely, but they are possibilities all the same. The point for our purposes is that, in order for the argument from environmental need to gain real traction, Hume owes us a reason for thinking that the deity is not likely to be in circumstances like the human.⁶ However, while we shall see that Hume can provide us with such a reason, this same basic reason could equally be given directly for the claim that the deity is not likely to have a sentimental psychology like ours—which is of course the ultimate conclusion we were after all along. So this argument from environmental need is at least somewhat redundant. Although it is not completely otiose (as I explain below), it does presuppose the soundness of an underlying argument that would ⁶ The traditional theist would of course endorse the premise that the deity is not likely to be in environmental circumstances like the human, and so Hume could always take up this concession and advance the argument from environmental need in ad hominem guise. But Hume also accepts this premise himself, as seems clear enough from the letter to Hutcheson and our passage from the Natural History. So the availability of an ad hominem strategy here does not obviate the need to answer the deeper interpretive question: why does Hume himself hold that the deity is unlikely to be in an environment like ours?

the argument from motivation

125

carry us to much the same conclusion more directly, short-circuiting the longer argument from environmental need. The best way to introduce this underlying argument from speciﬁcity (as I shall call it) is to consider, ﬁrst, the absurdity of maintaining an absolute or strict agnosticism in the face of every question regarding the deity’s intrinsic nature. Let us deﬁne strict agnosticism regarding some particular property P as the view that we ought to completely suspend judgment on the question of whether the deity has P. We can further distinguish between two varieties of strict agnosticism, depending on just how we understand this complete suspense of judgment. First, square agnosticism regarding some property P is the position that we ought to hold our degree of conﬁdence in the proposition that the deity has P at 0.5, and likewise hold our conﬁdence in the proposition that the deity lacks P at 0.5.⁷ According to the square agnostic, there is (epistemically speaking) a ﬁfty–ﬁfty chance that the deity has the particular property in question. Second, quietist agnosticism regarding property P is the view that we ought not to take any position on the question of whether the deity has P, not even the probabilistically neutral position of the square agnostic. The quietist agnostic thus resists being drawn into any judgment about these matters, probabilistic or otherwise. Finally, we can deﬁne global strict agnosticism as the view that we ought to be strict agnostics regarding every property that conceivably might or might not characterize the deity; and so likewise for global square agnosticism and global quietist agnosticism.⁸ Given Hume’s skeptical attitude toward theological speculation, global strict agnosticism of one species or the other might at ﬁrst glance seem an attractive position for him to take. But as we have seen, Hume does not in fact adopt global strict agnosticism. Nor ⁷ Adapting it somewhat for current purposes, I borrow the term ‘square agnosticism’ from Stephen Wykstra, ‘Rowe’s Noseeum Arguments from Evil,’ in Daniel Howard-Synder (ed.), The Evidential Argument from Evil (Bloomington: Indiana University Press, 1996), 126–50: 146, by way of David O’Connor, ‘Skepticism and Philo’s Atheistic Preference,’ Hume Studies 29 (2003), 267–82: 270. ⁸ ‘Conceivably,’ because I am setting to one side those properties that, as a matter of aprioristic conceptual necessity, either apply to all beings (such as the property of beingextended-or-not-extended), or apply to none (such as the property of being-extended-andnot-extended). Here I am addressing only those properties that, given the a priori knowable relations among our ideas, conceivably might or might not characterize the ﬁrst cause of all.

126 the argument from motivation should he: global strict agnosticism is quite untenable. First of all, we can note that global square agnosticism, at least, is logically selfdefeating. For instance, if one is a square agnostic regarding the property of having some color or other, one can hardly be a square agnostic regarding the property of being red and a square agnostic regarding the property of being blue—or, for that matter, regarding the property of being some speciﬁc shade of red: crimson, scarlet, or vermilion, and so on. One cannot rationally assign a probability of 0.5 to all of these hypotheses. This is an obvious point, and it may suggest that global square agnosticism is something of a straw man. But it does draw our attention to a point that we can carry forward: a relatively determinate or speciﬁc property is less likely (epistemically speaking) to characterize the deity than a relatively determinable or general property that encompasses it: the epistemic probability that the deity is red is less than the epistemic probability that it has some color or other. Further, we can say this quite independently of any substantive knowledge of the deity’s intrinsic nature, and without making any particular numerical assignment of probability regarding either the more determinate property or the more determinable one. Second, while we do not need to insist on the problem here, it is worth noting that global quietist agnosticism may also face challenges of its own. If we can construct a case where speculative questions about the deity have implications for the satisfaction of our own desires and preferences, then we may ﬁnd ourselves in a forced choice situation where quietism is simply not sustainable, as when Pascal famously asks us to consider the possibility that the deity has the property of eternally rewarding or punishing humans according to whether or not they make a commitment of faith. Assuming that we are not entirely indifferent to such rewards and punishments, then if we are acting rationally our conduct will reveal just how likely we judge the proposition that God has this particular property, for it will reveal the sorts of cost that we are or are not willing to incur in order to ensure that our behavior would appease such a being. (‘[Y]ou must wager. There is no choice, you are already committed.’⁹) Insofar as we are talking about beliefs that have consequences for the satisfaction ⁹ Blaise Pascal, Pens´ees (1670) tr. A. J. Krailsheimer (Harmondsworth: Penguin, 1995), 122–3.

the argument from motivation

127

of our desires and preferences, then, the quietist refusal to take any stance—not even the judgment of ‘square’ ﬁfty–ﬁfty neutrality—may not be rationally possible. To be sure, such practical concerns will not always engage, and most speculative questions about possible divine properties will be theoretical matters with no immediate relevance to the satisfaction or frustration of our own preferences. But the current point is simply that global quietist agnosticism may not be a viable position for those who have at least some desires or preferences. Third and most important for our purposes, even if the divine nature is entirely unknown, both square agnosticism and quietist agnosticism are absurd positions to maintain with respect to certain highly speciﬁc properties. Consider for example the complex conjunctive property of supporting Leeds over any other football team, supporting Huddersﬁeld over any other rugby team, and, during odd years only, supporting North Carolina over any other basketball team. Strict agnosticism regarding this cooked-up property is, I take it, manifestly absurd. The property is clearly extremely narrow and parochial, and rather obviously gerrymandered for effect. To treat it as having a ﬁfty–ﬁfty chance of characterizing the original cause of all (as square agnosticism demands) is quite ridiculous, and to refuse to be drawn on whether or not the ﬁrst cause is likely to have this property (as quietist agnosticism demands) is little better: it surely takes modesty to the point of evasion. It is in fact overwhelmingly unlikely that the original cause has this particular property, and we can say this without claiming to know anything substantive about the positive nature of the deity. So global strict agnosticism is not a tenable position, for at least some properties are (epistemically speaking) highly unlikely to characterize the ﬁrst cause of all. Which properties? First, our example draws attention to the fact that if a property is extremely speciﬁc we can say that (in the absence of any further information) it is unlikely to characterize the ﬁrst cause of all—or, for that matter, to characterize any unknown being that is singled out in advance. Generality and speciﬁcity are difﬁcult to deﬁne in absolute terms, but I take it that we have at least an intuitive sense, sufﬁcient for current purposes, that certain properties are highly general, such as the properties of being subject to change, being extended, or being a thinking thing, whereas certain other properties

128 the argument from motivation are highly speciﬁc, such as the property of simultaneously supporting Leeds in football, Huddersﬁeld in rugby, and North Carolina in basketball, or being the exact shape and size of Madagascar. So long as the attributes of the ﬁrst cause remain unknown, strict agnosticism might be a reasonable stance to take with respect to certain very general properties. But it is much less reasonable in the case of highly speciﬁc ones. This point, I take it, forms the implicit background to Hume’s current argument for divine amorality. Given his theory of moral motivation, to be moved by moral concerns a being would have to be, not merely a thinking thing, and not merely a thinking thing with some sort of sentimental psychology, but a thinking thing with a highly particular array of conative and affective attitudes corresponding in all sorts of ﬁne-grained ways to our own human sentiments. So Hume is stressing this corollary of his theory of moral motivation: to have any concern with morality is to have an exceptionally speciﬁc property, and hence a property that is unlikely to characterize the ﬁrst cause of all, or indeed to characterize any given unknown being that we might pick out in advance.¹⁰ ¹⁰ Can Hume’s ofﬁcial theory of probabilistic reasoning accommodate the argument from speciﬁcity? The principle of probabilistic reasoning driving the argument—that any highly speciﬁc property that is picked out in advance is (epistemically speaking) unlikely to characterize any given unknown being—might seem plausible enough. But it might also seem aprioristic and un-Humean in character. According to Hume’s ofﬁcial theory, there are no purely aprioristic principles governing probabilistic reasoning, since all such reasoning presupposes some empirical knowledge of the relevant causal laws at work, even if (as when throwing dice) many of the particular determinative causes in action are still unknown and so must be regarded as matters of ‘chance.’ Thus, ‘it is remarkable, that . . . it is impossible for us to conceive [a] combination of chances, which is requisite to render one hazard superior to another, without supposing a mixture of causes among the chances, and a conjunction of necessity in some particulars, with a total indifference in others. Where nothing limits the chances, every notion that the most extravagant fancy can form is upon a footing of equality; nor can there be any circumstance to give one the advantage above another. Thus, unless we allow that there are some causes to make the dice fall, and preserve their form in their fall, and lie upon some one of their sides, we can form no calculation concerning the laws of hazard’ (T 1.3.11.6). So if the argument from speciﬁcity does indeed require the application of a purely aprioristic principle, then Hume’s ofﬁcial theory cannot allow it. (Of course, Hume may in practice employ such a principle, even if his ofﬁcial theory disallows it: perhaps he is simply more ﬂexible in his actual argumentation than his rather narrow background theory permits.) But do we have to think of the principle driving the argument from speciﬁcity as aprioristic in character? Perhaps we might see it as a highly abstract generalization drawn up from our prior speculations and subsequent discoveries regarding unknown objects, rather like the various ‘[r]ules by which to judge of causes and effects’ that Hume distills from our experience with causal reasoning—highly abstract

the argument from motivation

129

Hume’s underlying argument for the conclusion that the deity is unlikely to have a system of passions like the human thus simply exploits the fact that this sort of highly speciﬁc property is unlikely (epistemically speaking) to characterize any unknown being that is singled out in advance. And notice that this same sort of argument from speciﬁcity must implicitly underlie the claim—which was required as a premise for the argument from environmental need we ﬁrst examined—that the deity is unlikely to be in circumstances like the human. We can make this claim only because being in circumstances like the human is an enormously speciﬁc property, and hence a property that is unlikely to characterize any unknown being that is singled out in advance. That is why I characterized the argument from environmental need as at least somewhat redundant: it presupposes the soundness of a more direct argument from speciﬁcity, which could by itself already establish the conclusion that the deity is highly unlikely to share a sentimental psychology like ours. So Hume could get by without the argument from environmental need and his current case for moral atheism could proceed all the same. But this redundancy should not be overstated, for while the argument from speciﬁcity already shows that the deity is highly unlikely to share a system of passions like the human, the argument from environmental need might help to reduce that likelihood still further. Perhaps the empirical fact that human sentimental psychology is so ﬁnely adapted to human life, and occurs (in our experience) only in beings that require this sort of sentimental psychology to survive and ﬂourish, does (through negative analogical reasoning) provide additional evidence that the ﬁrst cause of all is unlikely to have such a system of passions, and thereby helps to compound this improbability the more. Given the emphasis on such premises about the utility of human sentimental psychology for life in our human environment in each of our three texts, along with the (admittedly speculative) rationalizing considerations sketched above, the most plausible view is that Hume is implicitly committed to and relying on both the argument from environmental need and the sorts of considerations sketched in the argument from speciﬁcity in order to drive his overall argument from motivation. principles that are ‘are form’d on the nature of our understanding, and on our experience of its operations in the judgments we form concerning objects’ (T 1.3.15 title, 1.3.13.11).

130 the argument from motivation

5.4. Rival Early Modern Perspectives and Hume’s Responses To help situate the argument from motivation in its original historical and philosophical context, we might consider the rival early modern perspectives that Hume is setting himself against. First, his analysis is clearly opposed to any system that treats a concern with morality as a fairly general property rather than a highly speciﬁc one. Here the most prominent camp at odds with Hume is the tradition of moral rationalism, which holds that moral constraints are simply a form of rational constraint, and hence that any being that is acting rationally is eo ispo acting with due regard for morality. On this view, human sentiments need not simply reﬂect the particular needs and pressures peculiar to human life, but might also reﬂect a rationally discernible system of eternal and immutable proper objects of the passions, objects with essential and intrinsic positive or negative worth, not merely objects of parochial and contingent human approval or disapproval. And on this view, any rational being qua rational would share at least some of our own motivating attitudes and dispositions, insofar as those attitudes and dispositions are dictated by pure reason alone. In the early modern British context, this traditional Thomistic position was revitalized and inﬂuentially promulgated (in opposition to the radical voluntarism of Calvinist puritans) in the ﬁrst book of the Elizabethan divine Richard Hooker’s Lawes of Ecclesiastical Politie (book 1, 1594); Cambridge Platonists such as Benjamin Whichcote and Ralph Cudworth sponsored the position during the seventeenth century, as did the rationalist theologian Samuel Clarke at the turn of the eighteenth. For all of these philosophers a concern for morality is a characteristic of all rational beings qua rational, and hence a much more general property than Hume would allow. This is what Hume is stressing in the letter to Hutcheson when he writes that ‘If morality were determind by Reason, that is the same to all rational Beings’ (L. i 39), and similarly, in book 3 of the Treatise, published later that year: According to the principles of those who maintain an abstract rational difference betwixt moral good and evil, and a natural ﬁtness and unﬁtness of things, ’tis not only suppos’d, that these relations, being eternal and

the argument from motivation

131

immutable, are the same, when consider’d by every rational creature, but their effects are also suppos’d to be necessarily the same; and ’tis concluded they have no less, or rather a greater, inﬂuence in directing the will of the Deity, than in governing the rational and virtuous of our own species. (T 3.1.1.22)

But Hume’s sentimentalist moral philosophy does away with all this. In founding morality not on abstract rational principles, but ‘entirely on the particular fabric and constitution of the human species’ (EPM 1.3), it makes a concern with morality a highly speciﬁc property, characteristic only of beings with a highly particular sentimental psychology. Thus, contra moral rationalism, it is much less likely to characterize any unknown being we might pick out, even if that being proves to have intelligence and reason.¹¹ Hume’s conclusion also presents a major breach with the nonrationalist theists in the philosophical culture of early modern Britain, and this presents us with a second sort of tradition to contrast with his analysis. This second camp insists that there are independent reasons to think that the ﬁrst cause of all is concerned with morality, even if this property proves to be (as Hume insists, in opposition to moral rationalists) highly speciﬁc.¹² This camp would include those sentimentalists who argue for a morally praiseworthy divinity, such as the deist Anthony Ashley Cooper (Lord Shaftesbury) and the Christian moralist Francis Hutcheson. As fellow sentimentalists, these philosophers agree with Hume that a concern with morality is a highly speciﬁc property, and hence that when (in Hutcheson’s words) ‘the original fabric of our nature was, by the divine art and plan, designed ¹¹ Hume also makes this point in an unpublished fragment on evil, tentatively dated by Stewart to the approximate period of the Treatise: ‘The attempt to prove the moral Attributes from the natural, Benevolence from Intelligence, must appear vain, when we consider, that these Qualities are totally distinct & separate. Reason & Virtue are not the same; nor do they appear to have any immediate Connexion, in the Nature of things. Even in Man, any degree of the one affords no presumption for an equal degree of the other. A sound Understanding & a hard Heart are very compatible.’ (See M. A. Stewart, ‘An early fragment on evil,’ in M. A. Stewart and John P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), 160–70: 165.) ¹² So this camp would include those who subscribe to the traditional view that the deity’s positive moral character can be established (or at least shown to be probable) through rational argument, either by way of natural reason alone, or by way of a supernatural revelation that is authenticated and ratiﬁed by natural reason. It will not include those radical ﬁdeistic thinkers, such as Pascal, Bayle, Jurieu, and Hamann, who reject the existence of rational support for the doctrine of a morally laudable divinity.

132 the argument from motivation for every virtue’ it was designed in an enormously speciﬁc way.¹³ But while a concern with morality is thus a highly speciﬁc property and hence a priori unlikely to characterize any given unknown being, Shaftesbury and Hutcheson each claim that the deity is not in fact an entirely unknown being. Each of these philosophers discerns a benevolent scheme in the order of creation and infers from thence that the deity is a paternalistic intelligence with moral sentiments much like ours.¹⁴ Hutcheson, as a Christian, can also appeal to the authority of scripture on this point. But as is well-known, Hume both doubts the credentials of scripture (EHU 10), and rejects the a posteriori inference from the order of the world to a moral ﬁrst cause (EHU 11.14–17; DNR 10.35–36, 11.2–4; for commentary see Section 6.3 below). In Hume’s view we lack any real evidence that the deity has a sentimental psychology like ours, and in fact know nothing whatsoever about the distinctive, species-speciﬁc intrinsic character of the divinity. And so long as we remain in this state of ignorance, the proper stance is neither moral theism nor even moral agnosticism, but a strong (though of course fallibilistic) presumption in favor of moral atheism. For Hume, the temptation to think that the ﬁrst cause of all acts out of humanly intelligible passions, or even to treat this as a live possibility worth treating seriously, is simply the result of our ‘egregious partiality’ in taking our own minds as the model for the original cause (DNR 7.11). This partiality in turn reﬂects our general tendency to anthropomorphize the beings around us, projecting our own attitudes and concerns onto the nonhuman world.¹⁵ Thus according to the Natural History: There is an universal tendency among mankind to conceive all beings like themselves, and to transfer to every object, those qualities, with which they ¹³ Francis Hutcheson, On the Natural Sociability of Mankind (1730), in Hutcheson, Logic, Metaphysics, and the Natural Sociability of Mankind, ed. James Moore and Michael Silverthorne (Indianapolis: Liberty Fund, 2006), 189–216: 200. ¹⁴ Francis Hutcheson, An Inquiry into the Original of Our Ideas of Beauty and Virtue (1725), ed. Wolfgang Leidhold (Indianapolis: Liberty Fund, 2004), 181 (treatise 2, section 7.5); Anthony Ashley Cooper, Lord Shaftesbury, ‘The Moralists; a Philosophical Rhapsody’, in Characteristics of Men, Manners, Opinions, Times, ed. Douglas Den Uyl, 3 vols (Indianapolis: Liberty Fund, 2001), ii. 101–247: 164. ¹⁵ For an illuminating discussion of Hume on the human tendency to anthropomorphize nature, see P. J. E. Kail, Projection and Realism in Hume’s Philosophy (Oxford: Oxford University Press, 2007), 7–12.

the argument from motivation

133

are familiarly acquainted, and of which they are intimately conscious. We ﬁnd human faces in the moon, armies in the clouds; and by a natural propensity, if not corrected by experience and reﬂection, ascribe malice or good-will to every thing, that hurts or pleases us. Hence the frequency and beauty of the prosopopœia in poetry; where trees, mountains and streams are personiﬁed, and the inanimate parts of nature acquire sentiment and passion. And though these poetical ﬁgures and expressions gain not on the belief, they may serve, at least, to prove a certain tendency in the imagination, without which they could neither be beautiful nor natural. . . . Nay, philosophers cannot entirely exempt themselves from this natural frailty; but have oft ascribed to inanimate matter the horror of a vacuum, sympathies, antipathies, and other affections of human nature. The absurdity is not less, while we cast our eyes upwards; and transferring, as is too usual, human passions and inﬁrmities to the deity, represent him as jealous and revengeful, capricious and partial, and, in short, a wicked and foolish man, in every respect but his superior power and authority. (NHR 3.2; see also T 1.4.3.11)

Hume’s emphasis here on the projection of vicious traits is no doubt colored by his revulsion at the ‘barbarous conceptions of the Divinity . . . multiplied upon us’ by popular religion.¹⁶ But he also invokes the same basic diagnosis to account for those occasions when we project our better natures onto the ﬁrst cause. Thus in the ﬁrst Enquiry, Hume levels the complaint of anthropomorphism directly against the natural theologian who attempts to reason his way to a wise and benevolent divinity: The great source of our mistake in this subject, and of the unbounded licence of conjecture, which we indulge, is, that we tacitly consider ourselves, as in the place of the Supreme Being, and conclude, that he will, on every occasion, observe the same conduct, which we ourselves, in his situation, would have embraced as reasonable and eligible. But, besides that the ordinary course of nature may convince us, that almost everything is regulated by principles and maxims very different from ours; besides this, I say, it must evidently appear contrary to all rules of analogy to reason, from the intentions and projects of men, to those of a Being so different, and so much superior. In human nature, there is a certain experienced coherence of designs and inclinations; ¹⁶ Popular religion in general and ‘the representations given us by some later [i.e. monotheistic] religions’ in particular (NHR 13.7; see also NHR 13.6, 14.9). Hume particularly singles out Calvinism (at least as an example), for in a note appended to this paragraph he pointedly quotes at length from Chevalier Ramsey’s jeremiad of moral disgust at this speciﬁc theological system (DNR 13.7 note).

134 the argument from motivation so that when, from any fact, we have discovered one intention of any man, it may often be reasonable, from experience, to infer another, and draw a long chain of conclusions concerning his past or future conduct. But this method of reasoning can never have place with regard to a Being, so remote and incomprehensible. . . (EHU 11.27)

So Hume allows that we have a natural tendency to project our human attitudes, good and ill, onto the cosmological horizon and higher spheres of being—a propensity to create ‘spectres of false divinity’ that are simply the shadows and after-images of our own human needs, fears, and desires (NHR 14.8; see also Section 3.4 above).¹⁷ But this ‘natural frailty’ (NHR 3.2) lacks any real justiﬁcation, and in Hume’s view there is no genuine evidence suggesting that the ﬁrst cause of all has anything like our human sentimental psychology. And in the absence of such evidence, it is practically certain that the deity lacks this sort of anthropomorphic system of passions, even while we know nothing at all regarding its own distinctive intrinsic nature.

5.5. Liminal Natural Theology and the Argument from Motivation We should now be able to see how Hume can prosecute the argument from motivation while simultaneously maintaining a robust skepticism regarding all core natural theology. The ﬁrst point to notice is that his conclusion is a negative one concerning a highly speciﬁc property, and as such conveys very little positive doctrine about the divine attributes. Anyone wondering how to begin to conceive of the original cause of all would hardly be helped to be told that this being or principle does not (in all likelihood) have the highly speciﬁc property of having a sentimental psychology like the human, any more than they would ¹⁷ This anthropomorphizing tendency had been lamented previously by various other early modern philosophers, including both those of a religious and an irreligious stripe. See, for instance, Antoine Arnauld and Pierre Nicole, Logic, or the Art of Thinking (1662, rev. 1683), tr. and ed. Jill Vance Buroker (Cambridge: Cambridge University Press, 1996), 29–30; Nicholas Malebranche, The Search After Truth (1674–5), tr. Thomas M. Lennon and Paul J. Olscamp (Cambridge: Cambridge University Press, 1997), 251 (book 3, part 2, ch. 9), and Benedict de Spinoza, Ethics (1677), tr. and ed. G. H. R. Parkinson (Oxford: Oxford University Press, 2000), 106–12 (appendix to part 1).

the argument from motivation

135

be helped to be told that it does not (in all likelihood) simultaneously support Leeds, Huddersﬁeld, and North Carolina. The point is that while Hume’s argument does aim to tell us something about the likely character of the deity, it does not begin to offer us a substantive positive conception of the deity’s intrinsic nature. Of course, the argument from motivation does aim to establish a highly controversial thesis—namely, that the deity is (most likely) indifferent to moral concerns. And since this thesis is so controversial, it might seem that it must convey signiﬁcant information. But Hume’s whole point is that, once we accept that morality is bound up with a highly speciﬁc sort of sentimental psychology, it follows that the thesis of (presumptive) divine indifference is not such a substantive one after all: indeed, it is a thesis that one can responsibly endorse while knowing nothing whatsoever about the deity’s distinctive intrinsic nature. This brings us to the second and most important point. Hume’s reasoning in the argument from motivation would apply not just to the unknown original cause, but to any unknown being that we might single out. Given any unknown being that is picked out in advance, that being is unlikely (epistemically speaking) to have sentiments like ours, or to experience the sort of environmental pressures that are associated with this sort of sentimental psychology. So Hume is not then claiming any special insight into the particular nature of the deity, nor is he asserting anything of it that he would not equally assert of any other unknown being.¹⁸ He is not then engaging in core natural theology: the ambitious form of cosmological theorizing that aims at a conception of the distinctive intrinsic nature of the original cause, a conception that provides some species-speciﬁc account of the divine attributes permitting us to distinguish it from other unknown beings. He is only practicing what I have called liminal natural theology: the form of reasoning about the nature of the deity that makes no species-speciﬁc claims about the intrinsic character of this original being or principle. ¹⁸ Since Hume’s argument proceeds without any empirical knowledge of the ﬁrst cause and could equally be applied to any other unknown being, it could in this sense be described as a priori (as it is by Bernard Williams in ‘Hume on Religion,’ 86–7). However, Hume does appeal to at least one empirical premise, namely, the fact that human sentiments are (in our experience) correlated with and adapted to a highly particular form of life in a highly particular environment, and so his reasoning is a posteriori in the stricter and more familiar sense.

136 the argument from motivation This point can be generalized across Hume’s approach to the philosophy of religion, and here we can ﬁll out our account of the liminal sort of natural theology he does in fact allow. As we saw in Chapter 2, Hume does not eschew all reasoning about the ﬁrst cause of all, notwithstanding his familiar denunciations of theological speculation. For instance, he permits himself aprioristic forms of reasoning that identify conceptual necessities governing the divine attributes—indicating, for instance, that the deity cannot be both immutable and a thinking thing, since this is a conceptual impossibility and a contradiction in terms (DNR 4.3). Moreover, he also countenances certain provisional and probabilistic forms of reasoning that identify combinations of properties that the deity is most unlikely to have—arguing, for instance, that the deity is unlikely to be both a mind-like being and yet, unlike all experienced minds, capable of acting without a body (DNR 8.11); and he argues (as we have now seen in detail) that the deity is unlikely to have a sentimental psychology anything like the human. We can now understand how Hume can permit such reasoning without violating a prohibition on core natural theology, at least, for in none of these cases does he appeal to premises invoking the speciﬁc nature of this original being or principle, but rather simply exploits a more general form of argument that would apply to any unknown being we might pick out in advance. In this sense, Hume’s assertion of our total ignorance of the divine nature is not at odds with his advancing certain claims about its intrinsic character, insofar as those claims could equally be made of any other unknown being. The ultimate cause of all remains a black box, but this does not mean that we cannot move beyond the extreme reticence of global strict agnosticism to hazard some probabilistic conclusions about its intrinsic character.

5.6. The Unique Cause Objection My interpretation faces one remaining problem. I have suggested that Hume does practice a modest form of reasoning about the deity’s intrinsic nature, notwithstanding his total rejection of core natural theology. Hume’s own liminal form of natural theology permits reasoning that would apply equally to any other unknown being we

the argument from motivation

137

might single out.¹⁹ It includes (1) arguments about the combinations of properties that the deity could not possibly have. But it also includes (2) arguments about the sort of properties that the deity is not likely to have, either (2a) because of their high degree of speciﬁcity, or (2b) because this would be out of line with an empirically observed pattern governing the occurrence of such properties (as in the argument from environmental need, and in the argument that the deity is not likely to be both intelligent and disembodied). It is this last, empirical form of argument, category (2b), that causes the trouble.²⁰ The problem is that both in the ﬁrst Enquiry and again in the Dialogues, Hume outlines an objection that might seem to tell against all empirically motivated arguments for conclusions about the nature of the ﬁrst cause, regardless of whether these arguments are part of the ambitious program of core natural theology or merely the relatively modest agenda of liminal natural theology: It is only when two species of objects are found to be constantly conjoined, that we can infer the one from the other; and were an effect presented, which was entirely singular, and could not be comprehended under any known species, I do not see, that we could form any conjecture or inference at all concerning its cause. If experience and observation and analogy be, indeed, the only guides which we can reasonably follow in inferences of this nature; both the effect and cause must bear a similarity and resemblance to other effects and causes, which we know, and which we have found, in many instances, to be conjoined with each other. (EHU 11.30) When two species of objects have always been observed to be conjoined together, I can infer, by custom, the existence of one wherever I see the existence of the other; And this I call an argument from experience. But how this argument can have place, where the objects, as in the present case, are single, individual, without parallel, or speciﬁc resemblance, may be difﬁcult to explain. (DNR 2.24)

Given that the universe is unique, or at least unique in our experience, it cannot be comprehended under any broader species of objects ¹⁹ It also permits species-speciﬁc arguments about the deity’s relational properties, insofar as these arguments take off from its deﬁning relational property of being the original cause or organizing principle of the ordered universe (see Section 2.4.3). ²⁰ If the speciﬁcity arguments of category (2a) also require empirical premises, then the same basic objection will apply to them too. See note 10 above on the question of whether such speciﬁcity arguments are aprioristic or empirical in character.

138 the argument from motivation known by way of experience, thereby hobbling any attempt to infer facts about its cause by way of analogy with other more familiar empirical phenomena. Hume states the objection in quite forceful terms, and he has widely been interpreted as claiming that the uniqueness of the universe poses an absolute, insurmountable barrier to any attempt to reason analogically from observed patterns in experience to a conclusion about the character of its original cause. This would include Cleanthes’s attempts to argue, in the spirit of core natural theology, from familiar empirical cause–effect regularities to some account of the distinctive intrinsic character of the cause of the universe. But it would also include the empirically motivated arguments within Hume’s own liminal natural theology. For instance, when Hume observes that human sentimental psychology is adapted to our particular environment and circumstances, and infers that the deity (like any given unknown being) is unlikely to share anything like this sort of psychology with us, he is reasoning analogically from empirical phenomena to a conclusion about the likely character of the cause of the universe (along with the likely character of any other given unknown being). Likewise, when he argues from the fact that all observed minds have been embodied to the conclusion that the deity (like any given unknown being) is unlikely to be both mind-like and disembodied, he is again using analogical reasoning to project observed regularities (provisionally and fallibilistically, to be sure) onto the ﬁrst cause of all. He is, in short, permitting reasoning about the likely character of the original cause based on analogical extrapolation from empirical phenomena. But such reasoning seems to run directly into his own unique cause objection, at least if that objection is given an absolutist interpretation and is taken to imply that no inference from experience can begin to support a conclusion about the cause of a being as ‘entirely singular’ as the universe.²¹ ²¹ For the absolutist interpretation of the objection, see Stephen Buckle, Hume’s Enlightenment Tract (Oxford: Clarendon Press, 2001), 290–2; Terence Penelhum, Themes in Hume: The Self, the Will, Religion (Oxford: Clarendon Press, 2000), 234; Keith E. Yandell, Hume’s ‘Inexplicable Mystery’: His Views on Religion (Philadelphia: Temple University Press, 1990), 175–8; Anthony Flew, Hume’s Philosophy of Belief: A Study of his First Inquiry (London: Routledge and Kegan Paul, 1961), 230–2; J. C. A. Gaskin, ‘Religion: The Useless Hypothesis,’ in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), 349–69: 363–5. (But for some reservations, see Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, NJ: Humanities Press International, 1988), 27.)

the argument from motivation

139

If Hume’s unique cause objection does indeed place an absolute prohibition on all experientially based reasoning concerning the divine nature, then there is an anomaly either in my interpretation of Hume’s system or in Hume’s system itself. But does Hume intend his unique cause objection this way? In fact there is reason to doubt that Hume took the unique cause objection to present an absolute, in-principle obstacle to all experientially grounded argument concerning the original cause of the universe. Interpreted in such an absolutist fashion, the unique cause objection would face the rather obvious response that the universe can in fact be comprehended under numerous species or categories (such as, for instance, the category of beings that show structure and organization in their parts). Consider how Cleanthes frames the issue: ‘Look round the world: Contemplate the whole and every part of it: You will ﬁnd it to be nothing but one great machine, subdivided into an inﬁnite number of lesser machines’ (DNR 2.5). Even if (as Philo suggests) this analogy with a machine proves rather attenuated, so long as the universe bears some resemblance to a machine (or to some other object of empirical experience)—even if that resemblance is quite weak—then it is not strictly ‘without parallel, or speciﬁc resemblance’ (DNR 2.24): it is not unique in the strong sense required to support an absolute, in-principle objection to all analogical reasoning concerning its cause. And in fact Hume himself does allow that the universe bears some analogy to certain objects of empirical experience, as when he has Philo argue that the universe ‘plainly resembles more an animal or a vegetable than it does a watch or a knitting-loom’ (DNR 7.3), that ‘the works of nature have a much greater analogy to the effects of our art and contrivance, than to those of our benevolence and justice’ (DNR 12.8), and again that the debate between the theist and atheist is ‘merely verbal,’ since it simply concerns the degree to which the cause of the universe resembles human mind and thought, not the question of whether there is any resemblance here whatsoever (DNR 12.7; see also DNR 12.33). It is true, of course, that we have had no experience of other objects that closely resemble the universe: we have not, for instance, witnessed other prior universes (DNR 2.24, 2.28). And it is true that analogical reasoning that relates the universe to other objects of empirical experience would be stronger if it proceeded from such closer resemblances. But so long as there is some resemblance between the universe

140 the argument from motivation and other empirical objects, then there may be room for some (most likely highly attenuated and provisional) analogical reasoning.²² For these reasons, I suggest that we might more charitably read EHU 11.30 and DNR 2.24 as somewhat hyperbolic statements stressing the great degree to which the universe differs from all other objects of experience, and hence the relative weakness of the evidential support provided by any analogical argument concerning its original cause, rather than as imposing an absolute, in-principle prohibition on all such arguments. This understanding of the unique cause objection would help to explain why Hume does not devote much space either in the Enquiry or in the Dialogues to prosecuting the objection as a free-standing criticism, but rather proceeds to spend most of his energy in bringing out the highly attenuated character of the particular analogy driving the argument from design: here he is developing his point about the relative ‘uniqueness’ of the universe—that is, the great degree to which it differs from other empirical phenomena like human artifacts.²³ Hume clearly holds that this relative ‘uniqueness’ is sufﬁcient to condemn core natural theology, for the analogy between the universe and any other speciﬁc type of empirical object is so strained that it precludes any serious probabilistic support for any species-speciﬁc conclusion ﬁxing the distinctive intrinsic character of the universe’s ²² According to Hume, analogical reasoning retains some force so long as there is any resemblance at all. In the Treatise Hume writes of ‘that species of probability, deriv’d from analogy, where we transfer our experience in past instances to objects which are resembling, but are not exactly the same with those concerning which we have had experience. In proportion as the resemblance decays, the probability diminishes; but still has some force as long as there remain any traces of the resemblance’ (T 1.3.13.8; see also 1.3.12.25). ²³ Consider the following passage, where Philo stresses that his real objection to the argument from design is the highly attenuated character of the analogy it draws between the universe and an artifact, and hence the weak evidential force the argument carries, rather than some sort of absolutist objection insisting that arguments from experience carry no force whatsoever in speculation about the divine nature: ‘What I chieﬂy scruple in this subject, said Philo, is not so much that all religious arguments are by Cleanthes reduced to experience, as that they appear not to be even the most certain and irrefragable of that inferior kind. That a stone will fall, that ﬁre will burn, that the earth has solidity, we have observed a thousand and a thousand times; and when any new instance of this nature is presented, we draw without hesitation the accustomed inference. The exact similarity of the cases gives us a perfect assurance of a similar event; and a stronger evidence is never desired nor sought after. But wherever you depart, in the least, from the similarity of the cases, you diminish proportionably the evidence; and may at last bring it to a very weak analogy, which is confessedly liable to error and uncertainty’ (DNR 2.7).

the argument from motivation

141

original cause. First, the various species-speciﬁc analogies that are available to us—such as the analogy between the universe and an artifact, or an animal, or a plant—are attenuated to the point where they can only sustain analogical inferences with a very weak degree of probabilistic force (DNR 7.1, 7.8). But, worse, each of the various analogies that we might draw between the universe and some speciﬁc kind of empirical object must be set against all the other competing analogies, delivering a vast range of mutually conﬂicting hypotheses regarding the distinctive nature of the ﬁrst cause, each of which helps to undercut the plausibility of the others. When we speculate from experience about the likely distinctive character of the ﬁrst cause of all, then, ‘a hundred contradictory views may preserve a kind of imperfect analogy; and invention has here full scope to exert itself. . . . I could, in an instant, propose other systems of cosmogony, which would have some faint appearance of truth; though it is a thousand, a million to one, if either yours or any one of mine be the true system’ (DNR 8.1). None of this amounts to an aprioristic prohibition on core natural theology: if we were acquainted with objects that resembled the universe to a sufﬁcient and sufﬁciently singular degree, and if we had sufﬁcient experience of the causal origins of these objects, then perhaps we could meaningfully speculate about the distinctive intrinsic character of the universe’s cause. However, given the empirical data that is actually available to humans, this sort of a posteriori speculation about the distinctive intrinsic character of the ﬁrst cause is (in Hume’s view) effectively worthless. Empirically-motivated liminal natural theology also takes our observation of empirical patterns in the world as some sort of guide as to what to expect in the deity. Its arguments will thus also be weakened by the great dissimilarity between the universe on the one hand and all other empirical phenomena on the other. However, in this case we are not focusing on some species-speciﬁc analogy between the universe and one particular subgroup of empirical objects rather than another. So here at least there is no danger of a multiplication of conﬂicting analogical arguments undercutting one another. And liminal natural theology is a more modest form of speculation than core natural theology: the liminal natural theologian’s claims about the deity’s likely intrinsic character are so weak and general that they could equally be applied to any other unknown object. So even if the claims of

142 the argument from motivation this sort of empirically-motivated liminal natural theologian can only enjoy a weak degree of probabilistic support, still, they may be able to enjoy more probabilistic support than the bolder speculations of core natural theology. This at least seems to be Hume’s analysis of the situation, for, while condemning ‘experimental’ core natural theology, he does in fact practice some highly tentative, fallibilistic sort of empirically motivated liminal natural theology. He does take some well-conﬁrmed empirical patterns (such as the correlation between having a mind and being embodied, and the correlation between having sentiments like ours and being in environmental circumstances like ours) as a provisional guide as to what to expect in other as-yet unknown beings, the deity not excepted.

5.7. Conclusion Notwithstanding his familiar skeptical critique of theological speculation, Hume rejects a strictly agnostic attitude toward the question of whether the deity has the sort of sentimental psychology that might ground humanly comprehensible behavior. Although we know nothing about the deity’s distinctive intrinsic character, it is nevertheless overwhelmingly likely that the ﬁrst cause or designer lacks this sort of anthropomorphic sentimental psychology, just as it is overwhelmingly likely that any other unknown being we picked out in advance would lack this complex and highly speciﬁc system of sentiments. Hume’s probabilistic reasoning here rests on two main claims. First, Hume is appealing to the fact that the possession of a sentimental psychology even somewhat like the human is an enormously speciﬁc property, and hence one that is a priori unlikely to characterize the ﬁrst cause of all. Second, Hume notes that our sentimental psychology is correlated with and adapted to a highly particular form of life in a highly particular environment, and takes this fact to further compound the improbability that the deity has a sentimental psychology like ours. This second stage of the argument implicates the empirical premise that our psychology is indeed correlated with and adapted to our environment, and it clearly involves analogical reasoning from this observation to a conclusion about the likely intrinsic character of the divinity. Such a speculative inference may seem to be out of line with

the argument from motivation

143

Hume’s usual skeptical attitude toward the ‘airy sciences’ (EHU 1.12), but I have argued both that Hume does permit a form of experientially grounded natural theology, and that he can do so without straying into the overreaching speculation exempliﬁed by core natural religion. His own arguments concerning the intrinsic character of the ﬁrst cause of all would apply equally to any other unknown being: they claim no particular insight into the distinctive intrinsic nature of the divinity, nor do they offer up the sort of substantive positive conception of the divine attributes that is promised by the traditional constructive natural theologian. Hume’s relatively modest arguments are at least consistent with his main skeptical agenda in this area, the total rejection of core natural theology.

This page intentionally left blank

6 The Arguments from Evil 6.1. Introduction In the previous chapter we examined the argument from motivation, our ﬁrst argument from empirical premises for the conclusion that the deity is indifferent to moral concerns. In the current chapter I look at the evidential argument from evil, a second sort of empirically grounded argument for the indifference of any ﬁrst cause or organizing principle. According to this latter argument, the evils of the world provide circumstantial evidence that the deity is not moved by moral concerns: in the light of such evil, as Philo puts the claim, the hypothesis of divine indifference ‘seems by far the most probable’ (DNR 11.14). The argument thus presents a probabilistic inference from empirical facts about this scurvy and disastrous world of ours to the thesis of weak moral atheism, the view that the deity is not morally praiseworthy. Hume’s treatment of the ‘old questions’ of evil (DNR 10.25) is subtle and complex—more so, perhaps, than is generally appreciated. While Hume certainly sets out the evidential argument sketched above, it is not clear that he means to endorse it himself. In fact, as I will argue, Hume’s version of the evidential argument is best understood, not as a serious and sincere case for divine amorality, but as a parody of traditional core natural theology. The issue is complicated by the other uses to which Hume puts the old questions, for he also sketches arguments concerning the logical problem of evil (the challenge to theists to show the bare possibility that there is a moral deity, given the evils of the world) and the inference problem of evil (the challenge to theists to show how we can infer the existence of a moral deity, given the evils of the world). I will also examine Hume’s purposes in raising the logical and inference problems, and maintain that in neither case does

146 the arguments from evil he intend to offer an argument for moral atheism. Combining this result with my main thesis that his version of the evidential argument is essentially parodic, we can see that no part of Hume’s discussion of evil presents a serious case for moral atheism. Hume is a moral atheist (as I have argued in Chapters 2–5), but his case for this position does not include the familiar species of argument that takes us from premises about the evils of the world to a conclusion asserting the amorality of the deity. Rather, in keeping with his skeptical critique of ‘airy sciences’ and ‘distant and high enquiries’ (EHU 1.12, 12.25), Hume rejects all core natural theology as straying quite beyond the limits of our faculties—and this includes irreligious arguments from the sufferings of the world as well as the more orthodox or pious forms of constructive natural religion. Hume’s interest in the complex of issues traditionally known as the problem of evil spans his philosophical career. Memoranda notes that he drew up between 1737 and the early 1740s¹ testify to a study of Pierre Bayle’s provocative treatment of the issue in the Dictionnaire historique et critique (1697/1702),² along with Archbishop William King’s response in De Origine Mali (1702).³ We also have a recently ¹ For Hume’s memoranda, see E. C. Mossner, ‘Hume’s Early Memoranda, 1729–1740: The Complete Text,’ Journal of the History of Ideas 9 (1948), 492–518. Mossner’s dating of the memoranda is corrected in M. A. Stewart, ‘The Dating of Hume’s Manuscripts’, in Paul Wood (ed.), The Scottish Enlightenment: Essays in Reinterpretation (Rochester: University of Rochester Press, 2000), 276–88. ² Bayle’s discussion of evil is centered in the footnotes to his articles on ‘Manicheens,’ ‘Paulicians,’ ‘Origen,’ and in the ‘Second Clariﬁcation.’ See Pierre Bayle, The Dictionary Historical and Critical of Mr. Peter Bayle, tr. Pierre des Maizeaux, 5 vols (London: J. J. and P. Knapton, 1734–38) (Reprint: New York: Garland Publishing, 1984), iv. 90–97, iv. 412–22, iv. 512–28. For commentary, see Susan Neiman, Evil in Modern Thought: An Alternative History of Philosophy (Princeton, NJ: Princeton University Press, 2002), 116–28; John-Pierre Jossua, Pierre Bayle ou l’obsession du mal (Paris: Aubier Montaigne, 1977). Bayle’s own theological position is the subject of much dispute. He is sometimes taken at his word as a sincere Calvinist who embraces the deity’s moral unintelligibility as providing yet another argument for a non-rational Christian ﬁdeism. Others regard his avowed ﬁdeism as ironic or evasive, and take his obvious relish in cataloging the moral incomprehensibility of the Christian god to reﬂect an underlying religious skepticism. For the former reading, see Richard H. Popkin, The History of Scepticism: From Savonarola to Bayle (Oxford: Oxford University Press, 2003), 290–300, and Thomas M. Lennon, Reading Bayle (Toronto: Toronto University Press, 1999). For the latter interpretation, see David Wootton, ‘Pierre Bayle, Libertine?’ in M. A. Stewart (ed.), Studies in Seventeenth Century Philosophy (Oxford: Oxford University Press, 1997), 197–22, and Gianluca Mori, Bayle Philosophe (Paris: Champion, 1999). ³ William King, An Essay on the Origin of Evil (1731) (Reprint: New York: Garland, 1978). On King and his intellectual context, see James Harris, Of Liberty and Necessity

the arguments from evil

147

discovered fragment on evil in Hume’s hand, which Stewart dates to the same approximate period, and provisionally hypothesizes ‘may be one of the ‘‘noble Parts’’ prudentially removed from the Treatise in 1739 to avoid offense.’⁴ In his published work, Hume directly addresses the issue of evil on three occasions. He touches on the topic in ‘Of Liberty and Necessity’ in the Enquiry concerning Human Understanding (1748), relating it to his immediate topic of determinism and predestination, and sketching out the threat of a deity that is ‘the ultimate author of guilt and moral turpitude in all his creatures’ (EHU 8.33). Second, in ‘Of a Particular Providence and a Future State,’ also in the ﬁrst Enquiry, Hume develops the inference problem of evil, and suggests that, given the character of the world, there is no prospect of proving the moral perfection of the deity from the evidence of the creation. Third, Hume considers the ramiﬁcations of evil at length in the Dialogues concerning Natural Religion (1779; published posthumously), a work he was still revising in his ﬁnal months. The discussion in the Dialogues runs through many pages and addresses each of the three main issues in turn: the logical problem in part 10, and the inference problem and the evidential problem in part 11.⁵ The inﬂuence of Bayle is evident in Dialogues part 10, both in the language Hume uses to frame the topic and in the arguments he employs in examining the logical challenge. Part 11 again echoes Bayle in style if not substance in using the Manichean hypothesis as a way into the evidential problem of evil. Finally, the arguments of part 11 also suggest close, if hardly deferential, attention to Bishop Joseph Butler’s treatment of evil in his celebrated apologetic work The Analogy of Religion, Natural and Revealed, to the Constitution and Course of Nature (1736). This chapter proceeds as follows. Following Hume’s own ordering of the issues in the Dialogues, I ﬁrst review his treatment of the (Oxford: Clarendon Press, 2005), 43–6, and David Berman, ‘Enlightenment and CounterEnlightenment in Irish Philosophy,’ Archiv für Geschichte der Philosophie 64 (1982), 148–65. ⁴ M. A. Stewart, ‘An early fragment on evil,’ in M. A. Stewart and John P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), 160–70: 164–5. ⁵ Hume does not use this modern terminology in distinguishing these three discrete questions within the overall problematic of evil and providence, but the current practice of separating them out in this way owes a good deal to his particularly clear and forceful disambiguation of the issues in the Dialogues.

148 the arguments from evil logical and inferential problems before moving to address his account of the evidential argument from evil. In examining his handling of the logical problem (in Section 6.2), I argue that Hume holds that the existence of evil does rule out the possibility of a god that is both inﬁnitely powerful and morally praiseworthy. But however scandalous this result might seem, Hume regards it as a side-issue of little real interest or consequence. It does not preclude some other sort of (ﬁnitely powerful) praiseworthy deity, and the conception of an inﬁnitely powerful god, however popular with traditional theists, is not a philosophically respectable one that Hume takes seriously in any case. In Section 6.3 I examine Hume’s endorsement of the inference problem of evil. His suggestion here is that no realistic survey of this world can support an a posteriori inference to a morally praiseworthy deity. This blocks a traditional sort of argument for moral theism, but does not of course amount to a positive case for moral atheism. Finally, in Section 6.4 I document Hume’s treatment of the evidential argument of evil. Hume has often been interpreted as using the evidential argument to advance a positive case for divine amorality, and given our interest in Hume’s case for moral atheism, it is the evidential problem that stands as the most important aspect of his overall account of the ramiﬁcations of evil. However, I argue that his real purpose in setting out a version of the evidential argument is simply to parody the traditional program of core natural theology.⁶ A note before I begin. Throughout his discussion of evil and providence, Hume uses the term ‘evil’ simply to refer to the suffering, pain, misery, or discomfort of conscious beings. This hedonic usage permits him to treat the claim that evil exists as an assertion of empirical fact rather than an evaluative judgment, though of course evaluative judgments will arise elsewhere in his discussion of the ramiﬁcations of ‘evil’ for the moral character of the deity. Most of the time, Hume focuses his attention on ‘natural’ rather than ‘moral’ evil: the pain and suffering brought on by natural causes such as famine, disease, ⁶ I examine one other argument from evil in the next chapter in the context of Hume’s account of the implications of determinism for God’s moral status (in ‘Of Liberty and Necessity’ in the ﬁrst Enquiry). This argument exploits determinism in order to show that any original cause not merely fails to prevent but actively causes the various evils of the world (Section 7.5).

the arguments from evil

149

and ﬂood, rather than the pain and suffering that stems from vice and malevolent human action. But this seems to be essentially a rhetorical decision, and on occasion Hume explicitly extends his reasoning to cover moral as well as natural evil (DNR 11.10, 11.16).

6.2. The Logical Problem of Evil The logical problem of evil (also known as the consistency problem of evil) is the challenge, addressed to the traditional theistic believer, of showing that the existence of a morally praiseworthy deity is compatible with the apparently undeniable fact that evil exists in the world. The issue here is not whether the theist can prove the existence of a praiseworthy divinity, or even whether it seems particularly likely: rather the issue is simply whether they can maintain its bare possibility given the numerous ills of our world. Hume nowhere suggests that there is a general logical problem in supposing that a praiseworthy original cause or organizing principle exists alongside the empirical facts of evil. Indeed in part 11 he explicitly states that, given our limited understanding of the world, we may be unaware of all sorts of reasons why even a ‘very good, wise, and powerful’ deity has to permit evil: thus ‘a limited intelligence must be sensible of his own blindness and ignorance, and must allow that there may be many solutions of these phenomena which will forever escape his comprehension’ (DNR 11.2, see also 11.4, 5.6). So some sort of morally praiseworthy deity is at least epistemically possible, even in the face of evil. However, in part 10 of the Dialogues Hume repeatedly and unambiguously insists, through his spokesman Philo, that the existence of evil does preclude the possibility that there is an inﬁnitely powerful and praiseworthy deity. Since this interpretation has been contested,⁷ it is well to have the essential texts before us. ⁷ Paul Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion (New York: Oxford University Press, 2007), 280; David O’Connor, Hume on Religion (London: Routledge, 2001), 175–7; William Lad Sessions, Reading Hume’s Dialogues: A Veneration for True Religion (Bloomington: and Indianapolis: Indiana University Press, 2002), 156; Klaas J. Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ Hume Studies 29 (2003), 283–304: 285; William H. Capitan, ‘Part X of Hume’s Dialogues,’ American Philosophical

150 the arguments from evil Philo’s ﬁrst main statement of the logical problem facing the hypothesis of an inﬁnitely powerful and praiseworthy god runs as follows: And is it possible, Cleanthes, said Philo, that after all these reﬂections [on the various evils that plague our world], and inﬁnitely more, which might be suggested, you can still persevere in your anthropomorphism, and assert the moral attributes of the deity, his justice, benevolence, mercy, and rectitude, to be of the same nature with these virtues in human creatures? His power we allow is inﬁnite: Whatever he wills is executed: But neither man nor any other animal is happy: Therefore he does not will their happiness. His wisdom is inﬁnite: He is never mistaken in choosing the means to any end: But the course of Nature tends not to human or animal felicity: Therefore it is not established for that purpose. Through the whole compass of human knowledge, there are no inferences more certain and infallible than these. In what respect, then, do his benevolence and mercy resemble the benevolence and mercy of men? Epicurus’ old questions are yet unanswered. Is he willing to prevent evil, but not able? then is he impotent. Is he able, but not willing? then is he malevolent. Is he both able and willing? Whence then is evil? (DNR 10.24–5)⁸

Hume’s language is quite categorical. (‘Through the whole compass of human knowledge, there are no inferences more certain and infallible than these.’) And here, as in the other places where he underlines the current argument, Hume is quite clear that this speciﬁc logical problem concerns a being of inﬁnite power (DNR 10.27, 10.34, 10.35). Consider his second main statement of the argument: But allowing you, what never will be believed; at least, what you never possibly can prove, that animal, or at least human happiness, in this life, exceeds its misery, you have yet done nothing; For this is not, by any means, what we Quarterly 3 (1966), 82–5: 84. These commentators fail to distinguish Hume’s discussion of an inﬁnitely powerful deity in part 10 (where he does insist that the logical problem of evil is quite unanswerable) and his discussion of a ﬁnitely powerful deity in part 11 (where he does not). ⁸ We have no text by Epicurus that sets out these questions, though Lactantius (c.250–c.325), the Christian apologist and imperial tutor at the court of Constantine, ascribes them to the ancient atomist in just these terms in book 13 of his De ira dei. Hume might well have encountered Lactantius’s statement of the Epicurean challenge in Bayle’s article ‘Paulicians’ (note E) in the 1697/1702 Dictionnaire historique et critique. (Pierre Bayle, Historical and Critical Dictionary, ed. and tr. Richard H. Popkin (Indianapolis: Hackett, 1991), 169.)

the arguments from evil

151

expect from inﬁnite power, inﬁnite wisdom, and inﬁnite goodness. Why is there any misery at all in the world? Not by chance surely. From some cause then. Is it from the intention of the deity? But he is perfectly benevolent. Is it contrary to his intention? But he is almighty. Nothing can shake the solidity of this reasoning, so short, so clear, so decisive . . . (DNR 10.34)

A praiseworthy being with inﬁnite power would not permit any evil, and again Hume uses quite forceful language to drive home this ‘decisive’ logical challenge to the hypothesis of an inﬁnitely powerful and praiseworthy deity. More texts will conﬁrm this interpretation. But ﬁrst I need to address a complication and possible objection. I just cut off Hume’s paragraph mid-sentence, and in fact it goes on to conclude with what might look like a retraction of the logical challenge: Nothing can shake the solidity of this reasoning, so short, so clear, so decisive; except we assert, that these subjects exceed all human capacity, and that our common measures of truth and falsehood are not applicable to them; a topic which I have all along insisted on, but which you [Cleanthes] have, from the beginning, rejected with scorn and indignation. (DNR 10.34, emphasis mine)

Here Hume might seem (contrary to my own interpretation) to be backing off his otherwise strongly expressed commitment to the logical objection to the hypothesis of an inﬁnitely powerful and praiseworthy god. The suggestion would be that Hume is granting that there is a way of avoiding this logical challenge after all. The theist can always claim that ‘these subjects exceed all human capacity’ and thus avoid admitting the absolute inconsistency of evil with the existence of a praiseworthy and inﬁnitely powerful god: perhaps there is some way of reconciling a praiseworthy god of inﬁnite power with the evils of the world, even if we (in our ignorance) cannot quite see how. However, we make better sense of Hume’s overall position and language if we read this qualiﬁcation simply as allowing that we remain free to say that such an inﬁnitely powerful deity is benevolent, merciful, and the rest, only if we admit that that these divine attributes bear no signiﬁcant analogy to human benevolence, mercy, and so on, and hence that we have no real conception of what we mean in ascribing these attributes to such a deity (i.e. ‘these subjects exceed all human capacity, and . . . our common measures of truth and falsehood are not applicable to them’). After all, in ﬁrst setting out the logical argument against an inﬁnitely

152 the arguments from evil powerful and praiseworthy being, Hume was quite clear that the problem would only arise if we ‘persevere in . . . anthropomorphism’ and take divine ‘justice, benevolence, mercy, and rectitude, to be of the same nature with those virtues in human creatures’ (DNR 10.24; see also 11.1). A few paragraphs further on, after rehearsing yet more woes of the world and shortly before the supposed retraction in DNR 10.34, he underlines the point once again: How then does the divine benevolence display itself, in the sense of you anthropomorphites? None but we mystics, as you were pleased to call us, can account for this strange mixture of phenomena, by deriving it from attributes, inﬁnitely perfect, but incomprehensible. (DNR 10.27, emphases mine)

Thus when Philo notes that the only possible reply to the logical problem is to embrace a ‘topic which I have all along insisted on, but which you [Cleanthes] have, from the beginning, rejected with scorn and indignation’ (DNR 10.34), he is surely referring to the difﬁculty in maintaining any anthropomorphic conception of the divine attributes—a difﬁculty that Philo has been pressing throughout the Dialogues, going back to his initial position statement back in part 2: [I] piously ascribe to [the original cause of this universe] every species of perfection. . . . But as all perfection is entirely relative, we ought never to imagine, that we comprehend the attributes of this divine being, or to suppose that his perfections have any analogy or likeness to the perfections of a human creature. . . . [L]et us beware, lest we think, that our ideas anywise correspond to his perfections, or that his attributes have any resemblance to these qualities among men. (DNR 2.4; see also 11.16, 12.8)

Given this interpretation of DNR 10.34, then, the ‘way out’ of the logical problem is no way out at all. We can call an inﬁnitely powerful deity good if we want to, but only if we join the ‘mystics’ in admitting that divine ‘goodness’ is a totally incomprehensible characteristic. This simply permits us to retain a ﬁg leaf of traditional pious language while conceding the crucial point that the deity cannot be both inﬁnitely powerful and good in any humanly recognizable sense. In the closing paragraphs of part 10, Hume switches his agenda from this logical problem to the inference problem, and it has been suggested that in so doing he backs off his endorsement of the former

the arguments from evil

153

challenge.⁹ Here commentators have cited an apparent concession on Philo’s (and hence Hume’s) part: ‘I will allow, that, pain or misery in man is compatible with inﬁnite power and goodness in the deity, even in your sense of these attributes’ (DNR 10.35). But in fact the full text here points quite the other way. It shows that Hume is merely prepared to temporarily suspend this logical objection while he makes a separate point about the inference problem of evil; and indeed it makes it clear that he really ﬁnds the logical objection perfectly irrefutable, an impregnable salient running straight through the theist’s defensive lines: But I will be contented to retire still from this entrenchment: For I deny that you can ever force me in it. I will allow, that, pain or misery in man is compatible with inﬁnite power and goodness in the deity, even in your sense of these attributes: What are you advanced by all these concessions? A mere possible compatibility is not sufﬁcient. You must prove these pure, unmixed, and uncontrollable attributes from the present mixed and confused phenomena, and from these alone. A hopeful undertaking! (DNR 10.35, initial emphasis mine)

No counterargument can ‘force’ Philo to abandon his position and retract the logical objection to an inﬁnitely powerful and praiseworthy divinity: it is just that he is prepared to voluntarily waive the point in order to draw our attention to another, entirely distinct issue—namely, the inference problem of evil. One ﬁnal text conﬁrms the current interpretation. Not only does Hume have Philo endorse the logical challenge to the existence of an inﬁnitely powerful and praiseworthy deity through the close of part 10; right at the start of part 11 he also has Cleanthes accept Philo’s assessment before switching topics over to the very different question of a ﬁnitely powerful deity. I scruple not to allow, said Cleanthes, that I have been apt to suspect the frequent repetition of the word, inﬁnite, which we meet with in all theological writers, to savour more of panegyric than of philosophy, and that ⁹ Russell, The Riddle of Hume’s Treatise, 280; O’Connor, Hume on Religion, 175–6; Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ 285. Kraay also cites evidence from part 11 (DNR 11.2, 11.4, 11.8, 11.12), but there Hume has explicitly switched the discussion from a being of inﬁnite power to one of ﬁnite power (DNR 11.1–2).

154 the arguments from evil any purposes of reasoning, and even of religion, would be better served, were we to rest contented with more accurate and more moderate expressions. The terms, admirable, excellent, superlatively great, wise, and holy; these sufﬁciently ﬁll the imaginations of men; and anything beyond, besides that it leads into absurdities, has no inﬂuence on the affections or sentiments. Thus, in the present subject, if we abandon all human analogy . . . I am afraid we abandon all religion, and retain no conception of the great object of our adoration. If we preserve human analogy, we must for ever ﬁnd it impossible to reconcile any mixture of evil in the universe with inﬁnite attributes; much less can we ever prove the latter from the former. (DNR 11.1, ﬁnal emphasis mine)

If we ‘preserve human analogy’—if we think of divine ‘goodness’ as having any resemblance to human goodness—then the logical objection to a praiseworthy and inﬁnitely powerful deity is quite unanswerable. So Hume fully endorses the logical problem of evil as a refutation of the speciﬁc hypothesis of an inﬁnitely powerful and praiseworthy god: his advocacy of this argument is in fact just as full-blooded as Bayle’s notorious enforcement of this logical challenge in the Dictionnaire historique et critique. But there remains something of a twist to Hume’s handling of the logical problem. While orthodox believers may be scandalized by his conclusion that the existence of evil rules out an inﬁnitely powerful and praiseworthy god, and while philosophers such as Bayle, King, and Leibniz may have devoted enormous tracts to the systematic dissection and cross-analysis of the issue, Hume in fact regards the whole topic as something of a philosophical sideshow, an academic dust-up of no real consequence. After all, as he emphasizes, it clearly does not preclude the possibility of an original cause that is ﬁnitely powerful and praiseworthy (DNR 11.1–2, 11.4)—and for Hume this hypothesis of a ﬁnitely powerful praiseworthy deity is the real issue at stake, the only question worth taking seriously. There are two main reasons for this. First, however central the notion of an inﬁnitely powerful creator-god has become in popular monotheistic traditions, there is not (in Hume’s accounting) the shadow of an argument for such a being. The only natural theology with even a semblance of plausibility is the sort of a posteriori ‘experimental theism’ hazarded by Cleanthes (DNR 5.2), since this program of reasoning from the character of the world to the probable character of its original cause at least accepts the regulative ideal of controlling our

the arguments from evil

155

cosmological speculations by empirical evidence (even if experimental theists like Cleanthes do not always live up to this ideal in practice). But experimental theism cannot hope to demonstrate the existence of an inﬁnitely powerful being, for it provides ‘no reason to ascribe to . . . celestial beings any perfection or any attribute, but what can be found in the present world’ (EHU 11.16, see also EHU 11.12–15). As Hume puts it in the Dialogues: as the cause ought to be proportioned to the effect, and the effect, so far as it falls under our cognizance, is not inﬁnite; what pretensions have we . . . to ascribe that attribute to the divine being? (DNR 5.5)

In fact the only reason that philosophers have to treat the hypothesis of an inﬁnitely powerful being as anything more than an arbitrary and rather whimsical speculation is due to the popularity of this conception in the various historical monotheistic traditions. But this popular historical conception is simply a reﬂection of the hyperbolic language used by pious ﬂatterers to curry favor with the divinity. Consider Hume’s account in the Natural History of Religion (1757) of the historical genesis of the notion of an inﬁnitely powerful god from an initial belief in a tutelary spirit or suzerain among heavenly beings: 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. (NHR 6.5) [T]he same anxious concern for happiness, which begets the idea of these invisible, intelligent powers, allows not mankind to remain long in the ﬁrst simple conception of them; as powerful, but limited beings; masters of human fate, but slaves to destiny and the course of nature. Men’s exaggerated praises and compliments still swell their idea upon them; and elevating their deities to the utmost bounds of perfection, at last beget the attributes of unity and inﬁnity, simplicity and spirituality. (NHR 8.2; see also NHR 7.1, 13.2)

156 the arguments from evil Thus it is ﬂattery and panegyric, not argument or reﬂection, that brings the theory of an inﬁnite being to the fore. And while the Natural History focuses here on the popular monotheism of ‘the vulgar,’ in the ﬁrst Enquiry and Dialogues Hume similarly charges religious philosophers with indulging this sort of exuberant hyperbole and, in formulating sublime notions of the deity that far outstrip any possible empirical conﬁrmation, of ‘call[ing] in the assistance of exaggeration and ﬂattery to supply the defects of argument and reasoning’ (EHU 11.14; see also EHU 11.27). Cleanthes no less than Philo sees matters this way, suggesting (as we just saw) that ‘the frequent repetition of the word inﬁnite, which we meet with in all theological writers, . . . savour[s] more of panegyric than of philosophy’ (DNR 11.1).¹⁰ Second, it is in any case doubtful whether traditional monotheism—in either its popular or its philosophical forms—actually succeeds in presenting us with a coherent and intelligible concept of an inﬁnitely powerful being. Certainly Hume doubts whether ‘the vulgar’ have a genuine notion of an inﬁnite being corresponding to their hyperbolic language and pious panegyric.¹¹ But he is also more generally suspicious of the suggestion that the human mind can support any clear and self-consistent concept of the inﬁnite—as is evidenced by his hostility to mathematical arguments for the inﬁnite divisibility of space and time (T 1.2.1, EHU 12.18–20), and also, perhaps, by Cleanthes’s view that the use of inﬁnity in theological writings ‘leads into absurdities’ (DNR 11.1).¹² Taking these points together, we can ¹⁰ Cleanthes never claims that the ﬁrst cause of all is inﬁnitely powerful, but merely asserts that ‘the author of nature . . . is possessed of much larger faculties [than the human], proportioned to the grandeur of the work which he has executed’ (DNR 2.5). While Philo relishes pointing out that Cleanthes’s own method of ‘experimental theism’ (DNR 5.2) could never vindicate the traditional hypothesis of an inﬁnitely powerful divinity (DNR 5.5), this point need not disturb Cleanthes himself. ¹¹ ‘[W]e may observe, that the assent of the vulgar [in complying with ‘the higher encomiums’] is . . . merely verbal, and that they are incapable of conceiving those sublime qualities, which they seemingly attribute to the Deity. Their real idea of him, notwithstanding their pompous language, is still as poor and frivolous as ever’ (NHR 7.1). ¹² On Hume’s travails with inﬁnities, see Dale Jacquette, David Hume’s Critique of Inﬁnity (Leiden: Brill Academic Publishers, 2001). For the wider early modern debates over the paradoxes of inﬁnity, see Thomas Holden, The Architecture of Matter: Galileo to Kant (Oxford: Clarendon Press, 2004), 40–4, and A. W. Moore, The Inﬁnite, 2nd edn (London and New York: Routledge, 2001), chs 5 and 6.

the arguments from evil

157

see that, for Hume, the great debates over the logical problem of evil and the possibility of an inﬁnitely powerful and praiseworthy god are really a monumental distraction. The issue that so vexed philosophers from Origen and Augustine through to Bayle, King, and Leibniz has been elaborated and analyzed at baroque and tedious length, but is really a tempest in a teacup. For Hume, the empirical facts of human suffering quite obviously preclude the existence of a being of inﬁnite power and goodness: Bayle is exactly right in his assessment of the logic of the situation. But for sober philosophers such as Philo and Cleanthes who are free from prejudice and the contagion of popular opinion, the hypothesis of an inﬁnitely powerful being is in any case empirically unjustiﬁed, absurdly extravagant, and of doubtful intelligibility. Serious debates over evil and the moral character of the ﬁrst cause will not get bogged down in this side-issue, then, but proceed to the wider question of how the facts of evil might inform a genuine experimental theism, and what sort of evidence they provide for or against the hypothesis of any moral deity, be it of ﬁnite or inﬁnite power. The few pages of the Dialogues devoted speciﬁcally to the logical version of the problem of evil (DNR 10.24–35) are just a warm-up for these more pressing questions.¹³ Is Hume right to hold that the existence of evil precludes the possibility of an inﬁnitely powerful and praiseworthy god? This might seem a rather dogmatic position for him to take, particularly given his own doubts about whether humans have an adequate grasp of inﬁnity, along with his usual tendency toward skepticism and difﬁdence in cosmological and theological enquiry. As we have seen in previous chapters, Hume is certainly entitled to advance conceptual arguments about the combinations of properties that no original cause could possibly have. (Notice that the logical argument against an inﬁnitely powerful and praiseworthy deity simply involves the sort of liminal natural theology sketched out in Section 2.4, for it turns on the wider claim that no being could simultaneously be inﬁnitely powerful, ¹³ The remainder of part 10—which precedes the discussion of the logical problem of evil—is devoted to cataloging the various evils of the world, i.e. the data that drives all the versions of the problem of evil, logical, inferential, and evidential. Hume explicitly shifts the discussion to the consideration of a merely ﬁnitely powerful deity in part 11 (DNR 11.1), and there the discussion is limited to the inferential and evidential versions of the problem of evil.

158 the arguments from evil morally praiseworthy, and yet tolerate the existence of evil, here applied to the particular case of the ﬁrst cause of all.) But we might wonder whether this particular conceptual argument involves claims that go beyond the limits of human knowledge. True, insofar as Hume understands evil simply in terms of the suffering of conscious beings, it does seem plausible to say that an inﬁnitely powerful being would be able to completely prevent it. But it is less obvious that any morally praiseworthy being would necessarily want to end all suffering: perhaps there is a morally justifying reason to permit some such evil, even if we cannot think what it would be. (‘And, spite of Pride, in erring Reason’s spite / One truth is clear, ‘Whatever is, is right’.’¹⁴) The question takes us beyond the conﬁnes of this study, but it is at least worth noting that the consensus view among philosophers has in recent years moved sharply away from Hume’s (and Bayle’s) assessment that the logical argument against an inﬁnitely powerful and praiseworthy being is quite unanswerable.¹⁵

6.3. The Inference Problem of Evil The inference problem of evil is the challenge, leveled against those who claim that the universe attests to a morally praiseworthy creator, to show just how we can infer the deity’s moral attributes from a world plagued by evil. The inference problem is not addressed to those who would justify belief in a moral god without employing such empirical evidence: exclusively by aprioristic metaphysical argument, say, or purely by appeal to the authority of scripture. Nor will it disturb those who are prepared to make a leap of faith in the absence of any rational support. But there are many early modern philosophers who ¹⁴ Alexander Pope, An Essay on Man (London: J. Wilford, 1733–4), epistle 1, lines 293–4. ¹⁵ Nelson Pike, ‘Hume on Evil,’ The Philosophical Review 72 (1963), 180–97: 182–92; John Hick, Evil and the God of Love, rev edn: (New York: Harper & Row, 1977); Alvin Plantinga, God, Freedom and Evil (Grand Rapids: Eerdmans, 1977), 29–59; Eleonore Stump, ‘The Problem of Evil,’ Faith and Philosophy 2 (1985), 392–423. For a defense of the view that the existence of evil does preclude an inﬁnitely powerful and praiseworthy god, see John Mackie, The Miracle of Theism (Oxford: Clarendon Press, 1982), 150–76. Compare also John Stuart Mill’s particularly strong statement of this logical challenge in his essay ‘Theism,’ in Collected Works, 33 vols (Toronto: University of Toronto Press, 1963), x. 455–6.

the arguments from evil

159

argue from the character of the world to the moral attributes of its original cause, as when George Berkeley infers that the deity is ‘wise, powerful, and good, beyond comprehension’ from ‘the variety, order, and manner’ of empirical phenomena; when Francis Hutcheson argues that ‘The whole structure of the world . . . seem[s] to have been built to a benevolent design and to have been intended to create or to preserve life and happiness’; and when Joseph Butler appeals to ‘the known constitution and course of things’ as attesting to God’s moral attributes.¹⁶ The inference problem is targeted against such thinkers and their a posteriori case for moral theism. According to Hume, the sheer quantity of evil in the world embarrasses any attempt to reason from the character of the creation to the moral nature of its original cause (DNR 11.2, 11.4). He grants, certainly, that it is epistemically possible that our world is the creation of a deity that is morally praiseworthy but only ﬁnitely powerful: a being whose power is limited in certain ways, and thus cannot prevent all evil, or at least has to tolerate some evil in the service of other moral goals (DNR 11.1–2). As Cleanthes says, ‘A less evil may then be chosen, in order to avoid a greater: Inconveniences be submitted to, in order to reach a desirable end’ (DNR 11.1). But the possibility that such a god exists is one thing, and positive evidence that it does so is quite another. In his ﬁrst argument intended to block any inference from the character of the world to the moral praiseworthiness of its creator, Hume maintains that, in advance of the fact, we would not have expected a moral deity to produce a world such as this; and thus that, after the fact, we can hardly take this sort of world to provide positive evidence for the moral nature of its primary cause. Is the world, considered in general, and as it appears to us in this life, different from what a man, or such a limited being, would, beforehand, expect from a very powerful, wise, and benevolent deity? It must be strange prejudice to ¹⁶ George Berkeley, Three Dialogues between Hylas and Philonous, in The Works of George Berkeley, Bishop of Cloyne, ed. A. A. Luce and T. E. Jessop, 9 vols (London: Nelson and Sons, 1948–57), ii. 215; Francis Hutcheson, A Synopsis of Metaphysics Comprehending Ontology and Pneumatology (1742), in Francis Hutcheson, Logic, Metaphysics, and the Natural Sociability of Mankind, ed. James Moore and Michael Silverthorne (Indianapolis: Liberty Fund, 2006), 57–187: 174; Joseph Butler, The Analogy of Religion Natural and Revealed to the Constitution and Course of Nature (1736) in The Works of Bishop Butler, ed. J. H. Bernard, 2 vols (London: Macmillan and Co., 1900), ii. 8.

160 the arguments from evil assert the contrary. And from thence I conclude, that, however consistent the world may be, allowing certain suppositions and conjectures, with the idea of such a Deity, it can never afford us an inference concerning his existence. The consistency is not absolutely denied, only the inference. Conjectures, especially where inﬁnity is excluded from the divine attributes, may perhaps be sufﬁcient to prove a consistence; but can never be foundations for any inference. (DNR 11.4, see also 11.2)

Hume’s point here is surely quite trenchant. If the world is radically different from that which we might reasonably have expected from a moral creator, then however compatible it might be with a moral deity, it can hardly provide good evidence for this sort of original cause. This inference on Hume’s part seems correct, and he clearly views the antecedent proposition as quite secure. Still, perhaps to overcome any lingering ‘strange prejudice,’ Hume goes on to enumerate the various apparent shortcomings of the creation: those aspects of the world where we might have expected better from a benevolent creator. Thus ‘the greatest part of the ills, that molest sensible creatures’ stem from causes which, so far as we humans can tell, are not ‘in the least degree, necessary or unavoidable; nor can we suppose them such, without the utmost license of imagination’ (DNR 11.5). Degrees of pleasure alone could have guided animal conduct without recourse to pain (DNR 11.6); general laws could have given way to a system of particular divine volitions to avoid suffering (DNR 11.6–8); creatures could have been endowed more generously with powers and faculties to protect themselves (DNR 11.9–10); and the ‘springs and principles’ of the universe could have been devised without the ‘coarse . . . strokes’ and ‘inaccurate workmanship’ that result in natural disasters, ill-health, and self-destructive human passions (DNR 11.11). So far as humans can judge of such things, then, it seems that much of the suffering plaguing our world might easily have been done away with. This bolsters the premise that we would not have anticipated a world like ours from a moral deity, and thus the conclusion that we cannot infer that the deity is morally praiseworthy from the evidence of the world. As would have been evident to his original eighteenth-century audience, Hume is here directly confronting Joseph Butler’s case in his well-known apologetic masterpiece The Analogy of Religion (1736). This work is particularly interesting for our purposes, since

the arguments from evil

161

it takes the existence of an intelligent designer for granted (Butler is arguing against deists who are already agreed on this point) and instead concentrates on arguing that the order of nature betokens a moral providence—on presenting, cautiously, probabilistically, and incrementally, an empirical case for moral theism. In fact the Analogy (which C. D. Broad characterizes as ‘perhaps the ablest and ﬁnest argument for theism that exists’¹⁷) provides the most systematic and sustained defense of the a posteriori inference to the moral attributes of the deity in the early modern period. It also attempts to preempt the sort of inference problem of evil that Hume would subsequently prosecute—and Hume, who knew Butler’s work, in turn provides his own counterarguments intended to block such possible maneuvers.¹⁸ On the current point, Butler is quite forceful in his criticism of those who would presume to speculate—as Hume later would—about the ways in which the world might have been better constructed to avoid evil: there are [those] who indulge themselves in vain and idle speculations, how the world might possibly have been framed otherwise than it is; and upon supposition that things might, in imagining that they should, have been disposed and carried on after a better model, than what appears in the present disposition and conduct of them. Suppose now a person of such a turn of mind, to go on with his reveries, till he had at length ﬁxed upon some particular plan of Nature as appearing to him the best. One shall scarce be thought guilty of detraction against human understanding, if one should say, even beforehand, that the plan which this speculative person would ﬁx upon, though he were the wisest of the sons of men, probably would not be the very best, even according to his own notions of best. . . . However, it may not be amiss once for all to see, what would be the amount of these emendations and imaginary improvements upon the system of Nature, or how far they would mislead us. And it seems there could be no stopping, till we came to some such conclusions as these: that all creatures should at ﬁrst be made as perfect and as happy as they were capable of ever being: that nothing, to be sure, of hazard or danger should be put upon them to do; some indolent persons would perhaps think nothing at all: or certainly, that effectual care ¹⁷ C. D. Broad, Five Types of Ethical Theory (London: Routledge, 1930), 5. For a catalog of further plaudits, see Paul Russell, ‘Butler’s ‘‘Future State’’ and Hume’s ‘‘Guide of Life’’ ,’ Journal of the History of Philosophy 42 (2004), 425–48: 427–8. ¹⁸ On Hume’s familiarity with Butler and the Analogy, see Paul Russell, ‘Butler’s ‘‘Future State’’ ,’ 428–32.

162 the arguments from evil should be taken, that they should, whether necessarily or not, yet eventually and in fact, always do what was right and most conducive to happiness . . . Now, without considering what is to be said in particular to the several parts of this train of folly and extravagance; what has been above intimated, is a full and direct general answer to it, namely, that we have not faculties for this kind of speculation.¹⁹

In short, ‘the idle and not very innocent employment of forming imaginary models of a world’ quite overreaches the limits of our human faculties: we are really in no position to judge if our proposed ‘improvements’ would in fact improve the creation.²⁰ But as an advocate of the a posteriori inference to the deity’s moral character, Butler is on thin ice here, as Hume does not hesitate to point out. Hume is himself quite prepared to admit that all his speculations about ways of removing evil may be quite wrong: there may, quite possibly, be unknown reasons why a morally praiseworthy divinity cannot avoid such ills. But let us still assert, that as this goodness is not antecedently established, but must be inferred from the phenomena, there can be no grounds for such an inference, while there are so many ills in the universe, and while these ills might so easily have been remedied, as far as human understanding can be allowed to judge on such a subject. I am sceptic enough to allow, that the bad appearances, notwithstanding all my reasonings, may be compatible with such attributes as you suppose: But surely they can never prove these attributes. Such a conclusion cannot result from scepticism; but must arise from the phenomena, and from our conﬁdence in the reasonings, which we deduce from these phenomena. (DNR 11.12)²¹ ¹⁹ Butler, Analogy, 6–7 (introduction, paras 9–10). ²⁰ Butler, Analogy, 8 (introduction, para. 11); see also 119–29 (part 1, ch. 7), and Butler, Fifteen Sermons, ed. W. R. Matthews (London and Edinburgh: Morrison and Gibb, 1914), 232–5 (sermon 15 paras 5–7). On Butler’s account of and appeal to the limits of human understanding, see Terence Penelhum, ‘Butler and Human Ignorance,’ in Christopher Cunliffe (ed.), Joseph Butler’s Moral and Religious Thought: Tercentenary Essays (Oxford: Clarendon Press, 1992), 117–39. See also Penelhum, Butler (London: Routledge and Kegan Paul, 1985), 187–207. ²¹ Similarly, even if one is ‘fully convinced of the narrow limits of his understanding . . . this will not help him in forming an inference concerning the goodness of superior powers, since he must form that inference from what he knows, not from what he is ignorant of. The more you exaggerate his weakness and ignorance, the more difﬁdent you render him, and give him the greater suspicion, that such subjects are beyond the reach of his faculties’ (DNR 11.2).

the arguments from evil

163

So Butler in fact faces something of a dilemma. If we can form judgments, at least provisionally, about whether or not our world is well suited to avoid evil and promote good, then Hume’s catalog of apparent shortcomings in the world stands, and the fact is we would not have expected a world like this from a moral god. But if we cannot form such judgments, then Butler’s own argument from the character of the world to the moral attributes of its original cause is stalled from the outset. Of course, Butler could always try to jump the horns of the dilemma by claiming that, on the one hand, we know enough to infer a morally praiseworthy deity from the admirable aspects of the creation, but, on the other, we know too little to trust our judgment that there are less admirable aspects to the universe that might compromise such an inference.²² Hume has not shown that this particular balancing act cannot be performed, and so his argument falls thus far short of a categorical refutation of any inference from our world to a moral divinity. But he has at least clariﬁed the challenge that Butler (or any other advocate of the a posteriori inference to God’s moral attributes) faces in attempting to pull off this feat without special pleading or double standards. While Butler’s ﬁrst response to the inference problem of evil is to argue that we unable to tell whether the universe has genuine shortcomings, his second response is to argue that we might reasonably have anticipated a world rather like ours from a moral creator. If we can imagine some moral end served by suffering, then we might in fact reasonably expect (or at least not be surprised by) a pattern of such ‘evil’ in a world produced by a praiseworthy original cause. And, according to Butler, there are moral goals, including the punishment of vice and the cultivation of moral improvement, that may require pain and discomfort. While Hume takes the moral status of the deity to be essentially a reﬂection of whether this original being or principle willingly permits suffering, for Butler matters are more complicated: a moral deity might temper an ‘absolute’ ²² Butler’s sermon ‘Of the Ignorance of Man’ appears to suggest this sort of strategy: ‘we may know somewhat concerning the designs of Providence in the government of this world, enough to enforce upon us religion and the practice of virtue: yet, since the monarchy of the universe is a dominion unlimited in extent, and everlasting in duration; the general system of it must necessarily be quite beyond our comprehension’ (Butler, Fifteen Sermons, 234 (sermon 15, para. 6)).

164 the arguments from evil and undiscriminating benevolence with considerations of justice.²³ So pain and uneasiness in the world does not in and of itself preclude the inference to a moral deity. Further, Butler argues that, while the current distribution of suffering does not always appear just (with ‘virtue sometimes prosperous, sometimes depressed; vice sometimes punished, sometimes successful’²⁴), we can nevertheless discern a general tendency even in this life for the virtuous to ﬂourish and the vicious to suffer.²⁵ And given that the world is the production of an intelligent, purposive designer, this tendency points to a preference on the part of this deity for virtue, and serves as an ‘intimation’ of an increasingly just convergence of desert and reward over time, perhaps ultimately culminating in a ﬁnal justice in the afterlife. The notion . . . of a moral scheme of government, much more perfect than what is seen, is not a ﬁctitious, but a natural notion; for it is suggested to our thoughts, by the essential tendencies of virtue and vice. And these tendencies are to be considered as intimations, as implicit promises and threatenings, from the Author of Nature, of much greater rewards and punishments to follow virtue and vice, than do at present. And indeed every natural tendency, which is to continue, but is hindered from becoming effect only by accidental causes, affords a presumption, that such tendency will, some time or other, become effect: a presumption in degree proportionable to the length of the duration, through which such tendency will continue. And from these things together, arises a real presumption, that the moral scheme of government established in Nature shall be carried on much further towards perfection hereafter; and, I think, a presumption that it will be absolutely completed.²⁶

So we need not be surprised that a moral deity permits suffering, and in assessing the moral character of providence we must observe, not only the distribution of pain and pleasure in this world at any given ²³ ‘There may possibly be in the creation, beings, to whom the Author of Nature manifests himself under [the] most amiable of all characters, this of inﬁnite absolute benevolence; for it is the most amiable, supposing it not, as perhaps it is not, incompatible with justice: but He manifests Himself to us under the character of a righteous Governor. He may, consistently with this, be simply and absolutely benevolent . . . : but He is, for He has given us a proof in the constitution and conduct of the world that He is, a Governor over servants, as He rewards and punishes us for our actions’ (Butler, Analogy, 47 (part 1 ch. 3, para. 3)). ²⁴ Butler, Analogy, 64 (part 1, ch. 3, para. 23). ²⁵ Butler, Analogy, 37–8 (part 1, ch. 2, para. 6), 50, 66 (part 1, ch. 3, paras 7, 28). ²⁶ Butler, Analogy, 66–7 (part 1, ch. 3, para. 28).

the arguments from evil

165

moment, but the tendencies and patterns governing these phenomena across time. One rather indirect response to Butler’s scheme is afforded by Hume’s view that all such deontological constraints on benevolence are merely a reﬂection of ‘artiﬁcial’ (i.e. conventional) virtues that have their place and purpose only in speciﬁc human circumstances. Given Hume’s account of justice (in his narrow sense of respect for property and contract), for instance, there is simply no moral reason for the deity, while it stands outside the peculiar human circumstances of a competition over scarce resources and the ability of moral agents to make one another ‘feel the effects of their resentment’ (EPM 3.18), to temper whatever benevolence it might have with ‘the cautious, jealous virtue of justice’ (EPM 3.3). The same point will apply to the concerns of retributive punishment, and indeed any other considerations that might interfere with the free expression of an unlimited and absolute benevolence to all.²⁷ However, in the Enquiry Hume also provides another response more directly targeted at the claim that we can reason from the deity’s apparent preferences in this world to an expectation of better things to come. We can understand this as a second, entirely new argument from Hume intended to drive home the inference problem. Hume’s second argument draws attention to certain principles of empirical reasoning that limit any inference that we can make from the character of a unique effect (such as the universe) to the character of its speciﬁc cause (the deity): When we infer any particular cause from an effect, we must proportion the one to the other, and can never be allowed to ascribe to the cause any qualities, but what are exactly sufﬁcient to produce the effect. A body of ten ounces raised in any scale may serve as a proof, that the counterbalancing weight exceeds ten ounces; but can never afford a reason that it exceeds a hundred. . . . The same rule holds, whether the cause assigned be brute unconscious matter, or a rational intelligent being. If the cause be known only by the effect, we never ought to ascribe to it any qualities, beyond what are precisely requisite to produce the effect: Nor can we, by any rules of just reasoning, return back from the cause, and infer other effects from it, beyond those by which alone it is known to us. (EHU 11.12, 11.13, emphasis mine) ²⁷ See Chapter 5 note 3.

166 the arguments from evil This being the case, the experimental theist must limit his conclusions about the divine attributes to an exact correspondence with the rather mixed evidence before us: Allowing, therefore, the gods to be the authors of the existence or order of the universe; it follows, that they possess that precise degree of power, intelligence, and benevolence, which appears in their workmanship; but nothing farther can ever be proved. . . . So far as the traces of any attributes, at present, appear, so far may we conclude these attributes to exist. The supposition of farther attributes is mere hypothesis; much more the supposition, that, in distant regions of space or periods of time, there has been, or will be, a more magniﬁcent display of these attributes, and a scheme of administration more suitable to such imaginary virtues. . . . The knowledge of the cause being derived solely from the effect, they must be exactly adjusted to each other; and the one can never refer to any thing farther, or be the foundation of any new inference and conclusion. (EHU 11.14, see also 11.15)

So insofar as we are reasoning from the evidence of the universe to facts about its original cause, we can only assign properties to the creator that are fully displayed in the creation: ‘We never can have a reason to infer any attributes, or any principles of action in him, but so far as we know them to have been exerted and satisﬁed’ (EHU 11.21). And so if this world fails to exhibit marks of pure ‘power, intelligence, and benevolence’ throughout, we cannot infer such unalloyed attributes in its creator. The ills of the world must be fully reckoned alongside its advantages, and even a theodicy that successfully explains why a moral deity might allow such evil in the world cannot help us with the inference to divine goodness without presupposing that which we are supposed to be proving. ‘Let your gods, therefore, O philosophers, be suited to the present appearances of nature’ (EHU 11.15, my emphasis). Now, as we have seen, although Butler accepts the constraint of tightly controlling his conclusions by the empirical evidence, he would hold that Hume is not doing justice to the full range of evidence before us. Hume attends to the fact that there is suffering in the world, but he fails to adequately examine the patterns in which that suffering occurs. And for Butler, those patterns disclose a tendency for vice to suffer while virtue ﬂourishes, permitting us to discern the deity’s preference for virtue and look forward to a likely convergence of desert and reward in the future. We need to reﬂect on these anticipated future

the arguments from evil

167

phenomena as well as the current phenomena if we are to get a full picture of the deity’s moral character. Hume himself formulates this Butlerian objection to his own argument as follows: But allowing you to make experience (as indeed I think you ought) the only standard of our judgment concerning this, and all other questions of fact; I doubt not but, from the very same experience, to which you appeal, it may be possible to refute this reasoning [i.e. Hume’s own case against inferring a moral deity from the mixed phenomena of the world] . . . If you saw, for instance, a half-ﬁnished building, surrounded with heaps of brick and stone and mortar, and all the instruments of masonry; could you not infer from the effect, that it was a work of design and contrivance? And could you not return again, from this inferred cause, to infer new additions to the effect, and conclude, that the building would soon be ﬁnished, and receive all the further improvements, which art could bestow upon it? . . . Consider the world and the present life only as an imperfect building, from which you can infer a superior intelligence; and arguing from that superior intelligence, which can leave nothing imperfect; why may you not infer a more ﬁnished scheme or plan, which will receive its completion in some distant point of space or time? Are not these methods of reasoning exactly similar? And under what pretence can you embrace the one, while you reject the other? (EHU 11.24)

Hume’s reply is that the objection does not take sufﬁciently seriously the limitations of our ability to infer facts about the cause of the universe, which is (unlike a half-ﬁnished building) radically dissimilar to all other objects of experience: The inﬁnite difference of the subjects . . . is a sufﬁcient foundation for this difference in my conclusions. In works of human art and contrivance, it is allowable to advance from the effect to the cause, and returning back from the cause, to form new inferences concerning the effect, and examine the alterations, which it has probably undergone, or may still undergo. But what is the foundation of this method of reasoning? Plainly this; that man is a being, whom we know by experience, whose motives and designs we are acquainted with, and whose projects and inclinations have a certain connexion and coherence. . . . But did we know man only from the single work or production which we examine, it were impossible for us to argue in this manner; because our knowledge of all the qualities, which we ascribe to him, being in that case derived from the production, it is impossible they could point to any thing farther, or be the foundation of any new inference. (EHU 11.25)

168 the arguments from evil And this is of course our situation with respect to the ﬁrst cause of all: The Deity is known to us only by his productions, and is a single being in the universe, not comprehended under any species or genus, from whose experienced attributes or qualities, we can, by analogy, infer any attribute or quality in him. As the universe shews wisdom and goodness, we infer wisdom and goodness. As it shews a particular degree of these perfections, we infer a particular degree of them, precisely adapted to the effect which we examine. But farther attributes or farther degrees of the same attributes, we can never be authorised to infer or suppose, by any rules of just reasoning. (EHU 11.26)

Thus, when Butler would infer from the tendency for the vicious to suffer while the virtuous ﬂourish that the deity has preference for virtue, and looks forward to that preference resulting in a still more just distribution of pain and pleasure in the next world, he is assuming a good deal more about the character of the ﬁrst cause than the empirical evidence permits. The universe is sufﬁciently different from all other objects of experience that we cannot place any real degree of conﬁdence in predictions of the behavior of the ﬁrst cause based on analogical reasoning from the behavior of other known causes—and in particular, we ought not tacitly assume that the behavior of this original cause will (like the behavior of a construction team behind a half-ﬁnished building) reﬂect humanly comprehensible sentiments, goals, and attitudes, in addition to the workings of a rational intelligence (EHU 11.27).

6.4. The Evidential Problem of Evil So far Hume has not appealed to the existence of evil to argue for either weak or strong moral atheism. His version of the logical problem of evil is merely intended to show that there could be no inﬁnitely powerful and praiseworthy creator behind this world: he has no pretensions here to provide a general argument for the amorality of any sort of ﬁrst cause or organizing principle. And his version of the inference problem is, of course, simply intended to block one sort of argument for the moral praiseworthiness of the original cause, not to establish moral atheism in its own right. However, at the end of

the arguments from evil

169

Dialogues part 11 Hume suddenly switches tack, and, going beyond the inference problem of evil, appears to positively endorse the theory of divine amorality as ‘by far the most probable’ hypothesis regarding the moral status of the deity (DNR 11.15). He has his spokesman Philo declare that, given ‘the strange mixture of good and ill, that appears in life,’ the most plausible view is that ‘the original source of all things . . . has no more regard to good above ill than to heat above cold, or to drought above moisture, or to light above heavy’ (DNR 11.14). If we take this passage in isolation, Hume appears to be endorsing a version of the evidential argument from evil, indicating that the evils of our world provide probabilistic evidence for weak moral atheism, the view that the deity is not morally praiseworthy. The presentation of this evidential argument runs through just three short paragraphs (DNR 11.14–16). The crucial passage goes as follows: There may four hypotheses be framed concerning the ﬁrst causes of the universe; that they are endowed with perfect goodness, that they have perfect malice, that they are opposite, and have both goodness and malice, that they have neither goodness nor malice. Mixed phenomena can never prove the two former unmixed principles. And the uniformity and steadiness of general laws seem to oppose the third. The fourth, therefore, seems by far the most probable. (DNR 11.15)

So Hume distinguishes four rival hypotheses concerning the moral status of the original cause (or causes) and works through them in turn weighing their respective plausibility. The ‘[m]ixed phenomena’ of the world tells against both the theory of an all-good god and the theory of an all-malevolent god; and the uniformity of nature renders the Manichean theory of some sort of division between good and malevolent principles in the ﬁrst causes of the universe quite unlikely. So we are left with the theory of divine indifference as the most plausible hypothesis. In my view, Hume is not seriously endorsing this argument for divine amorality.²⁸ As we shall see, his mode here is quite ironic. But before examining his real purposes in presenting this sort of evidential ²⁸ For the traditional view that Hume is sincerely endorsing this argument, see J. C. A.Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, New Jersey: Humanities Press International, 1988), 72; Stanley Tweyman, ‘Hume’s Dialogues on

170 the arguments from evil argument, we should ﬁrst consider just how out of character such reasoning would be for Hume. To begin with, the argument is an instance of just the sort of theological speculation that Hume strongly disavows in both the ﬁrst Enquiry and the Dialogues. It purports to take us from facts about the creation to a quite speciﬁc conception of the divine attributes: thus, given the character of its works, we can infer that the original cause of all is (most likely) neither wellintentioned nor malevolent, but quite indifferent to moral concerns. But this argument would seem to be in clear violation of the principles of mitigated skepticism promulgated by Hume in the Enquiry, and apparently endorsed by his spokesman Philo at the beginning of the Dialogues. The mitigated skeptic is supposed to reject all substantive cosmological and theological speculation as straying quite beyond the ‘narrow capacity of human understanding’ (EHU 12.25, see also DNR 1.10–11). And Philo reafﬁrms his commitment to just this sort of epistemic modesty near the beginning of Dialogues part 11, just a few paragraphs before the statement of the evidential argument. We know so little beyond common life, or even of common life, that, with regard to the economy of a universe, there is no conjecture, however wild, which may not be just; nor any one, however plausible, which may not be erroneous. All that belongs to human understanding, in this deep ignorance and obscurity, is to be sceptical, or at least cautious; and not to admit of any hypothesis whatever. (DNR 11.5)

This seems to be in obvious tension with the evidential argument set out subsequently in part 11, and in particular with its methodological presupposition that we are competent to judge which hypothesis concerning the deity’s moral status is ‘the most probable’ (DNR 11.15), or perhaps even the ‘true conclusion’ (DNR 11.14). But is there no way of reconciling the evidential argument of DNR 11.15 with Hume’s own skeptical critique of core natural theology? I do not think there is. First, we cannot explain away this tension by Evil,’ Hume Studies 13 (1987), 74–85: 83–5; David O’Connor endorses this interpretation of Philo, without extending it to Hume himself, in his ‘Skepticism and Philo’s Atheistic Preference,’ Hume Studies 29 (2003), 267–82. Compare also Keith E. Yandell, Hume’s ‘Inexplicable Mystery’: His Views on Religion (Philadelphia: Temple University Press, 1990), 268–75.

the arguments from evil

171

understanding the current argument simply as an instance of (what I have called) liminal natural theology, the relatively unexceptionable form of reasoning about the divine nature that avoids claims about the deity’s distinctive intrinsic character.²⁹ Rather, the evidential argument clearly invites us to move from particular facts about this world to a conclusion about the distinctive intrinsic character of its original cause: a conclusion that ﬁlls in some distinguishing detail about this particular being or principle, and cannot be applied equally to any other unknown object. Second, we cannot reconcile the evidential argument with Hume’s critique of theological speculation by pointing out that it restricts itself to empirical premises, or that it claims to provide only probabilistic support for its conclusion rather than demonstrative certainty.³⁰ Perhaps it does, but so does much of the core natural theology that Hume clearly rejects—including, most obviously, Cleanthes’s argument for the view that there is a close and religiously signiﬁcant analogy between the deity and human intelligence. Third, the tension here cannot be resolved by stressing that the argument of DNR 11.15 involves the assessment of just four rival possible hypotheses, rather than the indeﬁnite number of rival hypotheses that compete with Cleanthes’s conclusion in the argument from design.³¹ Even if the four hypotheses were collectively exhaustive, this would hardly show that we are entitled to judge between them. Beyond this general tension with mitigated skepticism, there is another difﬁculty in supposing that Hume is seriously endorsing this evidential argument from evil. The argument would clearly fall prey to some of the more speciﬁc objections that Philo previously pressed against Cleanthes’s version of the argument from design—objections which still stand unanswered at this stage of the Dialogues, and which Hume apparently regards as quite unanswerable. For instance, in part 2 and again in part 7 Philo objected that our experience is simply too incomplete to permit any reliable conjecture about the overall ²⁹ On liminal versus core natural theology, see Sections 2.4 and 2.5. ³⁰ The reconciliation proposed by David O’Connor, ‘Skepticism and Philo’s Atheistic Preference,’ 273–4. ³¹ The reconciliation proposed by Stanley Tweyman, ‘Hume’s Dialogues on Evil,’ 84–5.

172 the arguments from evil character of the universe, and hence any speculation about the likely character of its original cause: we have no data to establish any system of cosmogony. Our experience, so imperfect in itself, and so limited both in extent and duration, can afford us no probable conjecture concerning the whole of things. (DNR 7.8; see also DNR 2.22–3)

This would seem to be a problem for the evidential argument of DNR 11.15, which would have us weigh what little and parochial evidence we have regarding the overall nature of the universe against the explanatory power of each of the four competing cosmological hypotheses. Further, even if we did have a more comprehensive view of the entire universe, we also lack a wider understanding of the principles of construction for universes in general. We have not seen the workshops of gods and the creation of prior worlds (DNR 2.28), and cannot begin to gauge the sorts of constraints and production difﬁculties facing even a very well-intentioned creator. Once again, this would seem to compromise our ability to evaluate the merits of this particular creation: it is impossible for us to tell, from our limited views, whether this system 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? (DNR 5.6)

Or consider another sort of objection that Philo pressed against Cleanthes in part 5: But were this world ever so perfect a production, it must still remain uncertain, whether all the excellences of the work can justly be ascribed to the workman. If we survey a ship, what an exalted idea must we form of the ingenuity of the carpenter who framed so complicated, useful, and beautiful a machine? And what surprise must we feel, when we ﬁnd him a stupid mechanic, who imitated others, and copied an art, which, through a long succession of ages, after multiplied trials, mistakes, corrections, deliberations, and controversies, had been gradually improving? Many worlds might have been botched and bungled, throughout an eternity, ere this system was struck out: Much labour lost: Many fruitless trials made: And a slow, but continued improvement carried on during inﬁnite ages in the art of world-making. In such subjects,

the arguments from evil

173

who can determine, where the truth; nay, who can conjecture where the probability, lies; amidst a great number of hypotheses, which may be proposed, and a still greater number, which may be imagined? (DNR 5.7, my emphasis)

Once again it seems as if an analogous objection would tell against the current argument for divine amorality. If a perfect work does not license even a probabilistic ‘conjecture’ in favor of a perfect creator, similarly a mediocre or mixed work would not seem to license an inference to a lack of concern on the creator’s part. Perhaps, indeed, this world is simply one of the ‘botched and bungled’ efforts of a well-intentioned but incompetent architect on the way to better productions. So in sum, not only does the evidential argument for divine amorality distilled in DNR 11.15 conﬂict with Hume’s general epistemological critique of cosmological and theological speculation. It also seems to run afoul of some of his own more speciﬁc criticisms of Cleanthes’s ‘experimental theism’ earlier in the Dialogues. Certainly one would not have expected Hume to endorse this sort of empiricallymotivated argument regarding the divine nature, just pages after condemning a relevantly similar form of core natural theology. One ﬁnal point. In addition to conﬂicting with mitigated skepticism and with certain of his own more speciﬁc objections to Cleanthes’s program of experimental theism, the evidential argument of DNR 11.15 seems quite weak on its own terms. Some commentators have argued that the four candidate hypotheses (perfect goodness, perfect malice, an opposition of goodness and malice, and the absence of both goodness and malice) are not exhaustive: there might be an original cause that is mainly good, or mainly malicious, or perhaps a confederacy or competition of multiple causes with various moral characters.³² This is, I think, somewhat uncharitable, for Hume presumably means all such intermediate or mixed positions to be covered under the rubric of the third (very loosely, ‘Manichean’) hypothesis (DNR 11.14). However, the argument also faces more damaging objections. Hume’s case against the ﬁrst two hypotheses is that ‘[m]ixed phenomena can never prove the . . . two unmixed principles’ (DNR 11.15). But ³² Keith Yandell, Hume’s ‘Inexplicable Mystery’, 268–9; William Lad Sessions, Reading Hume’s Dialogues, 174–5; Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ 291.

174 the arguments from evil this seems to be just another version of the inference problem: the now-familiar point that the mixed evidence of the world cannot justify an inference to purely good or purely evil ﬁrst cause. But the issue here in DNR 11.15 is not whether we can prove that the ﬁrst cause is purely good (or purely malicious) from the evidence before us, or even show that one of these theories is probably true. Rather, it is whether, in our impoverished state of information, we can establish that these two hypotheses are less likely than their competitors, and that is quite a different thing. Second, Hume’s case against the third (‘Manichean’) hypothesis is weak. His claim is that ‘the uniformity and steadiness of general laws seems to oppose’ the hypothesis of some sort of mixture of goodness and malice in the ﬁrst causes of the universe (DNR 11.15). But even if we focus on the classic Manichean picture of two warring divinities, one good, one evil, it is not clear that such a conﬂict is likely to manifest itself in a fragmentary, oscillating, or discordant system of laws.³³ Perhaps the good principle is responsible for some uniform laws, the malicious principle for others. Perhaps one is responsible for the laws, and the other for the initial conditions of the universe. Perhaps the uniform laws reﬂect an uneasy truce between stalemated forces. Perhaps the one power or the other has been temporarily subdued, but is gathering its strength in order to reassert itself in the next cosmic age. Perhaps, as Borges wryly speculates, ‘the history of the universe—and in it, our lives and every faintest detail of our lives—is the handwriting of a subordinate god trying to communicate with a demon.’³⁴ One could go on and on. And notice that Philo had previously reveled in generating just such a cascade of competing hypotheses when arguing against Cleanthes’s case for a unitary, intelligent, and purposive deity (see, for instance, DNR 8.1), which makes it all the more surprising that his imagination suddenly fails him now. Third, there is the standing question of whether we are in any position to determine the relative probability of the four competing hypotheses. Philo has already admitted that, given human ignorance, the ﬁrst hypothesis is at least epistemically ³³ Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ 291; Dale Jacquette, ‘Analogical Inference in Hume’s Philosophy of Religion,’ Faith and Philosophy 2 (1985), 287–94: 289. ³⁴ Jorge Luis Borges, Collected Fictions, tr. and ed. Andrew Hurley (New York: Viking Penguin, 1998), 74.

the arguments from evil

175

possible, and presumably he would also allow this of the second and third theories as well as the fourth. But perhaps we should be still more modest: we might wonder whether humans are capable of making even a probabilistic sort of judgment about such enormous cosmological questions. Certainly, one need not be as skeptical as Philo usually is to wonder whether we are really in a position to judge it improbable that there is a morally justifying reason for a praiseworthy deity to tolerate the suffering we ﬁnd in the world. For all these reasons, it seems quite unlikely that Hume seriously endorses the evidential argument from evil in part 11, or that he genuinely holds that we can establish such substantive results about the distinctive intrinsic character of the ﬁrst cause of all. The argument cannot, I think, be completely sincere.³⁵ But how are we then to understand it? One possibility is that Hume is simply waiving some of his usual objections to core natural theology purely for the sake of the argument, in order to show that, even on its own terms, the ‘experimental’ form of core natural theology would point to an amoral deity rather than a praiseworthy one.³⁶ This reading goes a good way toward disarming our textual problem. It grants the crucial point that Hume does not really endorse the evidential argument, and it does a reasonable job of explaining away the apparent textual inconsistencies. However, it still leaves Hume endorsing an embarrassingly weak argument, for even if we bracket all concerns about inconsistencies with his own previous positions, the case he actually provides for preferring the indifference hypothesis still seems quite hopeless. Moreover, it is rather odd that Hume would present this reasoning purely for the sake of the argument without openly ﬂagging the fact that he is suspending his usual skeptical objections. So it is better, I think, to understand a different sort of distancing mechanism to separate the current argument from Hume’s own view. We can more plausibly interpret Hume’s statement of the evidential argument as parody of core natural theology in general, and of Cleanthes’s experimental theism in particular.³⁷ Hume is ³⁵ Here I agree with Nelson Pike, ‘Hume on Evil,’ 194–5, and Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ 288–92. ³⁶ David O’Connor, Hume on Religion, 190. ³⁷ Kraay makes this suggestion in his paper ‘Philo’s Argument for Divine Amorality Reconsidered,’ 292–8. While I previously criticized Kraay’s interpretation of Hume on the

176 the arguments from evil indeed ignoring the principles of mitigated skepticism and his previous criticisms of Cleanthes’s reasoning, but not to show that, once we overlook these earlier objections, the hypothesis of divine amorality has in fact the best ﬁt with the empirical evidence. Rather the point is that, if we do ignore the skeptical objections to core natural theology—freeing our speculative theorizing from a proper difﬁdence about the limitations of our faculties and a proper recognition of how meager our empirical evidence for any ultimate cosmogony really is—then we can easily generate any number of theories that have a superﬁcial air of plausibility. Hume is showing, by provocative example, that natural theology at this point simply degenerates into a wide-open game of analogy-mongering and pseudoscience, driven not by real evidence but by superﬁcially plausible rhetoric, parochial views, and the arbitrary whims of a free-associating imagination. Philo had previously made this basic point in the face of Cleanthes’s preference for the hypothesis of a single intelligent deity with a mind somewhat like the human: In subjects, adapted to the narrow compass of human reason, there is commonly but one determination, which carries probability or conviction with it; and to a man of sound judgment, all other suppositions, but that one, appear entirely absurd and chimerical. But in such questions as the present, a hundred contradictory views may preserve a kind of imperfect analogy; and invention has here full scope to exert itself. Without any great effort of thought, I believe that I could, in an instant, propose other systems of cosmogony, which would have some faint appearance of truth . . . (DNR 8.1)

Of course, in this earlier discussion Cleanthes failed to take Philo’s enumeration of equally viable hypotheses seriously, and clung to his essentially arbitrary preference for the hypothesis of a unitary, mindlike deity all the same. And this blinkered dogmatism is just what Philo is parodying in his own peremptory dismissal of the alternative theories of divine goodness, divine malevolence, and Manicheanism. So, again, his point in sketching out this version of the evidential argument is not that, if we suspend our skeptical scruples and allow ourselves to speculate about the moral status of the ﬁrst cause of all, then the empirical logical problem of evil (see note 9 above), I should stress that on the current point—the interpretation of the evidential argument—his paper is important, textually painstaking, and (to my mind) thoroughly persuasive.

the arguments from evil

177

evidence at some level favors the theory of divine indifference. It is rather that the skeptical objections to natural theology show that our empirical evidence is essentially worthless. We can only transport ourselves ‘into the celestial regions’ and reach such conclusions about the distinctive intrinsic character of the original cause of all if we forget how little we truly know and ‘[aid] the ascent of reason by the wings of the imagination’ (EHU 11.16). With the deity’s moral attributes as with its natural attributes, core natural theology simply exceeds our grasp. This is then the central case for understanding the evidential argument of part 11 as a parody of Cleanthes’s experimental form of core natural theology. It explains the inconsistency between Hume’s usual mitigated skepticism and his presentation of this piece of speculative theological reasoning; it explains why Hume ignores his previous speciﬁc objections that would tell against the current argument; and it explains (as the previous, non-parodic, for-the-sake-of-the-argument interpretation cannot) why the argument is transparently weak even on its own terms. It also helps us see why Hume does not explicitly ﬂag the fact that he is ignoring his usual skepticism about core natural theology, for it is of the nature of a parody to let the point dawn on the reader without explicitly spelling it out. Klaas Kraay has also identiﬁed other points of the text that help to conﬁrm the current interpretation.³⁸ For instance, the spectacularly impertinent passage at DNR 11.13, where Philo shifts from the inference problem and the skeptical, cautious tone of DNR 11.12 toward the coming assertiveness of the evidential argument in DNR 11.14–16, seems to burlesque Cleanthes’s brash conﬁdence in stating the design argument back at DNR 2.5, even down to using the same opening words (‘Look round the world . . . ’; ‘Look round this universe . . . ’).³⁹ Or again, despite ³⁸ Kraay, ‘Philo’s Argument for Divine Amorality Reconsidered,’ 293–7. ³⁹ The two parallel passages: ‘Look round the world: Contemplate the whole and every part of it: You will ﬁnd it to be nothing but one great machine, subdivided into an inﬁnite number of lesser machines, which again admit of subdivisions, to a degree beyond what human senses and faculties can trace and explain. All these various machines, and even their most minute parts, are adjusted to each other with an accuracy, which ravishes into admiration all men who have ever contemplated them’ (DNR 2.5, Cleanthes speaking); ‘Look round this universe. What an immense profusion of beings, animated and organized, sensible and active! You admire this prodigious variety and fecundity. But inspect a little more narrowly these living existences, the only beings worth regarding. How hostile and destructive to each other! How insufﬁcient all of them for their own happiness! How

178 the arguments from evil having condemned Cleanthes’s analogy between the universe and a house as too remote to support any plausible analogical argument (no less than three times!), Philo adopts precisely the same analogy himself here in part 11 in maintaining that we can draw probabilistic conclusions about the moral character of the designer from the empirical evidence before us (DNR 2.8, 2.18, 2.28, 11.3). The point is that if the core natural theologian will permit himself to speculate about the divine attributes from empirical evidence as meager and analogies as attenuated as this, then we can generate theological systems and cosmogonies willy-nilly, be they religiously traditional (like Cleanthes’s hypothesis of a designing intelligence and a moral god) or scandalously unorthodox (like Philo’s hypothesis of an indifferent ﬁrst cause).

6.5. Conclusion Hume does not use the empirical fact of human suffering to argue for either strong or weak moral atheism. Yes, he holds that such suffering logically precludes the existence of an inﬁnitely powerful and praiseworthy original cause (or, indeed, the existence of any inﬁnitely powerful and praiseworthy being). And he holds that the quantity of suffering in the world subverts any attempt to infer a moral deity from the evidence of the created universe. But in neither case is Hume providing a general argument for the impossibility or even the improbability of a morally praiseworthy ﬁrst cause or organizing principle. Hume’s treatment of the evidential problem of evil is a little different, for here he does at least sketch out a case from the existence of suffering to the conclusion that the original cause is most likely amoral. But I have argued that Hume’s statement of this argument is entirely ironic: it is not a seriously intended piece of ‘experimental’ core natural theology, but rather a satire upon such reasoning. While Hume is a moral atheist, he does not hold that we can establish moral atheism by arguing from facts about the world to contemptible or odious to the spectator! The whole presents nothing but the idea of a blind nature, impregnated by a great vivifying principle, and pouring forth from her lap, without discernment or parental care, her maimed and abortive children.’ (DNR 11.13, Philo speaking).

the arguments from evil

179

a substantive, species-speciﬁc conception of the intrinsic character of the deity, a characterization of the divine nature that would distinguish this original cause or organizing principle from any other unknown being. In the recent literature on evil and providence, there has been something of a shift of attention away from the logical problem toward the evidential problem as a more interesting and perhaps more serious challenge to traditional theism. In these discussions, Hume is often mentioned, pro and con, as if he were the original heresiarch behind the evidential argument from evil, or at least its most prominent historical partisan. But if my interpretation is correct, there is a double irony in this. First, while Hume’s Dialogues have been enormously inﬂuential in drawing attention to the evidential argument, he does not actually endorse it himself. And second, the most common response to the evidential argument in the recent philosophical literature is to claim that humans are unable to judge whether evil provides even probabilistic evidence against the hypothesis of a morally praiseworthy god. But (as we can now see) this reply simply echoes Hume’s own skeptical position on the topic.⁴⁰ ⁴⁰ Alvin Plantinga, ‘Epistemic Probability and Evil,’ in Daniel Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington and Indianapolis: Indiana University Press, 1996), 69–96; William P. Alston, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ in Howard-Snyder (ed.), The Evidential Argument from Evil, 97–125; Keith Yandell, Hume’s ‘Inexplicable Mystery’, 275–8.

This page intentionally left blank

7 The Arguments from Determinism 7.1. Introduction In this ﬁnal chapter I examine Hume’s account of the implications of determinism for God’s moral status. Hume sketches out two arguments from this starting place, though a careful reading suggests that he only means to endorse one of them. The shared assumption driving each of the arguments is Hume’s deterministic ‘doctrine of necessity’ (EHU 8.3): his view that exceptionless natural laws govern all human thought and behavior, and that all events, including all human actions, fall out according to an unfolding deterministic causal system. It follows that if there is a deity—an original cause or organizing principle responsible for the entire ordered universe—then this being is the remote, predetermining cause of every human action. Our two arguments each start out from this picture of a deterministic general providence, and each purports to take us to a (different) negative conclusion concerning God’s moral status. We can preview them here. First the authentically Humean argument: call it the argument from determinism against duties to God. According to this argument, if there is an original cause or organizing principle, then it predetermines all human actions, and so none of those actions can be invasions of its rights and prerogatives. The result is that this primary being or principle cannot make moral claims on our behavior, and nor, correlatively, can we owe it any moral duties. This thesis is distinct from both weak and strong moral atheism: it says nothing one way or

182 the arguments from determinism the other about whether the original cause is morally praiseworthy, or whether it is a possible object of moral assessment. Nevertheless, the thesis that the deity cannot make moral claims on human behavior clearly connects to Hume’s wider program of locating the original cause outside the moral community, and thereby sealing off theological speculation from matters of morality and practice. The second argument—which Hume sets out for his own dialectical purposes, but does not ultimately endorse—does present a case for (weak) moral atheism. According to this argument from determinism and evil the thesis that the deity deterministically foreordains everything that ever happens in combination with the premise that evil exists in the world entails that this original being actively perpetrates evil. Given that it not merely fails to prevent evil (the weaker premise motivating the various arguments from evil discussed in Chapter 6), but positively causes it by way of deterministic foreordination, this original cause or organizing principle is, inescapably, ‘the author of sin’ (EHU 8.36)—and, no doubt, the author of natural evil as well. It follows, according to the current argument, that the deity is not morally praiseworthy. The chapter proceeds as follows. In Sections 7.2 and 7.3 I document and examine Hume’s argument from determinism against duties to God. I situate this piece of reasoning in its original polemical context, explain why Hume rather uncharacteristically employs a deontic language of rights and duties rather than his usual vocabulary of virtue (Section 7.2), and examine the logical structure of the argument itself (Section 7.3). In Section 7.4 I defend Hume from the charge that his argument would take away all human responsibility for behavior and leave the original cause of all as the only responsible agent. I then turn (in Section 7.5) to the argument from determinism and evil, documenting Hume’s presentation of this line of reasoning in both the Enquiry and the Dialogues. Hume does not mean to endorse the argument himself, I suggest, but merely uses it as an ad hominem device to bring out the internal tensions in Christian theology over the notion of a ‘meticulous’ predestinarian divine providence. Finally, I relate Hume’s respective treatment of these two arguments from determinism to his overall attitude to natural theology and cosmological speculation (Section 7.6).

the arguments from determinism

183

7.2. ‘Of Suicide’ and Duties to God, Self, and Neighbor The argument from determinism against duties to God appears in ‘Of Suicide’ and ‘Of the Immortality of the Soul’, the two incendiary ‘obnoxious Dissertations’ that Hume originally planned to publish in 1756, then attempted to suppress out of an ‘abundant Prudence’ (a few copies did slip out), and then ﬁnally lobbied from his deathbed to have published posthumously (L i. 444, ii. 253).¹ In ‘Of the Immortality of the Soul’ his reasoning is highly compressed and hence rather cryptic, but it emerges more clearly in ‘Of Suicide’ where it is developed over a few paragraphs.² Hume’s overall goal in ‘Of Suicide’ is to establish the moral innocence of self-killing—to prove that ‘That action may be free from every imputation of guilt or blame’ (E 580). He demarcates the problem space before him in remarks toward the start of the essay: ‘If Suicide be criminal, it must be a transgression of our duty, either to God, our neighbour, or ourselves’ (E 580). So there are three putative bases of the moral prohibition on suicide, each of which Hume attacks in turn. The argument from determinism against duties to God then occurs in the ﬁrst stage of this three-part scheme, under the rubric of Hume’s critique of the claim that we owe a duty to the deity not to commit suicide. ¹ His publisher William Strahan refused, although unauthorized editions appeared in 1777 and again in 1783. For details of the suppression and eventual publication of the essays, see E. C. Mossner, The Life of David Hume, 2nd edn (Oxford: Oxford University Press, 1980), 322–31. ² ‘Of Suicide’ is often regarded as a rather problematic text on account of its ﬂamboyantly rhetorical and somewhat unsystematic character. Since the time of its original publication, certain commentators have gone so far as to suggest that the essay is not so much a work of philosophy as it is an exercise in polemic, propagandist suasion, or mere clericbaiting—successful in its own dramatic terms, perhaps, but scarcely belonging in Hume’s philosophical corpus proper. But this is an overreaction. The essay hardly presents Hume at his most cautious and exact: he does indulge in a certain rhetorical grandstanding and the occasional undeniable hyperbole. But serious lines of argument can be retrieved from the essay without excessive charity or rationalization, as I will at least show in the case of the current argument. (I also make this case with respect to certain of the essay’s other arguments in Thomas Holden, ‘Religion and Moral Prohibition in Hume’s ‘‘Of Suicide’’,’ Hume Studies 31 (2005), 189–210.)

184 the arguments from determinism This ﬁrst sort of objection to suicide asserts that this sort of action is impermissible since it breaches a theological species of duty. The objection is typically expressed with reference to the rights of the creator, and suicide’s status as a violation of those rights. To kill oneself is said to infringe on the sovereignty of the maker: we do not have the right to dispose of our lives, for that prerogative belongs to God alone. The view is captured in the traditional metaphor, revisited in the early modern period by writers such as Montaigne, Pufendorf, Locke, and Clarke, of the passage of our earthly lives as the watch duty of a sentinel, bound to stay at his post through good times and bad.³ Hume’s argument against duties to God is not the only response he offers to the theological objection to suicide, but only with this argument does he provide a generalizable piece of reasoning: an argument that purports to show, not merely that suicide does not violate this ﬁrst species of duty, but that no action does. One ﬁnal preliminary regarding the overall master argument of ‘Of Suicide.’ It is important to appreciate that, in using the language of ‘duties owed to’ such and such a person, and in working through the tripartite scheme of duties owed to God, neighbor, and self, Hume is adopting the terminological framework of his dialectical opponents. This tripartite system runs through Christian moral philosophy ³ Michel de Montaigne, The Complete Essays, tr. and ed. M. A. Screech (Harmonsdworth: Penguin, 1987), 394; Samuel Pufendorf, On the Whole Duty of Man and Citizen, tr. Michael Silverthorne, ed. James Tully (Cambridge: Cambridge University Press, 1991), 47; John Locke, Two Treatises of Government, ed. Peter Laslett (Cambridge: Cambridge University Press, 1988), 271 (book 2, section 6); Samuel Clarke, A Discourse concerning the Unchangeable Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation, in Samuel Clarke, The Works, 4 vols (London, 1738; reprint New York, Garland Publishing Co., 1978), ii. 623. (Hume himself puts the theological challenge to suicide in terms of this metaphor at E 585.) Kant would provide a particularly vivid statement of the view in his lectures from the 1780s (some decades after Hume composed his essay): ‘We have been placed in this world for certain destinies and purposes; but a suicide ﬂouts the intention of his Creator. He arrives in the next world as one who has deserted his post; and must therefore be seen as a rebel against God. . . . [The responsibility to resist suicide] lies upon us until such time as God gives us His express command to depart this world. / Men are stationed here like sentries, and so we must not leave our posts until relieved by the beneﬁcent hand of another’ (Immanuel Kant, Lectures on Ethics, ed. Peter Heath and J. B. Schneewind, tr. Peter Heath (Cambridge: Cambridge University Press, 1997), 27: 367. Here I cite the pagination of the standard edition, Kant’s Gesammelte Schriften, edited by the Royal Prussian (later German) Academy of Sciences (Berlin: Georg Reimer, later Walter deGruyter & Co., 1900–), which is given in the margins of the Cambridge edition.)

the arguments from determinism

185

from Thomas Aquinas to early modern thinkers such as Pufendorf, Malebranche, Locke, Clarke, Hutcheson, and Butler.⁴ Thomas Reid illustrates the point nicely in his Essays on the Active Powers of Man, a work ﬁrst published in 1788, twelve years after Hume’s death. Here Reid contrasts the Christian tripartite scheme and the systematization of virtues provided by Greek and Roman moralists, and indicates his own preference for the three-part division as the more ‘natural’ moral taxonomy: Morals have been methodised in different ways. The ancients commonly arranged them under the four cardinal virtues of Prudence, Temperance, Fortitude, and Justice; Christian writers, I think more properly, under the three heads of the Duty we owe to God—to Ourselves—and to our Neighbour.⁵

But Hume’s own moral theory centers on the concepts of virtue and vice rather than the concept of duty: as Hume himself notes, his account here is closer to the virtue ethics of the ‘antient moralists’ than the scheme of moral laws and duties promulgated by ‘modern philosophers’ (‘or rather divines under that disguise’) (EPM Appendix 4.21; compare also T 3.3.4.4). His own system treats the notion of a duty as a secondary or derivative moral concept, a rather black-andwhite deontic construction from our more ﬁnely textured sentimental responses to the richly variegated world of virtue and vice. An excessive emphasis on rule-based moral categories such as duty and obligation thus has the potential to distort or oversimplify the true ⁴ As Tom L. Beauchamp notes, Aquinas explicitly deploys the threefold taxonomy in maintaining that suicide is morally impermissible since it breaches each of the three kinds of duty. Thomas Aquinas, Summa Theologica, 3 vols (New York: Benziger Brothers, 1947–8), part 2.2, question 64, article 5; see Tom L. Beauchamp, ‘An Analysis of Hume’s Essay ‘‘Of Suicide’’,’ Review of Metaphysics 30 (1976), 73–95: 75. On Pufendorf’s, Malebranche’s, Locke’s, Clarke’s, and Butler’s use of this taxonomy, see J. B. Schneewind, The Invention of Autonomy: A History of Modern Moral Philosophy (Cambridge: Cambridge University Press, 1998), 155–6, 234, 315, 343, 440. See also Francis Hutcheson, Philosophiae Moralis Institutio Compendiaria, ed. Luigi Turco (Indianapolis: Liberty Fund, 2007), book 1, ch. 3, section 5. ⁵ Thomas Reid, Essays on the Active Powers of Man, in Thomas Reid, Philosophical Works, ed. William Hamilton, 2 vols (Edinburgh 1895; facsimile reprint Hildesheim: Georg Olms, 1967), 642 (essay 2, ch. 5, para. 22). For commentary on the competing claims of the ethics of virtue and the ethics of duty, with particular reference to Reid, see Knud Haakonssen, Natural Law and Moral Philosophy: From Grotius to the Scottish Enlightenment (Cambridge: Cambridge University Press, 1996), 197–200; see also Schneewind, The Invention of Autonomy, 287.

186 the arguments from determinism complexity of moral experience (T 3.2.1.2, T 3.3.4.4, EPM Appendix 4.21). However, here in ‘Of Suicide’ Hume enters the debate with the theistic critics of suicide on their own terms, working not only with the concept of duty, but the relational concept of a duty owed by one person to another.⁶ No doubt Hume’s position could be framed in his own preferred terminology, but here it is presented in the language of rights and duties.

7.3. The Argument from Determinism against Duties to God As in the Treatise and ﬁrst Enquiry, in ‘Of Suicide’ Hume maintains that all events, including all human choices and actions, fall out according to an unfolding deterministic system. Just as deterministic laws govern the inanimate world, so too our own human actions and attitudes are ‘impelled or regulated in that course of life, to which they are destined’ by the prior state of the world and the various causal laws (E 580). Here in ‘Of Suicide’, Hume simply asserts that some such version of determinism is true, though he does of course sketch out a case for the ‘doctrine of necessity’ in the Treatise and Enquiry (T 2.3.1, EHU 8.4-22). Given that the deity is responsible for these laws and for the original state of the universe, it follows that everything that happens is a result of general divine providence: all events, including all human actions, fall out according to the causal system ﬁxed by the creator. The providence of the deity appears not immediately in any operation, but governs every thing by those general and immutable laws, which have been established from the beginning of time. All events, in one sense, may be pronounced the action of the almighty. . . . When the passions play, when the judgment dictates, when the limbs obey; this is all the operation of God; and upon these animate principles, as well as upon the inanimate, has he established the government of the universe. (E 581) ⁶ For a useful exposition of the system of relational (or ‘Hohfeldian’) duties owed by one party to another, and the correlative relational system of liberties and claim rights, see L. W. Sumner, The Moral Foundation of Rights (Oxford: Oxford University Press, 1987), 18–53.

the arguments from determinism

187

There is no one event, however important to us, which he has exempted from the general laws that govern the universe, or which he has peculiarly reserved for his own immediate action and operation. (E 581)

This result can then be applied to the particular case of suicide. Acts of suicide (like all human actions) occur as a result of the general deterministic laws governing nature, and so are not disruptions but rather manifestations of those laws. They do not break in on the overall causal system, but are part of it. It therefore makes no sense to say that suicide encroaches on the divinely ﬁxed order, or (adopting the theist’s own language of divine purposes) to speak of suicide as if it frustrates the deity’s plans or disturbs its intentions.⁷ There is no being, which possesses any power or faculty, that it receives not from its creator; nor is there any one, which, by ever so irregular an action can encroach upon the plan of his providence, or disorder the universe. . . . When the horror of pain prevails over the love of life: When a voluntary action anticipates the effects of blind causes; it is only in consequence of those powers and principles, which he has implanted in his creatures. Divine providence is still inviolate, and placed far beyond the reach of human injuries. (E 584)

‘Injuries’ here has a double sense, indicating the ﬁnal move Hume needs in order to make the transition to his conclusion that suicide does not violate a duty to the deity. It is not merely that suicide fails to ‘injure’ divine providence in the sense of materially damaging or disrupting the deity’s unfolding plans. It is further that that action fails to ‘injure’ divine providence in the sense of invading its rights or prerogatives. Since any act of suicide is ﬁxed and predetermined by the ﬁrst cause of all, it makes no sense (Hume is suggesting) to speak of that action as a moral offense against this original being. ⁷ Hume presents an interestingly related argument in his essay ‘Of the Original Contract.’ While ‘Of Suicide’ argues from the premise that God causes everything to happen deterministically to conclude that no event disturbs divine providence, in ‘Of the Original Contract’ Hume argues from this same premise to conclude that, since all events are part of the unfolding deterministic plan, no event can be interpreted to particularly reﬂect divine providence more than any other. This is meant to embarrass the special pleading of divine right theories of political authority: ‘Whatever actually happens is comprehended in the general plan or intention of providence; nor has the greatest and most lawful prince any more reason, upon that account to plead a peculiar or inviolable authority, than an inferior magistrate, or even an usurper, or even a robber and a pyrate’ (E 467).

188 the arguments from determinism ‘What is the meaning, then, of that principle, that a [suicide incurs] the indignation of his creator, by encroaching on the ofﬁce of divine providence, and disturbing the order of the universe?’ (E 582). If the deity foreordained and predetermined a particular act of self-slaughter, it makes no sense to say that that act is any sort of offense against its sovereign prerogatives over our lives. The underlying principle Hume is relying on here, then, is that (the puppet master principle) if X’s behavior is completely orchestrated by some other person Y, then X’s behavior cannot be a moral offense against Y. The claim that Jack’s action wronged Jill simply loses all traction if we discover that Jill was pulling the strings all along. But the deity is always pulling the strings. So Hume’s argument can be summarized as follows: the fact that the original cause of all is responsible for the entire deterministic causal order precludes any possibility that our actions are moral offenses against it: it precludes the possibility that they invade its rights or prerogatives, and it precludes the possibility that we owe this being a duty not to act as we actually do. Of course, this is not yet sufﬁcient to show that suicide is morally permissible. Hume has yet to canvass the other sorts of duties we might have (duties to our neighbors, perhaps, or to ourselves), which might in the end render suicide impermissible. But the current argument does show (if it is successful) that acts of suicide cannot morally offend against the creator. Moreover, since the argument deploys a perfectly general point that would apply to any given action, it will equally show that no human action can offend against the creator: we have no duty to the deity to resist any action. Various commentators have noted that Hume’s argument can be generalized beyond the case of suicide to apply to any arbitrary action. However, since they have taken Hume to be arguing from determinism to the conclusion that suicide is morally permissible (rather than simply not prohibited out of a duty owed to God), they have taken this generalization to amount to a reductio of his reasoning, since equivalent reasoning would then show (absurdly) that murder or any other act is morally permissible.⁸ However, the text simply does not ⁸ Thus Kenneth R. Merrill has suggested that the appeal to determinism is meant to take Hume to the conclusion that humans cannot be morally responsible for acts of suicide, since

the arguments from determinism

189

substantiate this interpretation of Hume’s intended conclusion; nor is it consistent with Hume’s settled philosophical views.⁹ These points aside, the most obvious problem with such a reading is that it neglects the fact that Hume’s arguments against the theological objection to suicide—including his argument from determinism—are merely the ﬁrst plank in a three-part master argument: they purport to show that we have no duty to God to resist suicide, not that there is no duty to resist suicide simpliciter. Once this is understood, we can see that Hume would have no problem in allowing that the argument from determinism is generalizable, for then it will show only that we have no duty to God to resist any action. Hume’s language indicates that he is well aware that the current argument is generalizable beyond the original case of suicide (‘All events, in one sense, may be pronounced the action of the almighty’; ‘There is no being . . . which by ever so irregular an action can encroach upon the plan of his providence, or disorder the universe’). And in ‘Of the Immortality of the Soul’, the second of Hume’s suppressed essays, he presents the same basic argument expanded to its widest scope: As every effect implies a cause, and that another, till we reach the ﬁrst cause of all, which is the Deity; every thing that happens, is ordained by him; and nothing can be the object of his punishment or vengeance. (E 594)

To make this inference intelligible, we need to take ‘punishment’ and ‘vengeance’ as moralized terms: Hume must be asserting that nothing any such acts are foreordained by a causal order beyond their ultimate control (Kenneth R. Merrill, ‘Hume on Suicide,’ History of Philosophy Quarterly 16 (1999), 395–412: 402–3). And Keith E. Yandell and G. R. McLean each suggest that Hume invokes determinism in order to argue that, since any act of suicide is predetermined by the deity, and the deity does no wrong, it follows that whenever an act of suicide occurs it is not wrong (Keith E. Yandell, Hume’s ‘Inexplicable Mystery’: His Views on Religion (Philadelphia: Temple University Press, 1990), 286–95: 290; G. R. McLean, ‘Hume and the Theistic Objection to Suicide’, American Philosophical Quarterly 38 (2001), 99–111: 106). Hume was also read in this latter way by certain of his eighteenth-century reviewers (George Horne, Letters on Inﬁdelity. By the Author of a Letter to Doctor Adam Smith (1784), in James Fieser (ed.), Early Responses to Hume’s Writings on Religion, 2 vols (Bristol: Thoemmes Press, 2001), ii. 367; Johann Georg Heinrich Feder’s review of ‘Of Suicide’ in G¨ottingische Anziegen von gelehrten Sachen, December 31, 1784, No. 210, 2100–03, translated by Curtis Bowman and anthologized in Fieser (ed.), Early Responses to Hume’s Writings on Religion (Bristol: Thoemmes Press, 2001) ii. 324.) ⁹ For details, see Holden, ‘Religion and Moral Prohibition in Hume’s ‘‘Of Suicide’’.’

190 the arguments from determinism can be the object of justiﬁed or legitimate vengeance on the part of the deity. And if we do interpret the inference this way, we then have a second statement of the argument from determinism against duties to God: given that the deity has causally ‘ordained’ all our actions in advance, none of those actions can be a moral offense against it.¹⁰ In doing away with the notion that our actions can be moral offenses against the creator and thereby rejecting the view that we can owe any duties to the deity, Hume’s argument will void the ﬁrst category of the traditional tripartite taxonomy of duties entirely: it leaves us only with duties to neighbor and self. The argument will remove the deity from the community of beings that can enter into moral relations of claim rights, prerogatives, duties, and obligations with one another. Such reasoning will clearly have sweeping consequences for traditional theocentric moral systems. These systems not only typically assert the existence of various duties owed to the ﬁrst cause of all (along with its correlative rights and prerogatives over us); they also commonly regard this as the most important of the three categories of duty. This theological species of duty is often regarded as primary in the sense that ¹⁰ At a stretch, we might also hear an echo of the argument from determinism in Hume’s comment (via Philo) in the Dialogues that, ‘were [the] divine being disposed to be offended at the vices and follies of silly mortals, who are his own workmanship; ill would it surely fare the votaries of most popular superstitions’ (DNR 12.32). In the antecedent clause Hume seems to be hinting at the view that the ﬁrst cause of all could not be justiﬁably aggrieved by our behavior, since it is responsible for creating us with all our various shortcomings. However, this remains merely a hint, and the Dialogues make no connection between this sort of intuition and the doctrine of determinism. Compare also Hume’s account of the moral absurdity of predestinarian religious systems in the ﬁrst volume of the History of England. (This ﬁrst volume was written in 1754, within a few years of the other texts discussed in this chapter.) ‘In tracing the coherence among the system of modern theology we may observe, that the doctrine of absolute decrees has ever been intimately connected with the enthusiastic spirit; as that doctrine forms the highest subject of joy, triumph, and security to the supposed elect, and exalts them, by inﬁnite degrees, above the rest of mankind. All the ﬁrst reformers adopted these principles; and the Jansenists too, a fanatical sect in France, not to mention the Mahometans in Asia, have ever embraced them. As the Lutheran establishments were subjected to episcopal jurisdiction their enthusiastic genius gradually decayed, and men had leisure to perceive the absurdity of supposing God to punish, by inﬁnite torments, what he himself, from all eternity, had unchangeably decreed’ (HE v. 131, my emphasis). One irony here is that in the Enquiry Hume had previously characterized his own deterministic ‘doctrine of necessity,’ in combination with the admission of a ﬁrst cause, as amounting to a version of the doctrine of ‘absolute decrees’ (EHU 8.36). So it is clearly not the doctrine of absolute decrees itself that Hume ﬁnds ‘absurd’ so much as the attendant Christian doctrines of a ﬁnal judgment, a retributive lawgiving god, ‘inﬁnite torments,’ and the rest.

the arguments from determinism

191

duties owed to the creator are supposed to trump any other competing obligations (consider, for instance, the story of Abraham and Isaac). In at least some systems it is also regarded as logically primary, in the sense that duties to neighbor and self turn out to be derivable from the more fundamental array of duties owed to God. For instance, Locke regards the duty not to take one’s neighbor’s or one’s own life simply as a particular speciﬁcation of our more fundamental duty to respect the deity’s property: the prohibition on both murder and suicide derives whatever force it has from this more basic duty owed to the creator.¹¹ And clearly one could go on in charting the destabilizing consequences of Hume’s heretical thesis across the various traditional theocentric moral systems.¹² How persuasive is the argument from determinism against duties to God? If we grant the truth of determinism and allow the existence of an original cause that is responsible for the entire deterministic system, then the argument will appear compelling if we both (i) accept that the antecedent clause of the puppet master principle accurately describes the relationship between that original cause and humans, and (ii) ﬁnd this principle persuasive. On the ﬁrst point, we should note that the antecedent of the puppet master principle will capture the deity–human relationship only if the deity can indeed be classed as a person. If it is merely a blind force, then it is much less clear that the intuitions behind the puppet master principle have any purchase. ¹¹ John Locke, Two Treatises of Civil Government, book 2, section 6. For commentary, see A. J. Simmons, The Lockean Theory of Rights (Princeton: Princeton University Press, 1992), 49–50; see also 30–1, 35–6; George Windstrup, ‘Locke on Suicide,’ Political Theory 8 (1980), 169–82. ¹² For instance, Hume would surely draw some satisfaction from the fact that his argument would dispatch the traditional doctrine, endorsed by Malebranche and others, that since the reprobate have offended against the inﬁnite majesty of the deity, eternal torment in Hell is an entirely proportionate form of punishment. (‘The offense increases in proportion to the majesty offended. It is just to condemn an insolent subject who has insulted his Prince to be a galley-slave for life. Compare and judge whether God, without belying what he is, can be satisﬁed with a transitory vengeance’ (Nicolas Malebranche, quoted in Dom Sinsart, Defense du Dogme Catholique ser l’Eternit´e des Peines (Strasbourg: 1748), 154–5). Pierre Bayle reports this traditional doctrine as follows: ‘that demerit increases in proportion to the dignity of the person offended; from which [orthodox theologians] conclude that sin deserves inﬁnite punishments, since it offends an inﬁnite Being, but that, since punishments cannot be inﬁnite in degree, they must be so in duration’ (Pierre Bayle, Oeuvres Diverses, 5 vols (The Hague, 1727–31) (Facsimile: Hildesheim: G. Olms, 1964), iv. 31). I draw both of these quotations from D. P. Walker’s invaluable study The Decline of Hell: Seventeenth-Century Discussions of Eternal Torment (London: Routledge and Kegan Paul, 1964), 43.)

192 the arguments from determinism On the other hand, if the deity is not some sort of person, then at least according to most early modern moralists, it is not in any case the sort of being to which we could owe duties. The driving conceit of the traditional tripartite framework, at least, is that our moral duties can be exhaustively classiﬁed according to whether they are owed to this or that sort of person: either to a personal god, to one’s neighbor, or to oneself. So here we can present the traditional theist with something of a dilemma: if the deity is a person, then the argument from determinism against duties to God can proceed; if it is not a person, then it is ipso facto doubtful whether we could owe it duties in any case. However, there is still the second question, concerning the plausibility of the puppet master principle itself. The principle states that if X’s behavior is completely orchestrated by some other person Y, then X’s behavior cannot be a moral offense against Y. I take it that the principle has some intuitive plausibility, at least if we read ‘orchestrated’ to involve full conscious knowledge on Y’s part of what they are doing, such that Y foresees that their actions will cause X to act in this way. In such a scenario, the intuitions driving the puppet master principle are (I suggest) at their height: since the puppet master completely foresees everything that he will cause the puppet to do, he surely cannot reasonably complain when the puppet acts accordingly. However, I take it that our intuitions are less clear in the alternative scenario where Y causes X’s behavior without foresight. We could multiply examples and cases, but much of the intuitive force of the puppet master principle is dissipated once we give up the thought that the puppet master is aware (or at least ought to be aware) of the ensuing consequences of its actions. The argument from determinism against duties to God may not then succeed in showing that, given determinism, no ﬁrst cause could be legitimately offended by the behavior of its creatures. But the argument does seem more powerful in the case of a traditional theistic deity, an omniscient agent that foresees all the consequences of its creative acts. Although Hume is not at all explicit about this, perhaps we might then best understand the argument from determinism against duties to God as an ad hominem critique directed at those who accept the existence of an all-knowing personal deity, rather than as a more general argument against the possibility of any sort of original cause to which humans

the arguments from determinism

193

might owe moral duties. If the original cause is a person (and so in that respect capable of exerting moral claims upon us) and yet cannot foresee all that it causes in creating the deterministic order (and so does not obviously forfeit its moral claims on our conduct by way of the puppet master principle), then perhaps we might still owe duties to this sort of fumbling or short-sighted divinity. Such a concession will hardly appease the traditional religious believer, but it does constitute an important exception that limits the scope of the argument from determinism against duties to God, and prevents Hume’s conclusion from applying to any conceivable ﬁrst cause or designer.

7.4. Divine and Human Responsibility So far I have been arguing that in Hume’s view we cannot owe any duties to the deity, at least if this being is understood as the traditional all-knowing personal god. Given that the deity has predetermined all our behavior, none of our actions can invade its rights or prerogatives; nor can this being make moral claims on our conduct. But at the same time, nothing that Hume has said is supposed to tell against the possibility that certain of our actions are moral offenses against other humans. Our behavior may still be virtuous or vicious, and we remain responsible for our own conduct, its predetermination by the ﬁrst cause of all notwithstanding: the argument from determinism against duties to God in no way jeopardizes Hume’s ‘reconciling project with regard to the question of liberty and necessity’ (EHU 8.23). To some this might seem an impossible combination of commitments. If the deity is responsible for all our actions, this may appear to take away any responsibility that humans might have for their own behavior. Indeed, an objector might go further, and suggest that the actual existence of an original cause is really beside the point: whether or not there is a deity, Hume’s picture of a deterministic universe where responsibility would accrue to the ﬁrst cause if there were one already seems to cut away at human responsibility. Barry Stroud puts the objection as follows: If the theological version [of determinism] can seem plausibly to lead to the conclusion that God is the only responsible agent, as Hume admits, then a

194 the arguments from determinism fully secularized version could equally plausibly lead to the conclusion that there are no responsible agents at all. Leaving ‘the original cause’ out of the picture, it is easy to see ‘the doctrine of necessity’, when applied to everything that happens, as implying a description of the world in which ‘there is a continued chain of necessary causes’ extending from time immemorial to ‘every single volition of every single creature’ and thus leaving no alternatives, ‘no indifference, no liberty’ anywhere in the universe. Surely that is the worry of many who see the ‘the doctrine of necessity’ as a threat to man’s liberty—their opposition comes from more than an embarrassment at having to admit that God is the place where the buck stops.¹³

This may indeed be the worry of incompatibilists, who see a general inconsistency between determinism (or ‘the doctrine of necessity’) on the one hand and human liberty and responsibility on the other. But Stroud is wrong to suggest that Hume has given any hostages to this point of view, or that there is any inconsistency in his endorsing the current argument from determinism against duties to God alongside his familiar view that determinism is compatible with human responsibility. For Hume the existence of an original cause that is responsible for all our actions need not preclude human responsibility: pace Stroud, he does not admit that the theological version of determinism makes the deity the only responsible agent. Rather, in Hume’s moral accountancy, it is quite possible for the deity to be responsible for every human action, while individual human beings are simultaneously responsible for many of those self-same actions. According to Hume, an agent is responsible for an action insofar as that action expresses his stable and enduring character. It is a basic fact of human sentimental psychology that if an action stems from an agent’s enduring character, our moral sentiments will respond to that agent as accountable for that action, praising or condemning them accordingly. And our sentimental responses, once corrected with reference to a calm and disinterested point of view, are authoritative insofar as attributions of responsibility go: they constitute the very ¹³ Barry Stroud, Hume (London and New York: Routledge, 1977), 152–3. (Stroud is quoting from EHU 8.32.)

the arguments from determinism

195

framework in which talk of moral responsibility has its place and purchase. So the attributions of responsibility embedded in our natural human sentimental reactions settle the facts about moral responsibility, and, according to Hume, those attributions reveal that an agent is responsible insofar as his behavior causally expresses his enduring character. There are two important consequences. First, Hume has said nothing about the question of whether the agent’s enduring character is itself causally predetermined: our sentimental responses and hence our attributions of responsibility are not sensitive to this fact one way or the other.¹⁴ There is therefore no conﬂict, on Hume’s account, between the doctrine of determinism and the claim that humans are often responsible for their actions: his theory of moral responsibility is quite compatibilist in outlook.¹⁵ Second, given this account of moral responsibility, it is perfectly possible for both the deity and a human agent to be responsible for one and the same action: all that is required is that that action result from the enduring character of a human agent as its immediate cause, while also being traceable to the deity’s enduring character as a remote cause. Don Garrett has made this point in the context of Hume’s discussion of determinism and divine providence in ‘Of Liberty and Necessity’ in the ﬁrst Enquiry.¹⁶ But it applies equally here in the context of ‘Of Suicide,’ ‘Of the Immortality of the Soul,’ and the speciﬁc question of whether we can owe duties to a predetermining god. The result is that Hume can endorse the argument from determinism against duties to God quite consistently with holding individual humans responsible for their actions and behavior. ¹⁴ In the ﬁrst Enquiry (which I address in the next section) Hume is quite explicit that the theory of a deterministic universal providence would not jeopardize the authority of our sentimental responses of approbation and blame. The distinction between ‘vice and virtue’ is, he remarks in this very context, ‘founded in the natural sentiments of the human mind,’ which react to the immediate tendencies of character, and ‘are not to be controuled or altered by any philosophical theory or speculation whatsoever’ (EHU 8.35). ¹⁵ Paul Russell, Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility (New York and Oxford: Oxford University Press, 1995), esp. chs 4, 7, 8, 9; George Botterill, ‘Hume on Liberty and Necessity,’ in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), 277–300. ¹⁶ Don Garrett, Cognition and Commitment in Hume’s Philosophy (New York: Oxford University Press, 1997), 132.

196 the arguments from determinism

7.5. The Argument from Determinism and Evil We have examined Hume’s argument from determinism to the conclusion that humans cannot owe duties to the original cause of all. But there is also another closely related argument in this vicinity, starting out from the same picture of a deterministic general providence and working toward the conclusion that this original being or principle cannot be morally praiseworthy, given that it deterministically foreordains all the evil in the world. This argument from determinism and evil would present a new case for weak moral atheism: a new sort of argument from evil, one that exploits the fact that the deity not merely fails to prevent evil in the universe, but actively causes it by way of a deterministic system. In the previous chapter we saw that Hume holds that the occurrence of any evil is incompatible with the existence of an inﬁnitely powerful and all-good god; but we also saw that he regards this logical problem of evil as a sideshow of little real interest, since the hypothesis of an inﬁnitely powerful original cause is in any case totally unmotivated and perhaps even unintelligible (Section 6.2). The more serious question for Hume then concerned the implications of evil for the moral status of any ﬁnitely powerful deity, and here we saw that he holds that the bare fact that evil exists in the world offers no real evidence one way or the other (Section 6.4). So the question now before us is whether it makes a difference if we add the further premise that the deity deterministically causes this evil. In this section, I explore this suggestion that determinism coupled with the existence of evil provides Hume with a new argument for the view that the deity is not morally praiseworthy. Hume raises the specter of this argument in both the ﬁrst Enquiry and the Dialogues (which is not to say that he ultimately endorses it). In the ﬁrst Enquiry (in section 8, ‘Of Liberty and Necessity’), he approaches the issue in considering the ramiﬁcations of the view that there is a continued chain of necessary causes, pre-ordained and predetermined, reaching from the original cause of all, to every single volition of every human creature. . . . The ultimate Author of all our volitions is the

the arguments from determinism

197

Creator of the world, who ﬁrst bestowed motion on this immense machine, and placed all beings in that particular position, whence every subsequent event, by an inevitable necessity, must result. (EHU 8.32)

This now-familiar picture marries the sort of determinism (or ‘doctrine of necessity’) that Hume has been arguing for in the ﬁrst part of section 8 with the thesis that there is an original cause or deity responsible for the whole ordered system. The result, as Hume notes, is a deterministic version of ‘the doctrine of absolute decrees,’ the predestinarian theory that the deity mandates every last event that actually occurs and thereby establishes what current theologians call a ‘meticulous’ or ‘risk-free’ providence (EHU 8.36; compare also HE v. 131). In the Enquiry, Hume draws a couple of conclusions concerning this combination of determinism with the admission of a ﬁrst cause, but the current point of interest is his response to a possible objection to determinism. The objection is that, given determinism, the original cause would be accountable for all our actions, and thus that insofar as those actions have any turpitude, they must involve our Creator in the same guilt, while he is acknowledged to be their ultimate cause and author. For as a man, who ﬁred a mine, is answerable for all the consequences, whether the train he employed be long or short; so wherever a continued chain of necessary causes is ﬁxed, that Being, either ﬁnite or inﬁnite, who produces the ﬁrst, is likewise the author of all the rest, and must both bear the blame and acquire the praise, which belong to them. Our clear and unalterable ideas of morality establish this rule, upon unquestionable reasons, when we examine the consequences of any human action; and these reasons must still have greater force, when applied to the volitions and intentions of a Being, inﬁnitely wise and powerful. Ignorance or impotence may be pleaded for so limited a creature as man; but those imperfections have no place in our Creator. He foresaw, he ordained, he intended all those actions of men, which we so rashly pronounce criminal. (EHU 8.32)

In short, Hume’s doctrine of necessity would force us to attribute all our vicious behavior to the deity, ‘acknowledg[ing] him to be the ultimate author of guilt and moral turpitude in all his creatures’ (EHU 8.33). In the mouth of the objector, this is supposed to call into question the theological propriety of determinism. But rather than withdrawing his own commitment to that doctrine, Hume’s reply is simply that he is unable to disarm the objection—that ‘[it is

198 the arguments from determinism not] possible to explain distinctly, how the Deity can be the mediate cause of all the actions of men, without being the author of sin’—but that here we have entered into territory that ‘mere and unassisted reason is very unﬁt to handle,’ and thus that ‘philosophy must be sensible of her temerity, when she pries into these sublime mysteries.’ Indeed, ‘to defend absolute decrees, and yet free the Deity from being the author of sin, has been found hitherto to exceed all the power of philosophy,’ and so the deity’s apparent responsibility for all our actions must, perforce, remain one of the sacred mysteries of our religion (EHU 8.36). These pious clich´es are quite obviously disingenuous, however, and there can be little doubt that Hume not only agrees that any original cause must in some sense be considered ‘the author of sin,’ but wants to draw our attention to the fact.¹⁷ In the Dialogues, Philo starts to sketch out a similar argument, but is cut off in midstream by Demea’s anxious interruption: so long as there is any vice at all in the universe, it will very much puzzle you anthropomorphites [i.e. those, like Cleanthes, who hold that the deity has moral attributes somewhat like the human], how to account for it. You must assign a cause for it, without having recourse to the ﬁrst cause. But as every effect must have a cause, and that cause another; you must either carry on the progression in inﬁnitum, or rest on that original principle, who is the ultimate cause of all things . . . Hold! hold! cried Demea: Whither does your imagination hurry you? (DNR 11.16–17)

In one sense this is nothing new. We already know from the original argument from determinism against duties to God that the deity is causally responsible for all the evil that occurs, and that if this original being or principle is morally accountable for everything that it causes, then it is accountable for all the evil in the world. This much Hume clearly grants. But in both the Enquiry and the Dialogues he also seems to be insinuating that there is at least the threat of an additional argument that goes a step further and positively impugns the deity’s moral character. As the critic of determinism puts it in the Enquiry, the danger is that ‘criminal actions render criminal the original cause, if the ¹⁷ For commentary on Hume’s discussion in the Enquiry, see Russell, Freedom and Moral Sentiment, 161–3; Garrett, Cognition and Commitment in Hume’s Philosophy, 131–4.

the arguments from determinism

199

connexion between them be necessary and inevitable’ (EHU 8.32). It is this insinuation, of course, that so alarms Demea in the Dialogues. But does Hume really intend this further conclusion? Does the fact that the deity deterministically foreordains evil have to reﬂect badly on its moral character? Or can this original cause be ‘the author of sin’ without prejudicing its moral status? These might seem like rhetorical questions, but in fact there are good reasons to challenge the claim that a deity that predetermines vicious behavior in humans thereby shows its own moral shortcomings, and good reasons why Hume himself would resist positively endorsing this sort of deterministic case for weak moral atheism.¹⁸ The ﬁrst thing to emphasize here is that Hume is being intentionally provocative. To bring this out, we might consider two superﬁcially peculiar features of his allusions to the threatened argument. First, one might wonder why, in a shift from his usual tactical focus on the simpler problem of natural evil, Hume hints at the current problem solely in the speciﬁc context of moral evil. After all, wouldn’t there be an equivalent problem in the sphere of natural evil, with the deity equally responsible for causally predetermining (not merely permitting) natural evil in the form of famine, disease, earthquakes, and so on? Second, Hume’s repeated use of the phrase ‘the author of sin’ is clearly quite out of character. He does not usually regard the concept of sin a particularly helpful or illuminating one, and his ¹⁸ Some commentators hold that Hume does intend some sort of atheistic conclusion here (see J. C. A. Gaskin, Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, New Jersey: Humanities Press International, 1988), 73. Joseph Priestley also read Hume this way in his The doctrine of philosophical necessity illustrated, appended to Disquisitions relating to matter and spirit (1777); the relevant passage is anthologized in James Fieser (ed.), Early Responses to Hume’s Metaphysical and Epistemological Writings, 2 vols (Bristol: Thoemmes Press, 2005), i. 252–4). And of course it is quite true that Hume holds that an inﬁnitely powerful and all-good god at least would not permit any evil (see Section 6.2) and so a fortiori would not cause it (George Botterill, ‘Hume on Liberty and Necessity,’ 288–9; Peter Millican, ‘The Context, Aims, and Structure of Hume’s First Enquiry,’ in Peter Millican (ed.), Reading Hume on Human Understanding, 27–65: 36). Other commentators hold that Hume does not intend to endorse any implied atheistic conclusion, but rather leaves the threat of such a conclusion simply as an unanswered puzzle for the theist. Thus Buckle suggests that the real lesson Hume intends is skeptical rather than atheistic in character: it is another illustration of the embarrassment and confusion we invite when we permit philosophy to go beyond ‘her true and proper province, the examination of common life’ (EHU 8.36) (Stephen Buckle, Hume’s Enlightenment Tract: The Unity and Purpose of An Enquiry Concerning Human Understanding (Oxford: Clarendon Press, 2001), 229–30).

200 the arguments from determinism standard moral terminology avoids this sort of theologically charged language. What these points show—and the reason they are only superﬁcially peculiar—is that Hume is intentionally needling the pious. More speciﬁcally, he is alluding, in terms his eighteenth-century audience could hardly have missed, to a poisonous quarrel running through early modern theology, the controversy between Arminians with their creed of libertarian or radically free human will on the one hand, and Lutheran, Calvinist, and Jansenist predestinarians with their system of absolute decrees and meticulous divine providence on the other. The historian D. P. Walker distils the essential point from this heated early modern debate in the following terms: ‘From the time that Luther insisted that ‘‘I will harden Pharaoh’s heart’’ means what it says, [predestinarian] theologians have been said by their adversaries to make God the author of sin.’¹⁹ For the eighteenth-century reader, Hume’s use of this phrase would have been an unmistakable reference to the Arminian charge that predestinarian theology is the functional equivalent of devil worship; and Hume’s emphasis on tracing moral evil to the deity also similarly echoes the traditional Arminian way of framing the issue.²⁰ The charge that a predestinarian divinity would be the ‘author of sin’ was quite familiar in Hume’s day, and philosophers such as Hobbes, Leibniz, and Berkeley made use of this language in either attacking or defending the system of meticulous divine providence. We can ﬁnd a useful account of the incendiary power of these words in Pierre Bayle’s discursive skeptical masterpiece, the 1697/1702 Dictionnaire historique et critique, in an article (on the ‘Paulicians’) which was familiar to Hume:²¹ The style of the orthodox does not at all vary in this point: to be a Manichee and to make GOD the author of sin, are two expressions, which always signiﬁed the same thing; and when one Christian sect accuses another of making GOD the author of sin, it never fails to impute Manicheism to it in that respect. . . . ¹⁹ D. P. Walker, The Decline of Hell, 57 (my emphasis). ²⁰ For instance, Milton presses the essential Arminian charge in the following terms: ‘we should feel certain that God has not decreed that everything must happen inevitably. Otherwise we should make him responsible for all the sins ever committed, and should make demons and men blameless’ (John Milton, Prose Works, ed. Don M. Wolfe, 8 vols (New Haven, CT: 1953-82), vi. 164–5). ²¹ See Section 6.1, and Chapter 6, notes 2 and 8.

the arguments from determinism

201

[A] Manichee observing how carefully the several sects of Christianity invent such hypotheses as they think will acquit GOD, without ever owning that they make him the author of sin, will always boldly maintain that this is a more dangerous rock than any other.²²

According to Bayle’s article, his contemporary Pierre Jurieu, the ultraorthodox Calvinist controversialist and a severe critic of Bayle’s own speculative liberties, openly makes the deity the author of sin, permitting Bayle to press a version of the traditional Arminian criticism against his old persecutor in the strongest of terms: Here is the most monstrous doctrine, and the most absurd paradox in Divinity that was ever heard of. . . . All imaginable methods have been tried to explain the manner how God inﬂuences the action of sinners: the hypothesis of absolute Predestination was held, as long as it was thought to do no wrong to the holiness of God; but it has been laid aside, as soon as it was perceived that it struck at that Divine attribute. They who were not sensible that free-will is inconsistent with the physical predetermination, have taught that predetermination all along; but they who believed that it destroyed free-will, have rejected it, and admitted only of a simultaneous and indifferent concourse. They who believe that any concourse is contrary to man’s liberty, suppose that he only is the cause of his own actions. The only reason which moves them to make such a supposition, is because they think that all the decrees whereby providence should be concerned with our will, would make all events necessary, whereby all wicked actions would be no less an effect of God than of his creatures. . . . Lastly, some have gone so far as to maintain, that God cannot forsee the free actions of his creatures. Why so many suppositions? What was the reason of so many steps? It was the desire of clearing GOD; for it was plainly perceived that religion was at stake, and that men would be necessarily led to Atheism, if God was said to be the author of sin. Hence it is that all Christian sects, deny it as a horrible blasphemy, and a detestable impiety, and complain that they are devilishly calumniated.²³ ²² Pierre Bayle, The Dictionary Historical and Critical of Mr. Peter Bayle, tr. Pierre des Maizeaux, 5 vols (London: J. J. and P. Knapton, 1734–38) (Reprint: New York: Garland Publishing, 1984), iv. 522, 524 (‘Paulicians’ article, note I). ²³ Bayle, Dictionary Historical and Critical of Mr. Peter Bayle, iv. 523 (‘Paulicians’ article, note I), my emphasis. See also the article on ‘Synergists,’ note A: the relevant passage is anthologized in Bayle, Political Writings, ed. Sally L. Jenkinson (Cambridge: Cambridge University Press, 2000), 276. On the vexed question of Bayle’s own theological position, see Chapter 6, note 2.

202 the arguments from determinism So in pointing to the specter of a god that is the author of sin, Hume is referring his readers to a highly charged theological controversy, and reminding them that predestinarian systems—of which his own deterministic philosophy is one version—have been widely seen as posing a severe threat to the doctrine of a morally praiseworthy original cause.²⁴ Moreover, he is quite aware of how provocative this is, as we can see in the Dialogues, where Philo’s insinuations on the point prove the last straw for pious Demea, who expresses his alarm and disgust, and promptly ﬁnds an excuse to leave the conversation (DNR 11.17, 11.21). Hume means, then, to shock and scandalize, to cause a ﬂutter in the theological dovecotes. He means to discomﬁt his Christian readers by reminding them of the traditional Arminian challenge to the moral status of a predestinarian god, and he means to stress that his own case for determinism will rule out the libertarian picture of free will and hence the easy Arminian theodicies that prevailed in the liberal, forward-looking circles of mid-century Edinburgh, Glasgow, and London.²⁵ But I do not think that Hume himself actually accepts the argument that ‘criminal actions render criminal the original cause,’ or the view that the moral status of a deity that is responsible for a deterministic universe is compromised by any evil it thereby causes. ²⁴ This may be the reason why the Enquiry drops the claim, made previously in the Treatise, that the doctrine of necessity poses no threat to religion. In the Treatise, Hume had asserted that ‘the doctrine of necessity, according to my explication of it, is not only innocent, but even advantageous to religion and morality’ (T 2.3.2.3). But the analogous passage in the Enquiry merely states that the doctrine of necessity ‘is not only consistent with morality, but . . . absolutely essential to its support’ (EHU 8.26). The claim that this doctrine is of no danger to religion has quite disappeared. ²⁵ Hume’s History of England charts with some amusement what he takes to have be an entirely non-rational transition in the seventeenth-century Anglican Church away from ‘the more rigid principles of absolute reprobation and unconditional decrees’ toward ‘the milder theology of Arminius.’ ‘Some noise was, at ﬁrst, made about these innovations; but being drowned in the fury of factions and civil wars which ensued, the scholastic arguments made an insigniﬁcant ﬁgure amidst those violent disputes about civil and ecclesiastical power, with which the nation was agitated. And at the restoration, the church, though she still retained her old subscriptions and articles of faith, was found to have totally changed her speculative doctrines, and to have embraced tenets more suitable to the genius of her discipline and worship, without it being possible to assign the precise period, in which the alteration was produced’ (HE v. 131–2). Of course, the Calvinist predestinarian scheme still held its own in much of Scotland, including the Chirnside Parish Church that Hume attended as a boy. (See also note 10 above for more from the History’s account of the underlying psychological attitudes driving predestinarian religious systems.)

the arguments from determinism

203

First, Hume never explicitly endorses the thesis that criminal actions render criminal the original cause. These are the words of the critic of determinism, and neither in the Enquiry nor the Dialogues does Hume (or Philo) do more than indicate that such an argument might be raised. Of course, it might be said that his failure to explicitly endorse the threatened argument is yet another instance of his ironic and artfully insinuating manner, that he intends his astute readers to draw a conclusion that is merely hinted at in the text. But this ironic interpretation would need to be properly motivated, and there are reasons to doubt that such motivation is available. For consider, second, that Hume only regards actions as virtuous or vicious insofar as they are expressions of some underlying character trait in the agent: If any action be either virtuous or vicious, ’tis only as a sign of some quality or character. It must depend upon durable principles of the mind, which extend over the whole conduct, and enter into the personal character. Actions themselves, not proceeding from any constant principle, have no inﬂuence on love or hatred, pride or humility; and consequently are never consider’d in morality. (T 3.3.1.4) ’Tis evident that, when we praise any actions, we regard only the motives that produc’d them, and consider the actions as signs or indications of certain principles in the mind and temper. The external performance has no merit. We must look within to ﬁnd the moral quality. This we cannot do directly; and therefore ﬁx our attention on actions, as on external signs. But these actions are still consider’d as signs; and the ultimate object of our praise and approbation is the motive that produc’d them. (T 3.2.1.2)

So actions cannot be assessed in isolation from the underlying character traits that produce them, and there is no judging that such-and-such a particular action is either virtuous or vicious without entering into the motivational psychology of the agent that produced it. The implications for our argument concerning the original cause are evident. One and the same action might be considered either vicious or virtuous, depending on whether we relate it to the character of its immediate cause, a human agent, or to the character of its remote cause, the deity. Supposing that the deity predetermined Caligula’s vicious behavior out of some farsighted benevolent scheme, that behavior would not reﬂect badly upon its remote cause, however much the immediate source of Caligula’s conduct in his own vanity

204 the arguments from determinism or intemperance makes it vicious in relation to the emperor himself.²⁶ In fact Hume provides the crucial part of this reply just two paragraphs before he has the objector complain that determinism will entail that ‘criminal actions render criminal the original cause,’ when he asserts (in addressing a distinct, non-theological point) that ‘actions render a person criminal, merely as they are proofs of criminal principles in the mind; and when, by an alteration of these principles, they cease to be just proofs, they likewise cease to be criminal’ (EHU 8.30). So while we have seen (against Stroud) that a human and the deity can simultaneously bear moral responsibility for one and the same action, it does not follow that if that action reﬂects badly on the one, it must reﬂect badly on the other. Third, we have already seen Hume state (through the character of Philo) that the bare fact that evil exists is quite consistent with the world being produced by a deity of excellent moral character: the fact that the deity fails to prevent this evil, at least, need not tell against its praiseworthiness (DNR 11.2, 11.4; see also my discussion in Section 6.2). So the current deterministic argument will have traction only if Hume accepts some morally relevant distinction between acts and omissions, with the result that a deity that actively causes evil compromises its moral status in a way that a deity that merely fails to prevent evil need not. But this sort of distinction is quite alien to Hume’s thinking. The concepts of virtue and vice stand at the core of his moral theory, and the deﬁning feature that permits us to distinguish a virtuous character trait from a vice is simply the tendency of the trait in question to encourage rather than subvert human wellbeing (EPM 9.1, 9.12). This is an outcomes- or consequences-oriented moral theory, not one that could naturally accommodate a systematic moral distinction between acts on the one hand and omissions on the other.²⁷ ²⁶ Of course, Hume does not himself believe it likely that the original cause of all is in fact acting from any such benevolent scheme (see Sections 5.2–4). But that has no bearing on the current point. ²⁷ This helps to explain why Hume fails to draw any sharp distinction between the deity actively causing evil versus merely failing to prevent it when introducing the proposal that evil might be consistent with a morally impeccable deity of ﬁnite power. The language Cleanthes uses in sketching out this proposal—which Philo immediately accepts as at least an epistemic possibility (DNR 11.2, 11.4)—seems thoroughly ambiguous on the current point: ‘supposing the author of nature to be ﬁnitely perfect . . . a satisfactory account may

the arguments from determinism

205

Taking these various points together, we can see that in his lament that ‘to defend absolute decrees, and yet free the Deity from being the author of sin, has been found hitherto to exceed all the power of philosophy’ (EHU 8.36), Hume is either showing an implausible ignorance of his own philosophical system, or (more likely) is being somewhat disingenuous. True, he does agree that determinism entails that any original cause is the author of sin in the sense that (i) it is causally responsible for producing all vicious behavior in humans, and also in the sense that (ii) it is morally accountable for causing that behavior (on the assumption that this original cause is the sort of being that is morally accountable for what it causes). But given his own moral theory, Hume should not further accept that determinism entails that any original cause is the author of sin in the sense that (iii) in causing vicious behavior in humans, it thereby behaves viciously itself. Further, while Hume’s own moral theory invites these distinctions, the same basic discrimination between a theologically unproblematic sense in which the deity is the author of sin, and a more dangerous sense in which he is not (or at least need not be), can be found in the many of the leading predestinarian controversialists: in Hobbes,²⁸ in Jonathan Edwards,²⁹ and even (a little more obscurely) in Calvin then be given of natural and moral evil. . . . A less evil may then be chosen, in order to avoid a greater: Inconveniences submitted to, in order to reach a desirable end: And in a word, benevolence, regulated by wisdom, and limited by necessity, may produce just such a world as the present’ (DNR 11.1). ²⁸ ‘[T]hough God be the cause of all motion and of all actions, and therefore unless sin be no motion nor action, it must derive a necessity from the ﬁrst mover; nevertheless it cannot be said that God is the author of sin, because not he that necessitateth an action, but he that doth command and warrant it, is the author,’ Thomas Hobbes, The Questions Concerning Liberty, Necessity, and Chance, in The English Works of Thomas Hobbes ed. William Molesworth, 11 vols (London: 1839–45), v. 1–440: 138 (‘Animadversions on the Bishop’s Reply No. 12’). The same distinction occurs in the 1668 Latin version of Hobbes’s Leviathan (at ch. 46, para. 22), solving a theological puzzle that was simply left a matter of God’s incomprehensibility in the 1651 English version (see ch. 46, para. 31). (Both passages are included in Curley’s edition: see Thomas Hobbes, Leviathan, ed. Edwin Curley (Hackett: Indianapolis, 1994), 475–6, 463–4.) ²⁹ ‘If by ‘‘the author of sin,’’ be meant the sinner, the agent, or actor of sin, or the doer of a wicked thing; so it would be a reproach and a blasphemy, to suppose God to be the author of sin. In this sense I utterly deny God to be the author of sin. . . . But if, by ‘‘the author of sin’’, is meant the permitter, or not a hinderer of sin, and, at the same time, a disposer of the state of events, in such a manner, for wise, holy, and most excellent ends and purposes, that sin, if it be permitted, or not hindered, will most certainly and infallibly follow: I say, if this be all that is meant by being the author of sin, I don’t deny that God is

206 the arguments from determinism himself.³⁰ So while it may ‘exceed all the power of philosophy’ to show that the original cause of a deterministic universe is not the remote cause of sin, various philosophies have been developed—including Hume’s own—in which a predetermining god need not be committing the sins it causes: in which it need not be a sinner itself. So again, it seems that Hume is framing his conclusion in a manner calculated to shock and scandalize, and that he is at least as interested in unsettling the traditional Christian responses to the problems of providence and evil as he is in pointing out that the underlying philosophical puzzles here can in fact be neatly resolved in his own moral theory.

7.6. Conclusion In the previous chapter we saw that in Hume’s view the bare fact that evil exists provides no evidential support for either weak or strong moral atheism. Now we have seen that the addition of determinism makes no real difference to this picture. Hume no more endorses the argument from determinism and evil than he does the evidential argument from evil. Even if the deity actively causes all the evils of the world by way of remote deterministic foreordination, still (given Hume’s own philosophical principles) this does not give us a persuasive argument for any form of moral atheism. Hume is certainly quite happy to exploit the toxic controversy over predestinarian theology to bring out the internal tensions facing Christian philosophy, but he does not the author of sin (though I dislike and reject the phrase, as that which by use and custom is apt to carry another sense), it is no reproach for the Most High to be thus the author of sin,’ Jonathan Edwards, The Freedom of the Will (1754), ed. Paul Ramsey, in The Works of Jonathan Edwards, ed. Perry Miller, 23 vols (New Haven: Yale University Press, 1957–), i. 129–439: 399. ³⁰ ‘[T]he wicked man is motivated either by his avarice, or his ambition, or envy, or cruelty to do what he does, and he disregards any other end. Consequently, according to the root which motivates his heart and the end toward which he strives, his work is qualiﬁed and with good reason is judged bad. . . . But God’s intention is completely different. For His aim is to exercise His justice for the salvation and preservation of good, to pour out His goodness and grace on his faithful, and to chastise those who need it. Hence that is how we ought to distinguish between God and men; by separating in the same work His justice, His goodness and His judgment from the evil of both the devil and the ungodly,’ Jean Calvin, Against the Libertines, quoted in Paul Helm, John Calvin’s Ideas (New York: Oxford University Press, 2004), 128 (emphasis mine).

the arguments from determinism

207

himself think that there is a genuine argument for divine amorality from this picture of a deterministic general providence. However, at the same time, we have also seen that Hume does fully endorse the argument from determinism against duties to God. Given the sort of determinism that he himself accepts, the ﬁrst cause of all cannot (in Hume’s view) make moral claims upon us, and nor, correlatively, can we owe it any moral duties. Hume’s differing assessment of our two arguments is of a piece with his overall attitude to cosmological and theological speculation. It conforms to (and thus far helps to conﬁrm) our distinction between the liminal natural theology he permits and the core natural theology he disallows.³¹ The argument from determinism and evil is a clear instance of core natural theology. It purports to take us from facts about the character of the created world to a species-speciﬁc conclusion about the particular intrinsic character of its original cause, offering distinguishing knowledge of the divine attributes that goes beyond the sort of general claims that we might make of any unknown being whatsoever. But Hume rejects all such core natural theology on epistemological grounds, and in the current case (as with the original evidential argument from evil, discussed in Section 6.4), he holds that we are simply not in a position to make an informed guess at the deity’s motives (if indeed it has any) in creating our universe in the way that it did. The argument against duties to God, on the other hand, does conform to Hume’s epistemological strictures. It does not purport to tell us anything about the intrinsic character of the divinity, much less anything species-speciﬁc about that intrinsic character. Rather, it attempts to show that there is a perfectly general conceptual constraint on the way in which any original cause could relate to its causal productions. (The argument simply develops the implications of the deity’s relational property of being the cause or organizing principle responsible for the ordered universe—the deﬁning property by which the reference of the term ‘the deity’ is ﬁxed—in order to show that no such original being or principle could be justiﬁably aggrieved at human behavior, or at least could not be so aggrieved if it ought to ³¹ On the distinction between core and liminal natural theology, see Sections 2.4 and 2.5.

208 the arguments from determinism be able to foresee that behavior.) Finally, the argument against duties to God is really just a particular application of a wider argument that would apply to any given being, for according to Hume no being X could be fully responsible for causing the behavior of some person Y, and yet exert moral claims on Y’s behavior (at least, on the assumption that X ought to be able to foresee the behavior that they will bring about in Y). So the argument falls squarely within the domain of liminal natural theology, the relatively unproblematic form of reasoning about the deity that Hume can and does allow.

Conclusion Hume has two distinct lines of argument for strong moral atheism. One appeals to his account of the natural ambit of our human passions along with his sentimentalist metaphysics of morals in order to conclude that the ﬁrst cause or designer is beyond the projected, responsedependent world of moral properties. According to this argument from sentimentalism our natural feelings of approval and disapproval range only so far as the outer frontier of sense and imagination. The projected properties of virtue and vice are thus conﬁned to the immanent world, and cannot characterize any transcendental order beyond this permanent horizon (see Chapters 3 and 4). Hume’s second argument also turns on facts about sentimental psychology, though this time the emphasis is on the deity’s own sentiments and passions (if indeed it has any). This argument from motivation appeals to Hume’s account of the passions as the engines of action along with a form of probabilistic reasoning about the likely character of any ﬁrst cause or ultimate organizing principle responsible for the ordered universe. According to Hume, even though we know nothing substantive about the distinctive intrinsic character of the original cause of all, we can judge it highly unlikely that this being or principle has anything like our own sentimental psychology. But if the deity lacks some sort of approximately anthropomorphic sentimental psychology, it will not be moved by moral concerns: weak moral atheism follows. Moreover, without an approximately human-like system of passions, the deity’s behavior will be unintelligible in human terms—and for Hume this is sufﬁcient to place this being or principle beyond the sphere of moral assessable beings: strong moral atheism also follows (see Chapter 5). In addition to this two-pronged case for strong moral atheism, Hume also has an argument that cuts the original cause off from our

210 conclusion moral universe in one other way. Given his deterministic ‘doctrine of necessity,’ any ﬁrst cause or ultimate organizing principle will be the remote cause of every human action. It follows, according to Hume’s argument from determinism against duties to God, that this sort of original being or principle cannot make moral claims on human behavior, and nor, correlatively, can we owe it any moral duties. If there is some sort of ﬁrst cause or designer, it has in effect foreordained every human action, and so cannot reasonably be aggrieved at or offended by those actions. It cannot then enter into the sort of moralized relations that presuppose the possibility of legitimate grievance or moral offense (see Chapter 7). None of these arguments advances an account of the deity’s distinctive intrinsic character. Hume’s epistemology prohibits all such species-speciﬁc conclusions about the deity’s intrinsic nature, and his case for divine amorality is restricted to the sort of liminal natural theology that avoids such overreaching speculation (see Chapter 2). In particular, Hume does not endorse either the evidential argument from evil or the argument from determinism and evil, two well-known arguments for weak moral atheism that have often been attributed to him. Each of these arguments purports to take us from facts about empirical phenomena to a conclusion afﬁrming the moral indifference or malevolence of the original cause of all. But in Hume’s view this sort of inference from local observations to a conclusion about the distinctive intrinsic character of the ﬁrst cause totally exceeds the evidence: it is no better, in fact, than the traditional natural theologian’s attempt to prove that the deity is morally laudable from facts about the beauty and utility of the created world. Whether supportive of traditional religious views or subversive, each of these arguments fails to take seriously both the narrow limits of human understanding and the hopeless inadequacy of our empirical evidence (see Chapters 6 and 7). This then is Hume’s assessment of the logic of the situation: we know nothing of the deity’s distinctive intrinsic character, but we do know enough to place any such original cause or designer beyond the world of morality. But in addition to providing this case for moral atheism, Hume also offers a genealogical explanation or diagnosis of the human tendency to embrace moral theism despite the lack of any real evidence for the view. Humans have a deeply ingrained propensity to

conclusion

211

anthropomorphize the beings around them, a pre-reﬂective disposition to interpret the nonhuman world, particularly where it impinges on our own fates and fortunes, through a matrix of humanly intelligible passions, desires, and aversions. Unless checked by philosophical reﬂection, we project human attitudes onto the forces of nature and accordingly seek to curry favor with the spirits and animistic forces that impact our lives, understanding their aims and interests through rough analogy with our own. This piecemeal idolatry and polytheism, ‘the ﬁrst and most ancient religion of mankind’ (NHR 1.1), is then subsequently inﬂated through ﬂattery and panegyric, and eventually codiﬁed into a monotheistic theology of ultimate perfection associated with the original cause or creative principle responsible for the entire universe. We create a god in our own image, and then shout it up through exaggerated praise and the vanity of projective self-identiﬁcation. (‘While we look up to Heaven we confound / Knowledge with knowledge,’ John Webster’s character Flamineo laments in The White Devil (1612)—and on this point is a perfect Humean avant la lettre.) This anthropomorphic vision of the deity is then embellished still further as we dignify our own fears and desires with an ultimate cosmological signiﬁcance and tell ourselves that our passions are focused on an angry or a loving god, painting forth an original cause that is a rich and detailed reﬂection of our own misinterpreted emotional attitudes, while the real objects of our feelings are in fact more human, parochial, and mundane. For Hume, the notion of a praiseworthy personal god thus has its roots in some of our most characteristic human frailties: in our nervous regard for the vagaries of fate, in our absurd self-importance, and in the opacity of our own emotional states to ourselves. (On our tendency to project human attitudes onto the ﬁrst cause of all, see Section 3.4 and Section 5.4, and on the movement from an original polytheism to a monotheistic theology of inﬁnite perfection, Section 6.2.) I conclude with some remarks about the historical and philosophical signiﬁcance of Hume’s moral atheism. It might be asked: what difference does it make to our wider understanding of Hume’s philosophy if he is in fact a moral atheist? After all, no one could have taken this notoriously irreligious philosopher and relentless critic

212 conclusion of cosmological anthropomorphism for a moral theist.¹ The standard, received view is rather that he is a skeptic regarding all speculation about the divine attributes, and hence a moral agnostic, a skeptic regarding the question of the deity’s moral character. And such moral agnosticism might already seem to render belief in a deity quite irrelevant to practical questions of moral conduct. So what difference does it make to our overall understanding of Hume if we read him as positively denying the existence of a god with moral attributes, rather than simply remaining skeptical or noncommittal on the point? Does moral atheism add anything of real philosophical signiﬁcance or practical force to moral agnosticism? Put the question of moral practice to one side for the moment and consider ﬁrst the ramiﬁcations for our understanding of Hume’s theoretical philosophy. Hume is one of the most powerful and trenchant critics of religion, and if his critique extends beyond a strict moral agnosticism to a positive afﬁrmation of divine amorality this ought to be of some interest. Moral atheism is fundamentally at odds not only with Christianity and other traditional theistic systems, but also with the optimistic deistic philosophies that dominated ‘freethinking’ thought in Hume’s day (the systems, say, of Shaftesbury, Tindal, and Voltaire). It is also at odds with the sort of blanket theological skepticism that is often credited to Hume himself. If the Great Inﬁdel went so far, it would be an important moment in the history of irreligion. True, Hume would not be the ﬁrst early modern philosopher to endorse moral atheism. Here he is plainly anticipated by Spinoza, whose pantheistic philosophy presents us with a sublimely indifferent and totally impersonal ‘God-or-Nature’ quite beyond moral assessment. More controversially, perhaps we can also ¹ Never say never. I am neglecting those rare accounts that take Hume’s protestations of Christian piety at irony-free face value (see Section 1.4), or take the moral theist Cleanthes to speak for Hume on the current point. Dugald Stewart is perhaps the most forthright advocate of this latter reading, insisting that Cleanthes ‘is the hero of the Dialogue, and is to be considered as speaking Mr. Hume’s real opinions. . . . [T]he reasonings of Philo have often been quoted as parts of Hume’s philosophical system, although the words of Shylock or Caliban might, with equal justice, be quoted as speaking the real sentiments of Shakespeare’ (Dugald Stewart, Collected Works, ed. Sir William Hamilton, 11 vols (Edinburgh: T. Constable, 1854–60), i. 605). As I said in my preface, this interpretation hardly warrants direct refutation. But many of the interpretive arguments in essay provide additional cumulative evidence against it.

conclusion

213

see a form of moral atheism lurking unacknowledged in Luther, in Calvin, and in Hobbes, given that each of these thinkers holds that moral standards are created by the positive decrees of an untrammeled divine will,² and thus appears (intentionally or unintentionally) to place the deity beyond the possibility of moral assessment.³ However, Hume’s arguments for moral atheism show an epistemological caution and metaphysical restraint that is quite lacking in these earlier thinkers: here his case is quite unlike Spinoza’s grandiose rationalist vision of a necessarily imperturbable world-substance, or Luther’s, Calvin’s, and Hobbes’s revelations of an ultimate legislator whose arbitrary will is beyond all normative constraint. Hume is surely a more judicious critic of natural theology than any of these earlier thinkers, and his own cautiously developed brand of moral atheism—which simply draws out the implications of his sentimentalism and moral psychology for the ² Martin Luther, Martin Luther: Selections from His Writings, ed. John Dillenberger (Garden City, NY: Anchor, 1961), 196; John Calvin, Institutes of the Christian Religion, ed. John T.McNeil, tr. Ford Lewis Battles, 2 vols (London: 1941), 3.23.2. For this (controversial) interpretation of Hobbes as a voluntarist or divine command theorist about the moral law, see A. P. Martinich, The Two Gods of Leviathan: Thomas Hobbes on Religion and Politics (Cambridge: Cambridge University Press, 1992), chs 3 and 4. ³ Such at least was the objection of moral rationalists such as Cudworth, Clarke, and Leibniz: the voluntarist doctrine of a deity that creates all moral distinctions through acts of pure will and is not beholden to any higher standards would entail (disastrously) a form of strong moral atheism. The inference from voluntarism to an unconscionable moral atheism was widely accepted in Hume’s day, and a consequent ‘fear of voluntarism’ had taken hold among British philosophers and theologians by the 1730s (J. B. Schneewind, ‘Hume on the Religious Signiﬁcance of Moral Rationalism,’ Hume Studies 26 (2000), 211–23: 213). If we bear in mind this widespread critique and condemnation of voluntarism in the period, it is clear that there is an irreligious undercurrent to the following famous passage in which Hume explicitly notes the parallel between his own sentimentalist moral theory and the tradition of voluntarism: ‘[T]he distinct boundaries of reason and 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. . . . The standard of the one, being founded on the nature of things, is eternal and inﬂexible, even by the will of the Supreme Being: The standard of the other, arising from the internal frame and constitution of animals, is ultimately derived from that Supreme Will, which bestowed on each being its peculiar nature, and arranged the several classes and orders of existence’ (EPM Appendix 1.21). Behind the mock-piety of the voluntarist language, Hume is intimating that moral atheism is a consequence of his sentimentalist moral philosophy. On the background of early modern voluntarism and rationalist concerns about the voluntarist theory, see Frederick C. Beiser, The Sovereignty of Reason: The Defense of Rationality in the Early English Enlightenment (Princeton, NJ: Princeton University Press, 1996), 41–4; J. B. Schneewind, The Invention of Autonomy: A History of Modern Moral Philosophy (Cambridge: Cambridge University Press, 1998), 25–36.

214 conclusion moral standing of any ultimate cause or organizing principle, without claiming any understanding of its distinctive intrinsic character—is, I think, the more attractive. This brings me to my second point. Hume’s case for moral atheism is an important part of his philosophy of religion in its own right, but we might also see it as casting light on his general methodology in this area. It shows how Hume can make good on his assurance, offered right at the start of the Treatise, that the ‘science of man’ particularly promises improvements . . . in natural religion, as it [i.e. natural religion] is not content with instructing us in the nature of superior powers, but carries its views farther, to their disposition towards us, and our duties towards them; and consequently we ourselves are not only the beings, that reason, but are also one of the objects, concerning which we reason. (T Introduction 4)

Hume’s account of human nature—comprehending the natural objects of our passions, the basis of our moral claims in sentimental responses, and the role of our affective psychology in producing behavior—does turn out to support conclusions about the moral standing of the ﬁrst cause of all. This is not because it discloses anything about the distinctive intrinsic character of this ultimate being or principle, but because it tells us something about the way in which our own passions would relate to any original cause, and because it tells us what sort of (highly speciﬁc and hence a priori improbable) sentimental psychology is necessary to produce morally intelligible behavior. We might think of this as a small-scale Copernican revolution in natural theology. At least, we might be impressed by Hume’s ingenuity here, and by the unexpected methodological detour his natural theology takes through the theory of human nature. A ﬁnal point concerns the wider implications for our understanding of Hume’s epistemology—in particular, the much debated question of the precise scope and force of his skepticism. We have seen that Hume’s case for moral atheism is driven by doctrines developed within his experimentalist science of human nature, and that as a result his arguments here are able to keep within the disciplined caution imposed by his general epistemology. Perhaps Hume’s arguments for moral atheism are not always consistent with the letter of his more tubthumping statements of mitigated skepticism, for there is a clear sense

conclusion

215

in which they do trade in ‘distant and high enquiries’ beyond the sphere of ‘common life.’ But they are at least consistent with the general epistemological principles underpinning that skepticism, including the more fundamental injunction that we ‘[limit] our enquiries to such subjects as are best adapted to the narrow capacity of human understanding’ (EHU 12.25). So Hume can hazard certain nonskeptical views about the properties of any ﬁrst cause of all consistent with his deepest epistemological commitments and, I suggest, with the underlying spirit and motivation of his mitigated skepticism (see Section 2.5). Hume’s case for moral atheism thus illustrates the background epistemological theme of this book: that he can and does advance a (negative and irreligious) liminal natural theology of his own, notwithstanding his own skepticism regarding traditional (‘core’) natural theology. This should inform our understanding of the precise boundary between the ‘airy sciences’ and ‘false and adulterate’ metaphysics that Hume totally rejects on epistemological grounds, and the ‘true metaphysics’ that ‘we must cultivate . . .with some care’ (EHU 1.12). At least, it should serve as a corrective to the common view that Hume is a strict and total skeptic regarding literally all speculation about the ﬁrst cause of all. When we turn to the ramiﬁcations of Hume’s moral atheism for practice and conduct, there is (it must be said) less of a sharp contrast with moral agnosticism. Both positions reject any suggestion that we might be able to show that the deity has moral attributes—on the one view, because we already know that the ﬁrst cause cannot have moral attributes, and on the other, because we cannot know that it has. On neither view can religious metaphysics present us with an original cause that plays the role of a morally authoritative lawgiver or an advising voice of conscience; on neither view can it validate expectations of a ﬁnal justice and an afterlife of due punishment or reward; on neither view can it underwrite the moralized conception of religious practice as an attempt to ‘imitate and resemble’ a perfect God.⁴ So either moral atheism or moral agnosticism would ⁴ As the inﬂuential Cambridge Platonist Benjamin Whichcote (1609–83) characterizes the ‘business of religion’ (The works of the learned Benjamin Whichcote (London: J. Chalmers, 1751), i. 32, see also i. 54); and recall also John Tillotson’s view that ‘religion . . . consists in the imitation of [God]’ (Tillotson, Sermons on several subjects and occasions, 12 vols (London:

216 conclusion effectively neutralize any inﬂuence from theological speculation on morality and politics, shutting it off from all questions about how to live.⁵, ⁶ Is there then no practical difference at all between moral atheism and moral agnosticism? That would be a little too quick. Since moral atheism rules out the existence of a ﬁrst cause or designer with moral attributes (and not simply our knowledge of such a being), it will also tell against the ﬁdeistic proposal that our conduct be informed by a commitment of faith in the absence of knowledge. There have of course been religious thinkers who positively welcome Hume’s attack on the claims of natural theologians to be able to vindicate traditional 1742–4), vii. 2280. In the Christian tradition this optimistic conception of religious practice takes its lead from the Sermon on the Mount and Matthew 5:48: ‘Be ye therefore perfect, even as your Father which is in heaven is perfect.’ It has of course been ﬁercely contested by theologians inclined to the darker anthropology of Augustine and Calvin, since the imitation of divine perfection can seem to be out of reach for fallen human beings corrupted by original sin. For discussion, see John Passmore, The Perfectibility of Man, 2nd edn (London: Duckworth, 1971), chs 4 and 5; Michael Gill, The British Moralists on Human Nature and the Birth of Secular Ethics (Cambridge: Cambridge University Press, 2006), 18–20, 60–4. ⁵ Recall that Cleanthes reported the threat to traditional theism in terms of an agnosticism regarding the deity’s moral attributes, not a total denial: ‘For to what purpose establish the natural attributes of the Deity, while the moral are still doubtful and uncertain?’ (DNR 10.27, my emphases). Compare also Tillotson’s and Berkeley’s remarks quoted in Section 1.1, which similarly frame the challenge in terms of a skepticism regarding the moral attributes, not outright moral atheism. ⁶ The practical impotence of natural theology is an important result for Hume. It is stressed in DNR 12.33, and provides the main theme of section 11 of the ﬁrst Enquiry, which was provocatively titled ‘Of the Practical Consequences of Natural Religion’ in the ﬁrst edition of 1748 (though subsequently renamed, rather blandly, ‘Of a Particular Providence and a Future State’). In this dialogue, the suspiciously Humean ‘friend who loves sceptical paradoxes’ (EHU 11.1) summarizes the view in his closing peroration: ‘All the philosophy . . . in the world, and all the religion, which is nothing but a species of philosophy, will never be able to carry us beyond the usual course of experience, or give us measures of conduct and behavior different from those which are furnished by the reﬂections of common life’ (EHU 11.27). For a wide-ranging discussion of Hume’s emphasis on the practical impotence of natural religion, see Simon Blackburn, ‘Playing Hume’s Hand,’ in D. Z. Phillips and Timothy Tessin (ed.), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 3–16. For useful commentary on this theme in the particular context of Enquiry section 11, see J. C. A. Gaskin, ‘Religion: The Useless Hypothesis,’ in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 349–69, and Stephen Buckle, Hume’s Enlightenment Tract: The Unity and Purpose of An Enquiry Concerning Human Understanding (Oxford: Clarendon Press, 2001), 275–89.

conclusion

217

religious doctrines such as moral theism: the view is typically that (as in Kant’s famous formulation) we must tear down such knowledge to make way for faith. Consider two rather unlikely German admirers of Hume, the pietistic Romantic Johann Georg Hamann (1730–88) and the philosopher and religious controversialist Friedrich Heinrich Jacobi (1743–1819). Hamann—the Christian religious thinker, early ideologue of Sturm und Drang, and so-called ‘Magus of the North’—was an enthusiastic cheerleader for Hume’s skeptical attack on Enlightenment optimism about the powers of reason.⁷ (In all likelihood it was his translation of Treatise book 1, part 4, section7—Hume’s famous odyssey through philosophical melancholy and delirium to the equivocal comforts of natural belief—that ﬁrst introduced Hamann’s friend and fellow K¨onigsberger, Immanuel Kant, to the rousing effects of Hume’s philosophy.⁸) For Hamann, Hume’s critique of rational theology along with his emphasis on the animal instinct that underpins all our knowledge makes this ‘Attic Philosopher’⁹ an unintentional witness to the true power of Christian belief, an accidental ally of the man of faith.¹⁰ ‘Hume is always my man,’ Hamann writes; for all his own obvious unbelief, the great Scottish skeptic is nevertheless a ‘Saul among the Prophets.’¹¹ The writings of Jacobi present a similarly paradoxical combination of pietistic enthusiasm with reverence for Hume. In his David Hume on Faith, or Idealism and Realism (1787), a classic of ⁷ Isaiah Berlin, The Magus of the North: J. G. Hamann and the Origins of Modern Irrationalism (New York: Farrar, Strauss, and Giroux, 1993); Frederick C. Beiser, The Fate of Reason: German Philosophy from Kant to Fichte (Cambridge, MA: Harvard University Press, 1987), 16–43; M. Redmond, ‘The Hamann–Hume Connection,’ Religious Studies 23 (1987), 95–107. ⁸ Hamann’s translation was published in a K¨onigsberg newspaper in 1771. Hamann’s partial translation of Hume’s Dialogues in 1780 also seems to have provided Kant’s ﬁrst exposure to this later text (Paul Guyer, Knowledge, Reason, and Faith: Kant’s Response to Hume (Princeton University Press: Princeton, 2008), 6). ⁹ Quoted in Beiser, The Fate of Reason, 24. ¹⁰ ‘Know ye, philosophers, that between cause and effect, means and ends, the connection is not physical but spiritual, ideal; that is the nexus of blind faith.’ By his own admission Hume needs faith ‘to eat an egg, to drink a glass of water’; he can thus scarcely deny it to the committed Christian: since faith ‘is needed for eating or drinking, why does Hume break his own principle when judging of things higher than eating or drinking?’ (quoted in Berlin, The Magus of the North, 32, 33). ¹¹ Quoted in Berlin, The Magus of the North, 33.

218 conclusion early anti-Kantian philosophy, Jacobi expressly appeals both to Hume’s defense of instinctive belief and to his critique of the overreaching claims of metaphysics in order to set out his own case for the primacy of religious faith over the vanities of speculative reason.¹² So there are major religious thinkers who take Hume’s attack on rational religion to leave us with the possibility of trusting in a moral God through a leap of faith—or even, in sweeping away the fool’s gold of rational theology, to positively clear the way for such a ﬁdeistic commitment. The man of faith might then co-opt the skeptical side of Hume’s philosophy of religion and press it into service in the sort of skepticalﬁdeistic program endorsed (whether sincerely or otherwise) by the likes of Montaigne and Bayle.¹³ Or, without going so far as to enjoin full-blown ﬁdeistic belief , perhaps one might take the Kantian position that Hume’s skeptical attack on rational religion at least leaves us free to hope that there is a moral God without positively believing it so. Whether Hume would have been more amused at or alarmed by the selective appropriation of his philosophy found in CounterEnlightenment ﬁdeists such as Hamann and Jacobi, we can now see that they are not doing justice to the full subtleties of his system. Hume is not simply a pure skeptic regarding all natural theology, and while ﬁdeists may glean what they will from his skeptical critique of core natural theology, they should appreciate that Hume’s own philosophical system positively rules out a morally approvable God, and thus would render any ﬁdeistic belief in (or even hope for) such a deity quite impossible. So Hume’s moral atheism has at least this prophylactic consequence for practice and conduct over and above moral agnosticism: it closes off not just the possibility that ¹² Friedrich Heinrich Jacobi, David Hume u¨ ber den Glauben, oder Idealismus und Realismus, ein Gerspr¨ach (1787). For commentary, see Beiser, The Fate of Reason, 89–91; George di Giovanni, ‘Hume, Jacobi, and Common Sense: An Episode in the Reception of Hume in Germany at the Time of Kant,’ Kant-Studien, 89 (1998), 44–58. ¹³ Hume himself suggests this ﬁdeistic path at the close of the Dialogues, though no one aware of his wider irreligious commitments could think him sincere: ‘A person, seasoned with a just sense of the imperfections of natural reason, will ﬂy to revealed truth with the greatest avidity. . . . To be a philosophical sceptic is, in a man of letters, the ﬁrst and most essential step towards being a sound, believing Christian’ (DNR 12.33). On Montaigne’s use of skepticism in the service of ﬁdeism and the question of his sincerity, see Richard H. Popkin, The History of Scepticism: From Savonarola to Bayle (Oxford: Oxford University Press, 2003), 44–63; Fr´ed´eric Brahami, Le scepticisme de Montaigne (Paris: Presses Universitaires de France, 1998). For the literature on Bayle, see Chapter 6, note 2.

conclusion

219

rational theology has genuine implications for our behavior, but also the possibility that an arational faith or religious hope could shape our moral lives without falling into open irrationality. In positively ruling out a deity with moral attributes, Hume disallows practical implications from any religion, rational or arational, centered on the ﬁrst cause or organizing principle responsible for the ordered universe.

This page intentionally left blank

Bibliography Abbruzzese, John, ‘Garrett on the Theological Objection to Hume’s Compatibilism’, British Journal for the History of Philosophy 8 (2000), 345–52. Alanen, Lilli, ‘The Powers and Mechanisms of the Passions’, in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), 179–98. Allen, D. C., Doubt’s Boundless Sea: Skepticism and Faith in the Renaissance (Baltimore: Johns Hopkins Press, 1964). Alston, William P., ‘The Inductive Argument from Evil and the Human Cognitive Condition’, in Daniel Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington and Indianapolis: Indiana University Press, 1996), 97–125. Anscombe, G. E. M., ‘Modern Moral Philosophy’, Philosophy 33 (1958), 1–19. Appleby, Peter C., ‘On Religious Attitudes’, Religious Studies 6 (1970), 359–68. Aquinas, Thomas, Summa Contra Gentiles, 4 vols (New York: Benziger, 1924). Summa Theologica, 3 vols (New York: Benziger, 1947–8). ´ Ardal, P´all, Passion and Value in Hume’s Treatise (Edinburgh: Edinburgh University Press, 1966). Aristotle, The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941). Arnauld, Antoine, and Pierre Nicole, Logic, or the Art of Thinking, tr. and ed. Jill Vance Buroker (Cambridge: Cambridge University Press, 1996). Bacon, Francis, The Essayes or Counsels, Civill and Moral, ed. Michael Kiernan (Oxford: Clarendon Press, 1985). Badía Cabrera, Miguel A., Hume’s Reﬂections on Religion (Dordrecht: Kluwer, 2001). Baier, Annette C., A Progress of Sentiments: Reﬂections on Hume’s Treatise (Cambridge, MA: Harvard University Press, 1991).

222 bibliography Baier, Annette C., ‘Hume on Religion: Stopping the Ocean with a BullRush’, in J. J. MacIntosh and H. A. Meynell (eds.), Faith, Scepticism and Identity: A Festschrift for Terence Penelhum (Calgary: University of Calgary Press, 1994). Balguy, John, The Foundation of Moral Goodness (1728–9), 2 vols (Facsimile: New York: Garland, 1976). Battersby, Christine, ‘The Dialogues as Original Imitation: Cicero and the Nature of Hume’s Skepticism’, in David Fate Norton, Nicholas Capaldi, and Wade L. Robison (eds), McGill Hume Studies (San Diego: Austin Hill Press, Inc., 1979), 239–52. Bayle, Pierre, Oeuvres Diverses, 5 vols (The Hague: 1727–31) (Facsimile: Hildesheim: G. Olms, 1964). The Dictionary Historical and Critical of Mr. Peter Bayle, tr. Pierre des Maizeaux, 5 vols (London: J. J. and P. Knapton, 1734–38) (Reprint: New York: Garland Publishing, 1984). Historical and Critical Dictionary, tr. and ed. Richard H. Popkin (Indianapolis: Hackett, 1991). Political Writings, ed. Sally L. Jenkinson (Cambridge: Cambridge University Press, 2000). Beam, Craig, ‘Hume and Nietzsche: Naturalists, Ethicists, AntiChristians’, Hume Studies 22 (1996), 299–324. Beattie, James, An Essay on the Nature and Immutability of Truth, in opposition to sophistry and scepticism (1770) (facsimile reprint: StuttgartBad Cannstatt: Verlag, 1974). Beauchamp, Tom L., ‘An Analysis of Hume’s Essay ‘‘On Suicide’’ ’, Review of Metaphysics 30 (1976), 73–95. ‘Introduction’ to David Hume, An Enquiry Concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000), xi–civ. ‘Editor’s Annotations’ to David Hume, An Enquiry Concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000), 124–200. Beiser, Frederick, The Fate of Reason: German Philosophy from Kant to Fichte (Cambridge, MA: Harvard University Press, 1987). The Sovereignty of Reason: The Defense of Rationality in the Early English Enlightenment (Princeton: Princeton University Press, 1996).

bibliography

223

Bell, Martin, ‘Hume on Superstition,’ in D. Z. Phillips and Timothy Tessin (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 153–70. Bentley, Richard, The Folly and Unreasonableness of Atheism (London: 1692). Berkeley, George, The Theory of Vision Vindicated and Explained (London: J. Tonson, 1733). The Works of George Berkeley, Bishop of Cloyne, ed. A. A. Luce and T. E. Jessop, 9 vols. (London: Nelson and Sons, 1948–57). Alciphron, in David Berman (ed.), Alciphron, or the Minute Philosopher in Focus (London: Routledge, 1993), 17–161. A Treatise of the Principles of Human Knowledge, ed. Jonathan Dancy (Oxford: Oxford University Press, 1998). Berlin, Isaiah, The Magus of the North: J. G. Hamann and the Origins of Modern Irrationalism (New York: Farrar, Strauss, and Giroux, 1993). Berman, David, ‘Enlightenment and Counter-Enlightenment in Irish Philosophy’, Archiv für Geschichte der Philosophie 64 (1982), 148–65. ‘David Hume and the Suppression of ‘‘Atheism’’ ’, Journal of the History of Philosophy 21 (1983), 375–87. ‘Deism, Immortality, and the Art of Theological Lying’, in J. A. Leo Lemay (ed.), Deism, Masonry, and the Enlightenment (London and Toronto: Associated University Presses, 1987), 61–78. A History of Atheism: From Hobbes to Russell (London: Routledge, 1988). Alciphron, or the Minute Philosopher in Focus (London: Routledge, 1993). Bernard, Christopher, ‘Hume and the Madness of Religion’, in M. A. Stewart and John P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), 224–38. Blackburn, Simon, ‘Hume on the Mezzanine Level’, Hume Studies 19 (1993), 273–88. Ruling Passions (Oxford: Oxford University Press, 1998). ‘Playing Hume’s Hand’, in D. Z. Phillips and Timothy Tessin (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 3–16. Blackmore, Richard, Creation: A Philosophical Poem. Demonstrating the Existence and Providence of God, 2nd edn (London, 1712). Borges, Jorge Luis, Collected Fictions, tr. and ed. Andrew Hurley (New York: Viking Penguin, 1998).

224 bibliography Botterill, George, ‘Hume on Liberty and Necessity’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), 277–300. Brahami, Fr´ed´eric, Le scepticisme de Montaigne (Paris: Presses Universitaires de France, 1998). Broad, C. D., Five Types of Ethical Theory (London: Routledge, 1930). Buckle, Stephen, Hume’s Enlightenment Tract: The Unity and Purpose of An Enquiry concerning Human Understanding (Oxford: Clarendon Press, 2001). Buckley, George T., Atheism in the English Renaissance (New York: Russell and Russell, 1965). Buckley, Michael J., At the Origins of Modern Atheism (New Haven: Yale University Press, 1987). Butler, Joseph, Fifteen Sermons preached at the Rolls Chapel, 4th edn (London, 1749). The Analogy of Religion Natural and Revealed to the Constitution and Course of Nature, in The Works of Bishop Butler, ed. J. H. Bernard, 2 vols (London: Macmillan and Co., 1900). The Works of Bishop Butler, ed. J. H. Bernard, 2 vols (London: Macmillan and Co., 1900). Fifteen Sermons, ed. W. R. Matthews (London and Edinburgh: Morrison and Gibb, 1914). Butler, R. J., ‘Natural Belief and the Enigma of Hume’, Archiv für Geschichte der Philosophie 42 (1960), 73–100. Byrne, Peter, Religion and the Enlightenment: from Descartes to Kant (Louisville, KY: Westminster John Knox Press, 1997). Calvin, John, Institutes of the Christian Religion, ed. John T.McNeil, tr. Ford Lewis Battles, 2 vols (London, 1941). Capaldi, Nicholas, David Hume: The Newtonian Philosopher (Boston: G.K. Hall and Co., 1975). Capaldi, Nicholas, Norton, David Fate, and Robison, Wade L. (eds), McGill Hume Studies (San Diego: Austin Hill Press, 1976). Capitan, William H., ‘Part X of Hume’s Dialogues’, American Philosophical Quarterly 3 (1966), 82–5. Cicero, Marcus Tullius, The Nature of the Gods, tr. Horace C. P. McGregor (London: Penguin, 1972). Clarke, Samuel, A Discourse concerning the Unchangeable Obligations of Natural Religion, and the Truth and Certainty of the Christian Revelation, in

bibliography

225

Samuel Clarke, The Works, 4 vols (London, 1738; reprint New York: Garland Publishing Co., 1978). A Demonstration of the Being and Attributes of God, and Other Writings, ed. Ezio Vailati (Cambridge: Cambridge University Press, 1998). Cocking, J. M., Imagination: A Study in the History of Ideas (London and New York: Routledge, 1991). Coleman, Dorothy, ‘Interpreting Hume’s Dialogues’, Religious Studies 25 (1989), 179–90. ‘Introduction’ to David Hume, Dialogues Concerning Natural Religion, ed. Dorothy Coleman (New York: Cambridge University Press, 2007), xi–xl. Colie, Rosalie L., Light and Enlightenment: A Study of the Cambridge Platonists and the Dutch Arminians (Cambridge: Cambridge University Press, 1957). Cudworth, Ralph, The True Intellectual System of the Universe, wherein all the reason and philosophy of atheism is confuted and its impossibility demonstrated (London, 1678). Cunningham, Andrew S., ‘Hume’s Vitalism and its Implications’, British Journal for the History of Philosophy 15 (2007), 59–73. Delumeau, Jean, Sin and Fear: The Emergence of Western Guilt Culture: 13th to 18th Centuries, tr. Eric Nicholson (New York: St Martin’s Press, 1990). Descartes, Ren´e, The Philosophical Writings of Descartes, tr. and ed. John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, 3 vols (Cambridge: Cambridge University Press, 1984–91). Dixon, Thomas, From Passions to Emotions (Cambridge: Cambridge University Press, 2003). Edwards, Jonathan, A Treatise Concerning the Religious Affections (1746), ed. John E. Smith, in The Works of Jonathan Edwards, ed. Perry Miller, 23 vols (New Haven: Yale University Press, 1957–), ii. 84–461. The Freedom of the Will (1754), ed. Paul Ramsey, in The Works of Jonathan Edwards, ed. Perry Miller, 23 vols (New Haven: Yale University Press, 1957–), i. 129–439. The Works of Jonathan Edwards, ed. Perry Miller, 23 vols. (New Haven: Yale University Press, 1957–). Ellin, Joseph, ‘Streminger: ‘‘Religion a Threat to Morality’’ ’, Hume Studies 15 (1989), 295–300.

226 bibliography Emerson, Roger L., ‘Heresy, the Social Order, and English Deism’, Church History 37 (1968), 389–403. Evans, Arthur William, Warburton and the Warburtonians: A Study in Some Eighteenth-Century Controversies (London: Oxford University Press, 1932). Febvre, Lucien, The Problem of Unbelief in the Sixteenth Century: The Religion of Rabelais, tr. Beatrice Gottlieb (Cambridge, MA: Harvard University Press, 1982). Feder, Johann Georg Henirich, review of ‘Of Suicide’ in G¨ottingische Anziegen von gelehrten Sachen, December 31 1784, No. 210, 2100–03, tr. Curtis Bowman and anthologized in James Fieser (ed.), Early Responses to Hume’s Writings on Religion, 2 vols (Bristol: Thoemmes Press, 2001). Feiring, Norman, Jonathan Edwards’s Moral Thought in its British Context (Chapel Hill, NC: The University of North Carolina Press, 1981). Fieser, James, ‘Hume’s Concealed Attack on Religion and his Early Critics’, Journal of Philosophical Research, 20 (1995), 83–101. (ed.) Early Responses to Hume’s Writings on Religion, 2 vols (Bristol: Thoemmes Press, 2001). (ed.) Early Responses to Hume’s Metaphysical and Epistemological Writings, 2 vols (Bristol: Thoemmes Press, 2005). Flage, Daniel E., ‘Hume’s Relative Ideas’, Hume Studies 7 (1981), 55–73. ‘Relative Ideas Revisited: A Reply to Thomas’, Hume Studies 8 (1982), 158–71. ‘Relative Ideas Re-Viewed’, in Rupert Read and Kenneth A. Richman (eds), The New Hume Debate: Revised Edition (New York: Routledge, 2007), 138–55. Flew, Anthony, Hume’s Philosophy of Belief: A Study of his First Inquiry (London: Routledge and Kegan Paul, 1961). God and Philosophy (London: Hutchinson and Co., 1966). Foley, Rich, ‘Unnatural Religion: Indoctrination and Philo’s Reversal in Hume’s Dialogues Concerning Natural Religion’, Hume Studies 32 (2006), 83–112. Fosl, Peter S., ‘Doubt and Divinity: Cicero’s Inﬂuence on Hume’s Religious Skepticism’, Hume Studies 20 (1994), 103–120. Foster, Stephen Paul, ‘Different Religions and the Difference They Make: Hume on the Political Effects of Religious Ideology’, The Modern Schoolman 66 (1989), 253–74.

bibliography

227

Melancholy Duty: The Hume-Gibbon Attack on Christianity (Dordrecht: Kluwer, 1997). Garrett, Don, Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997). ‘A Small Tincture of Pyrrhonism: Skepticism and Naturalism in Hume’s Science of Man’, in Walter Sinnott-Armstrong (ed.), Pyrrhonian Skepticism (Oxford: Oxford University Press, 2004), 68–98. ‘Reasons to Act and Believe: Naturalism and Rational Justiﬁcation in Hume’s Philosophical Project’, Philosophical Studies 132 (2007), 1–16. Gaskin, J. C. A., ‘Hume, Atheism, and the ‘‘Interested Obligation’’ of Morality’, in David Fate Norton, Nicholas Capaldi, and Wade L. Robison (eds), McGill Hume Studies (San Diego: Austin Hill Press, 1976), 147–59. Hume’s Philosophy of Religion, 2nd edn (Atlantic Highlands, NJ: Humanities Press International, 1988). ‘Hume on Religion’, in David Fate Norton (ed.), The Cambridge Companion to Hume (Cambridge: Cambridge University Press, 1993), 313–44. ‘Religion: The Useless Hypothesis’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), 349–69. Gastrell, Francis, The Certainty and Necessity of Religion in general (London, 1697). Gauthier, David, ‘Why Ought One Obey God? Reﬂections on Hobbes and Locke’, Canadian Journal of Philosophy 7 (1977), 425–46. Gawlick, Günter, ‘Hume and the Deists: A Reconsideration’, in G. P. Morice (ed.), David Hume: Bicentenary Papers (Edinburgh: Edinburgh University Press, 1977), 128–38. Gill, Michael, The British Moralists on Human Nature and the Birth of Secular Ethics (Cambridge: Cambridge University Press, 2006). Giovanni, George di, ‘Hume, Jacobi, and Common Sense: An Episode in the Reception of Hume in Germany at the Time of Kant’, Kant-Studien, 89 (1998), 44–58. Guyer, Paul, Knowledge, Reason, and Faith: Kant’s Response to Hume (Princeton University Press: Princeton, 2008). Haakonssen, Knud, ‘Moral Philosophy and Natural Law: From the Cambridge Platonists to the Scottish Enlightenment’, Political Science 40 (1988), 97–110.

228 bibliography Haakonssen, Knud, Natural Law and Moral Philosophy: From Grotius to the Scottish Enlightenment (Cambridge: Cambridge University Press, 1996). Harris, James A., ‘Answering Bayle’s Question: Religious Belief in the Moral Philosophy of the Scottish Enlightenment’, in Daniel Garber and Steven Nadler (eds), Oxford Studies in Early Modern Philosophy: Volume I (Oxford: Oxford University Press, 2003), 229–53. Of Liberty and Necessity (Oxford: Clarendon Press, 2005). ‘Hume’s Use of the Rhetoric of Calvinism’, in M. Frasca-Spada and P. J. E. Kail (eds), Impressions of Hume (Oxford: Clarendon Press, 2005). Harris, John, The Atheistical Objections against the Being of a God, and his Attributes, fairly considered and fully refuted (London: printed by J. L. for Richard Wilkin, 1698). Harvey, Van. A., ‘Is There Anything Religious About Philo’s ‘‘True Religion’’?’, in D. Z. Phillips and Timothy Tessin (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 68–80. Hatﬁeld, Gary S., ‘The Cognitive Faculties,’ in Daniel Garber and Michael Ayers (eds), The Cambridge History of Seventeenth-Century Philosophy, 2 vols (Cambridge: Cambridge University Press, 1998), ii. 953–1002. Helm, Paul, Eternal God (Oxford: Clarendon Press, 1988). John Calvin’s Ideas (New York: Oxford University Press, 2004). Herdt, Jennifer, Religion and Faction in Hume’s Moral Philosophy (Cambridge: Cambridge University Press, 1997). Hick, John, Evil and the God of Love, rev edn, (New York: Harper & Row, 1977). Hobbes, Thomas, The Questions Concerning Liberty, Necessity, and Chance, in The English Works of Thomas Hobbes ed. William Molesworth, 11 vols (London, 1839–45), v. 1–440. The English Works of Thomas Hobbes ed. William Molesworth, 11 vols (London, 1839–45). Third Set of Objections, collected in Ren´e Descartes, The Philosophical Writings of Descartes, tr. and ed. John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, 3 vols (Cambridge: Cambridge University Press, 1984–91), ii. 121–37. Human Nature and De Corpore Politico, ed. J. C. A. Gaskin (New York: Oxford University Press, 1994). Leviathan, ed. Edwin Curley (Hackett: Indianapolis, 1994).

bibliography

229

On the Citizen, ed. Richard Tuck and Michael Silverthorne (Cambridge: Cambridge University Press, 1998). Holden, Thomas, The Architecture of Matter: Galileo to Kant (Oxford: Clarendon Press, 2004). ‘Religion and Moral Prohibition in Hume’s ‘‘Of Suicide’’ ’, Hume Studies 31 (2005), 189–210. ‘Hume on Religious Affect’, Archiv für Geschichte der Philosophie 89 (2007), 283–306. Horne, George, Letters on Inﬁdelity. By the Author of a Letter to Doctor Adam Smith (1784), in James Fieser (ed.), Early Responses to Hume’s Writings on Religion, 2 vols (Bristol: Thoemmes Press, 2001). Howard-Snyder, Daniel (ed.), The Evidential Argument from Evil (Bloomington and Indianapolis: Indiana University Press, 1996). Humber, James, ‘Hume on Liberty, Necessity, Morality and Religion’, Philosophical Inquiry 21 (1999), 17–31. Hume, David, The Letters of David Hume, ed. J. Y. T. Grieg, 2 vols (Oxford: Clarendon Press, 1931). New Letters of David Hume, ed. Raymond Klibansky and Ernest C. Mossner (Oxford: Clarendon Press, 1954). History of England, ed. William B. Todd, 6 vols. (Indianapolis: Liberty Fund, 1983). Essays, Moral, Political, and Literary, ed. E. F. Miller (Indianapolis: Liberty Fund, 1987). An Enquiry concerning Human Understanding: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2000). An Enquiry concerning the Principles of Morals: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2003). Dialogues concerning Natural Religion, in Dialogues concerning Natural Religion: And Other Writings, ed. Dorothy Coleman (Cambridge: Cambridge University Press, 2007). A Letter from a Gentleman to his Friend in Edinburgh, in A Treatise of Human Nature: A Critical Edition, ed. D. F. Norton and M. J. Norton, 2 vols (Oxford: Clarendon Press, 2007), 419–31. A Treatise of Human Nature: A Critical Edition, ed. D. F. Norton and M. J. Norton, 2 vols (Oxford: Clarendon Press, 2007). An Abstract of a book lately published, entitled, A Treatise of Human Nature, in A Treatise of Human Nature: A Critical Edition, ed. D. F.

230 bibliography Norton and M. J. Norton, 2 vols (Oxford: Clarendon Press, 2007), i. 403–17. ‘A Dissertation on the Passions’, in A Dissertation on the Passions; The Natural History of Religion: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2008), 1–29. ‘The Natural History of Religion’ in A Dissertation on the Passions; The Natural History of Religion: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 2008), 31–87. Hutcheson, Francis, An Essay on the Nature and Conduct of the Passions and Affections, with Illustrations on the Moral Sense, ed. Aaron Garrett (Indianapolis: Liberty Fund, 2002). An Inquiry into the Original of Our Ideas of Beauty and Virtue, ed. Wolfgang Leidhold (Indianapolis: Liberty Fund, 2004). On the Natural Sociability of Mankind, in Francis Hutcheson, Logic, Metaphysics, and the Natural Sociability of Mankind, ed. James Moore and Michael Silverthorne (Indianapolis: Liberty Fund, 2006), 189–216. Logic, Metaphysics, and the Natural Sociability of Mankind, ed. James Moore and Michael Silverthorne (Indianapolis: Liberty Fund, 2006). A Synopsis of Metaphysics Comprehending Ontology and Pneumatology, in Francis Hutcheson, Logic, Metaphysics, and the Natural Sociability of Mankind, ed. James Moore and Michael Silverthorne (Indianapolis: Liberty Fund, 2006), 57–187. Philosophiae Moralis Institutio Compendiaria, ed. Luigi Turco (Indianapolis: Liberty Fund, 2007). Immerwahr, John, ‘God and Morality in Hume’s Suppressed Essays’, International Studies in Philosophy 11 (1979), 91–102. ‘Hume’s Aesthetic Theism’, Hume Studies, 22 (1996), 325–37. Israel, Jonathan I., Radical Enlightenment: Philosophy and the Making of Modernity, 1650–1750 (Oxford: Oxford University Press, 2001). Enlightenment Contested: Philosophy, Modernity, and the Emancipation of Man, 1670–1752 (Oxford: Oxford University Press, 2006). Jacobi, Friedrich Heinrich, David Hume über den Glauben, oder Idealismus und Realismus, ein Gerspr¨ach (Breslau: 1787). Jacquette, Dale, ‘Analogical Inference in Hume’s Philosophy of Religion’, Faith and Philosophy 2 (1985), 287–94. David Hume’s Critique of Inﬁnity (Leiden: Brill Academic Publishers, 2001).

bibliography

231

James, Susan, Passion and Action: the Emotions in Seventeenth-Century Philosophy (New York: Oxford University Press, 1999). Jones, Peter, Hume’s Sentiments: Their Ciceronian and French Context (Edinburgh: Edinburgh University Press, 1982). Jossua, John-Pierre, Pierre Bayle ou l’obsession du mal (Paris: Aubier Montaigne, 1977). Kail, P. J. E., ‘Hutcheson’s Moral Sense: Skepticism, Realism, and Secondary Qualities’, History of Philosophy Quarterly 18 (2001), 57–77. Projection and Realism in Hume’s Philosophy (Oxford: Oxford University Press, 2007). ‘Understanding Hume’s Natural History of Religion’, Philosophical Quarterly 57 (2007), 190–211. Kant, Immanuel, Gesammelte Schriften, edited by the Royal Prussian (later German) Academy of Sciences (Berlin: Georg Reimer, later Walter deGruyter & Co., 1900–). Critique of Pure Reason, tr. and ed. Werner S. Pluhar (Indianapolis: Hackett, 1996). Lectures on Ethics, ed. Peter Heath and J. B. Schneewind, tr. Peter Heath (Cambridge: Cambridge University Press, 1997). Kenny, Anthony, The God of the Philosophers (Oxford: Clarendon Press, 1979). King, William, An Essay on the Origin of Evil (1731) (Reprint: New York: Garland, 1978). Kingsley, Charles, His Letters and Memories of His Life, ed. Frances Eliza Kingsley, 2 vols. (London: H. S. King and Co., 1877). Knox, R. A., Enthusiasm: A Chapter in the History of Religion (Oxford: Clarendon Press, 1950). Korsgaard, Christine, ‘The General Point of View: Love And Moral Approval in Hume’s Ethics’, Hume Studies 25 (1999), 3–41. Kraay, Klaas J., ‘Philo’s Argument for Divine Amorality Reconsidered’, Hume Studies 29 (2003), 283–304. Kremer, Elmar J., and Latzer, Michael J. (ed.), The Problem of Evil in Early Modern Philosophy (Toronto: University of Toronto Press, 2001). Kristellar, Paul Oskar, ‘The Myth of Renaissance Atheism and the French Tradition of Free Thought’, Journal of the History of Philosophy 6 (1968), 233–43. Laird, John, Hume’s Philosophy of Human Nature (London: Methuen & Co., 1932).

232 bibliography Lee, San Hyun, The Philosophical Theology of Jonathan Edwards (Princeton: Princeton University Press, 1988). Leechman, William, On the Nature, Reasonableness, and Advantages of Prayer (Glasgow, 1743). Lennon, Thomas M., Reading Bayle (Toronto: Toronto University Press, 1999). Livingston, Donald W., Hume’s Philosophy of Common Life (Chicago: Chicago University Press, 1984). Philosophical Melancholy and Delirium (Chicago: Chicago University Press, 1998). Locke, John, An Essay concerning Human Understanding, ed. P. H. Nidditch (Oxford: Oxford University Press, 1975). Two Treatises of Government, ed. Peter Laslett (Cambridge: Cambridge University Press, 1988). Logan, Beryl, ‘ ‘‘Aiding the Ascent of Reason with the Wings of Imagination’’: The Prospect of a Future State’, Hume Studies 25 (1999), 193–205. Luther, Martin, Martin Luther: Selections from His Writings, ed. John Dillenberger (Garden City, NY: Anchor, 1961). McCormick, Miriam, ‘A Change in the Manner: Hume’s Scepticism in the Treatise and the First Enquiry’, Canadian Journal of Philosophy 29 (1999), 431–48. McIntyre, Jane L., ‘Passion and Artiﬁce in Hume’s Account of Superstition’, in D. Z. Phillips and Timothy Tessin (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999), 171–84. ‘Hume’s ‘‘New and Extraordinary’’ Account of the Passions’, in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), 199–215. Mackie, J. L., Hume’s Moral Theory (London and New York: Routledge, 1980). The Miracle of Theism (Oxford: Clarendon Press, 1982). McLean, G. R., ‘Hume and the Theistic Objection to Suicide’, American Philosophical Quarterly 38 (2001), 99–111. Malebranche, Nicolas, Treatise on Nature and Grace, tr. and ed. Patrick Riley (Oxford: Clarendon Press, 1992). The Search After Truth, tr. and ed. Thomas M. Lennon and Paul J. Olscamp (Cambridge: Cambridge University Press, 1997).

bibliography

233

Martinich, A. P., The Two Gods of Leviathan: Thomas Hobbes on Religion and Politics (Cambridge: Cambridge University Press, 1992). Merrill, Kenneth R., ‘Hume on Suicide’, History of Philosophy Quarterly 16 (1999), 395–412. Merrill, Kenneth R., and Wester, Donald G., ‘Hume on the Relation of Religion to Morality’, The Journal of Religion 60 (1980), 272–84. Mill, John Stuart, Collected Works, 33 vols (Toronto: University of Toronto Press, 1963). Millican, Peter, ‘The Context, Aims, and Structure of Hume’s First Enquiry’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 27–65. (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002). Milton, John, Prose Works, ed. Don M. Wolfe, 8 vols. (New Haven, CT: 1953–82). Paradise Lost, ed. Christopher Ricks (New York: New American Library, 1968). Montaigne, Michel de, The Complete Essays, tr. and ed. M. A. Screech (Harmonsdworth: Penguin, 1987). Moore, A. W., The Inﬁnite, 2nd edn (London and New York: Routledge, 2001). Moore, James, ‘Hume and Hutcheson’, in M. A. Stewart and John P. Wright (eds.), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), 24–57. More, Henry, Enthusiasmus Triumphatus, in A Collection of Several Philosophical Writings of Dr. Henry More, 2nd edn (London, 1662). Mori, Gianluca, Bayle Philosophe (Paris: Champion, 1999). Mossner, E. C., ‘Hume’s Early Memoranda, 1729–1740: The Complete Text’, Journal of the History of Ideas 9 (1948), 292–518. The Life of David Hume, 2nd edn (Oxford and New York: Oxford University Press, 1980). Neiman, Susan, Evil in Modern Thought: An Alternative History of Philosophy (Princeton, NJ: Princeton University Press, 2002). Nielsen, Kai, ‘Some Remarks on the Independence of Morality from Religion’, Mind 70 (1961), 175–86. An Introduction to the Philosophy of Religion (London: Macmillan, 1982).

234 bibliography Norton, David Fate, David Hume: Common-Sense Moralist, Sceptical Metaphysician (Princeton: Princeton University Press, 1982). ‘Hutcheson’s Moral Realism’, Journal of the History of Philosophy 23 (1985), 397–418. ‘Hume, Atheism, and the Autonomy of Morals’, in Anthony Flew, Donald Livingston, George I. Mavrodes, and David Fate Norton (eds), Hume’s Philosophy of Religion (Winston-Salem, NC: Book Service Associates, 1986), 97–144. ‘Hume, human nature, and the foundations of morality’, in David Fate Norton (ed.), The Cambridge Companion to Hume (Cambridge: Cambridge University Press, 1993), 193–205. (ed.) The Cambridge Companion to Hume (Cambridge: Cambridge University Press, 1993). ‘Of the Academical or Sceptical Philosophy’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 371–92. Nussbaum, Martha, Upheavals of Thought: The Intelligence of Emotions (Cambridge: Cambridge University Press, 2001). O’Connor, David, Hume on Religion (London: Routledge, 2001). ‘Skepticism and Philo’s Atheistic Preference’, Hume Studies 29 (2003), 267–82. Pascal, Blaise, Pens´ees, tr. A. J. Krailsheimer (London: Penguin, 1995). Passmore, John, The Perfectibility of Man, 2nd edn (London: Duckworth, 1971). Penelhum, Terence, Hume (New York: St. Martin’s Press, 1975). God and Skepticism: A Study in Skepticism and Fideism (Dordrecht: D. Reidel, 1983). Butler (London: Routledge and Kegan Paul, 1985). ‘Butler and Human Ignorance’, in Christopher Cunliffe (ed.), Joseph Butler’s Moral and Religious Thought: Tercentenary Essays (Oxford: Clarendon Press, 1992), 117–39. Themes in Hume: The Self, the Will, Religion (Oxford: Clarendon Press, 2000). Phillips, D. Z., Religion and the Hermeneutics of Contemplation (Cambridge: Cambridge University Press, 2001). Phillips, D. Z., and Tessin, Timothy (eds), Religion and Hume’s Legacy (New York: St. Martin’s Press, 1999).

bibliography

235

Pike, Nelson, ‘Hume on Evil’, The Philosophical Review 72 (1963), 180–97. Plantinga, Alvin, God, Freedom and Evil (Grand Rapids: Eerdmans, 1977). ‘Epistemic Probability and Evil’, in Daniel Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington and Indianapolis: Indiana University Press, 1996), 69–96. Pope, Alexander An Essay on Man (London: J. Wilford, 1733–4). Popkin, Richard H., The History of Scepticism: From Savonarola to Bayle (Oxford: Oxford University Press, 2003). Porter, B. F., Deity and Morality (London: Allen and Unwin, 1968). Priestley, Joseph, The doctrine of philosophical necessity illustrated, appended to Disquisitions relating to matter and spirit (1777). Prinz, Jesse, The Emotional Construction of Morals (Oxford: Oxford University Press, 2007). Pufendorf, Samuel, Of the Law of Nature and Nations, tr. Basil Kennet et al. with the notes of Jean Barbeyrac (London, 1729). On the Whole Duty of Man and Citizen, tr. Michael Silverthorne, ed. James Tully (Cambridge: Cambridge University Press, 1991). Radcliffe, Elizabeth S., ‘Hutcheson’s Perceptual and Moral Subjectivism’, History of Philosophy Quarterly 3 (1986), 407–21. ‘Love and Benevolence in Hutcheson’s and Hume’s Theories of the Passions’, British Journal for the History of Philosophy 12 (2004), 631–53. Read, Rupert, and Richman, Kenneth A. (eds), The New Hume Debate: Revised Edition (New York: Routledge, 2007). Redmond, M., ‘The Hamann–Hume Connection’, Religious Studies 23 (1987), 95–107. Redwood, John, Reason, Ridicule and Revelation: The Age of Enlightenment in England, 1660–1750 (Cambridge, MA: Harvard University Press, 1976). Reid, Jasper, ‘The Metaphysics of Jonathan Edwards and David Hume’, Hume Studies 32 (2006), 53–82. Reid, Thomas, Essays on the Active Powers of Man, in Thomas Reid, Philosophical Works, ed. William Hamilton, 2 vols (Edinburgh 1895; facsimile reprint Hildesheim: Georg Olms, 1967). Philosophical Works, ed. William Hamilton, 2 vols (Edinburgh 1895; facsimile reprint Hildesheim: Georg Olms, 1967). Rivers, Isabel, Reason Grace, and Sentiment: A Study of the Language of Religion and Ethics in England, 1660–1780, 2 vols (Cambridge: Cambridge University Press, ﬁrst volume 1991, second volume 2000).

236 bibliography Rivers, Isabel, ‘Response to Hume on Religion by Anglicans and Dissenters’, Journal of Ecclesiastical History 52 (2001), 675–95. Robertson, John, The Case for the Enlightenment: Scotland and Naples 1688–1760 (Cambridge: Cambridge University Press, 2005). Russell, Paul, ‘Skepticism and Natural Religion in Hume’s Treatise’, Journal of the History of Ideas 49 (1988), 247–65. Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility (Oxford and New York: Oxford University Press, 1995). ‘Hume’s Treatise and the Clarke–Collins Controversy’, Hume Studies 21 (1995), 95–115. ‘Butler’s ‘‘Future State’’ and Hume’s ‘‘Guide to Life’’ ’, Journal of the History of Philosophy 42 (2004), 425–48. The Riddle of Hume’s Treatise: Skepticism, Naturalism and Irreligion (New York: Oxford University Press, 2008). Schneewind, J. B., The Invention of Autonomy: A History of Modern Moral Philosophy (Cambridge: Cambridge University Press, 1998). ‘Hume and the Religious Signiﬁcance of Moral Rationalism’, Hume Studies 26 (2000), 211–23. Sessions, William Lad, Reading Hume’s Dialogues: A Veneration for True Religion (Bloomington and Indianapolis, Indiana University Press, 2002). Shaftesbury, Earl of [Anthony Ashley Cooper], ‘The Moralists; a Philosophical Rhapsody,’ in Shaftesbury, Characteristics of Men, Manners, Opinions, Times, ed. Douglas Den Uyl, 3 vols (Indianapolis: Liberty Fund, 2001), 101–247. Characteristics of Men, Manners, Opinions, Times, ed. Douglas Den Uyl, 3 vols (Indianapolis: Liberty Fund, 2001). Siebert, Donald T., ‘Hume on Idolatry and Incarnation’, Journal of the History of Ideas 45 (1984), 379–96. The Moral Animus of David Hume (Newark, NJ: Delaware University Press, 1990). Simmons, A. J., The Lockean Theory of Rights (Princeton: Princeton University Press, 1992). Sinsart, Dom, Defense du Dogme Catholique ser l’Eternit´e des Peines (Strasbourg, 1748). Skinner, Quentin, Reason and Rhetoric in the Philosophy of Hobbes (Cambridge: Cambridge University Press, 1996).

bibliography

237

Smart, Ninian, The Concept of Worship (Edinburgh: Macmillan, 1972). Smith, Adam, The Theory of the Moral Sentiments, ed. Knud Haakonssen (Cambridge: Cambridge University Press, 2002). Smith, John, Select Discourses (London, 1660). Smith, Norman Kemp, ‘Introduction’ to David Hume, Dialogues concerning Natural Religion, ed. Norman Kemp Smith, 2nd edn (Indianapolis: Bobbs-Merrill, 1947), 1–75. The Philosophy of David Hume: A Critical Study of its Origins and Central Doctrines, with a new introduction by Don Garrett (New York: Palgrave Macmillan, 2005). Spinoza, Benedict de [Baruch], Ethics, ed. and tr. G. H. R. Parkinson (Oxford: Oxford University Press, 2000). Steele, Richard B., ‘Gracious Affection’ and ‘True Virtue’ According to Jonathan Edwards and John Wesley (Metuchen, NJ: The Scarecrow Press, 1994). Stewart, Dugald, Collected Works, ed. Sir William Hamilton, 11 vols (Edinburgh: T. Constable, 1854–60). Stewart, M. A., ‘An early fragment on evil’, in M. A. Stewart and John P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), 160–70. ‘The Dating of Hume’s Manuscripts,’ in Paul Wood (ed.), The Scottish Enlightenment: Essays in Reinterpretation (Rochester: University of Rochester Press, 2000), 276–88. ‘Two Species of Philosophy: The Historical Signiﬁcance of the First Enquiry’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), 67–95. ‘The Early British Reception of Hume’s Writings on Religion’, in Peter Jones (ed.), The Reception of Hume in Europe (London and New York: Thoemmes Continuum, 2005), 30–42. Strawson, Galen, The Secret Connexion (Oxford: Clarendon Press, 1989). Streminger, Gerhard, ‘Religion a Threat to Morality: An Attempt to Throw Some New Light on Hume’s Philosophy of Religion’, Hume Studies 15 (1989), 277–94. ‘A Reply to Ellin,’ Hume Studies 15 (1989), 301–6. Stroud, Barry, Hume (London and New York: Routledge, 1977). Stump, Eleonore, ‘The Problem of Evil’, Faith and Philosophy 2 (1985), 392–423. Sturgeon, Nicholas, ‘Moral Skepticism and Moral Naturalism in Hume’s Treatise’, Hume Studies 27 (2001), 3–83.

238 bibliography Sumner, L. W., The Moral Foundation of Rights (Oxford: Oxford University Press, 1987). Swinburne, Richard, The Coherence of Theism (Oxford: Oxford University Press, 1977). Taliaferro, Charles, Evidence and Faith: Philosophy and Religion since the Seventeenth Century (Cambridge: Cambridge University Press, 1995). Taylor, Jacqueline, ‘Virtue and the Evaluation of Character’, in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), 276–95. Tillotson, John, Sermons, ed. Ralph Barker, 14 vols (London: 1695–1704). Sermons on several subjects and occasions, 12 vols (London: 1742–4). Tillyard, E. M. W., The Elizabethan World Picture (London: Chatto and Windus, 1943). Traherne, Thomas, Centuries of Meditations (London: P. J. & A. E. Dobell, 1950). Traiger, Saul (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006). Tweyman, Stanley, ‘Hume’s Dialogues on Evil’, Hume Studies 13 (1987), 74–85. Wainright. William J., ‘Theological Determinism and the Problem of Evil: Are Arminians any better off?’, International Journal for the Philosophy of Religion 50 (2001), 81–96. Walker, D. P., The Decline of Hell: Seventeenth-Century Discussions of Eternal Torment (London: Routledge and Kegan Paul, 1964). Wallis, Jerry L., ‘Hume on Divine Amorality’, Religious Studies 26 (1990), 257–266. Warburton, William, A View of Lord Bolingbroke’s philosophy; in four letters to a friend. Letters ﬁrst and second (London, 1754). Remarks on Mr. David Hume’s Essay on the natural history of religion (London, 1757). Letters of a late Eminent Prelate to One of His Friends, ed. Richard Hurd (Kidderminster, 1808). Watts, Isaac, Discourses of the love of God, and its inﬂuence on all the passions (Philadelphia: Woodward, 1799). Whichcote, Benjamin, The works of the learned Benjamin Whichcote (London: J. Chalmers, 1751). Williams, Bernard, ‘Hume on Religion’, in D. F. Pears (ed.), David Hume: A Symposium (New York: St Martin’s Press, 1966), 77–88.

bibliography

239

The Sense of the Past: Essays in the History of Philosophy, ed. Myles Burnyeat (Princeton: Princeton University Press, 2006). Windstrup, George, ‘Locke on Suicide’, Political Theory 8 (1980), 169–82. Winkler, Kenneth P., ‘Hutcheson and Hume on the Color of Virtue’, Hume Studies 22 (1996), 3–22. Wootton, David, Paolo Sarpi: Between Renaissance and Enlightenment (Cambridge: Cambridge University Press, 1983). ‘Pierre Bayle, Libertine?’ in M. A. Stewart (ed.), Studies in Seventeenth Century Philosophy (Oxford: Oxford University Press, 1997), 197–226. Wright, John P., The Sceptical Realism of David Hume (Minneapolis: University of Minnesota Press, 1983). Wykstra, Stephen, ‘Rowe’s Noseeum Arguments from Evil’, in Daniel Howard-Synder (ed.), The Evidential Argument from Evil, (Bloomington: Indiana University Press, 1996), 126–50. Yandell, Keith E., Hume’s ‘Inexplicable Mystery’: His Views on Religion (Philadelphia: Temple University Press, 1990). Young, B. W., Religion and Enlightenment in Eighteenth-Century England: Theological Debate from Locke to Burke (Oxford: Clarendon Press, 1998). Zakai, Avihu, Jonathan Edwards’ Philosophy of History (Princeton, NJ: Princeton University Press, 2003).

This page intentionally left blank

Index agnosticism moral 19, 212, 215–16 quietist 125, 126–7 square 125, 126 strict 125–7 Alanen, Lilli 55 n. Allen, D. C. 9 n. Alston, William P. 179 n. amorality divine, see atheism, moral Aquinas, St Thomas 85, 185 ´ Ardal, P´all 55 n., 97 n. argument from design 29, 36–7, 45–6, 139 argument from determinism against duties to God 39, 181–2, 186–93, 210 argument from determinism and evil 40, 182, 196–206, 210 argument from environmental need 123–4, 129 argument from motivation 115, 117–23, 142–3, 209 argument from sentimentalism 44, 49–51, 64, 95–6, 113–14, 209 objections to 100–13 argument from speciﬁcity 125–9 Aristotle 85 Arminianism 200, 201, 202 n. Arnauld, Antoine 134 n. artiﬁcial lives, see Hume, David, on artiﬁcial lives atheism ancient polytheism as a form of 5 n. and libertinism 9, 10 contemplative 10 early modern use of the term ‘atheism’ 9–14 moral 2, 4–9, 11, 12, 28, 50, 99–100, 113, 209–19 implications of for human life 3, 215–16, 218–19

strong 8–9, 19, 99–100, 113, 115, 209 and voluntarism 213 n. weak 7–8, 115, 145, 169, 182, 196, 210 practical 10 speculative 10, 11 attributes of the deity 5 n., 10 analogy with human attributes 17, 30, 35–7, 133, 151–2, 168, 191–2 see also God, human tendency to anthropomorphize immutability 30 inﬁnitude 87, 154–8, 196 intelligence 30, 33, 35–6, 192 intrinsic 29–36, 100, 116 moral 1–2, 4, 7–9, 12, 15, 115–16, 119–20, 202, 209–10, 215–16 omniscience 192 passions 34–5, 115–16, 117–18 relational 36–44 Augustine, St, of Hippo 157 Bacon, Francis 10, 85 n. Baier, Annette 97 n. Balguy, John 43 n. Battersby, Christine 14 n. Bayle, Pierre 7 n., 24, 131 n., 146, 147, 150 n., 154, 157, 158, 191 n., 200–1, 218 Beattie, James 11 Beauchamp, Tom L. 60 n. Beiser, Frederick 74 n., 213 n., 217 n., 218 n. Bentley, Richard 13, 45 Berkeley, George 1, 3, 12, 43 n., 91 n., 159, 200, 216 n. Berlin, Isaiah 217 n. Berman, David 11 n., 12 n., 16 n., 147 n.

242 index Blackburn, Simon 3 n., 42 n., 102 n., 216 n. Blackmore, Richard 11 Bolingbroke, Lord (Henry St John) 14 n. Borges, Jorge Luis 174 Botterill, George 16 n., 195 n., 199 n. Boyle lectures 13 Boyle, Robert 45 Brahami, Fr´ed´eric 218 Broad, C. D. 161 Buckle, Stephen 3n., 23 n., 138 n., 199 n., 216 n. Buckley, George T. 13 n. Buckley, Michael J. 13 n. Butler, Joseph 13, 45, 57 n., 147, 159, 160–8, 185 Calvin, Jean 205–6, 213 Calvinism 63–4, 81–2, 90, 92, 133 n., 200, 202 n. Capaldi, Nicholas 3n. Catholicism 79 causal reasoning 110–11 and the uniqueness of the universe 112, 136–42, 165–6, 167–8 Cheke, John 13 n. Church history and iconography 81 n. and predestinarianism 190 n., 202 n. Scottish 63, 81–2, 202 n. Cicero, Marcus Tullius 14 n., 85 n., 122 n. Clarke, Samuel 11, 45, 130, 184, 185, 213 n. Coleman, Dorothy 117 n. compatibilism, see determinism and human responsibility cosmological argument 45 Cudworth, Ralph 13, 130, 213 n. Darwin, Charles 12 deism 10 deity, see God

Descartes, Ren´e 67, 87 n. design, see argument from design determinism 38–40, 181–2, 186–7, 190 n., 193–5, 196–8, 206–7 and divine responsibility 193–5, 197–8, 200, 203–4, 205–6 and human responsibility 193–5 see also argument from determinism against duties to God see also argument from determinism and evil Diogenes 59 n. Dixon, Thomas 92 n. Dominic, St 106 Donne, John 86 duties 183, 185 owed to God 183–5, 189, 190–1 ‘Hohfeldian’ or relational concept of 186 n. tripartite scheme of 183, 184–5 Edwards, Jonathan 89–93, 205 Elliot, Gilbert 25 empiricism 67, 85 enthusiasm 63, 74, 75 n., 90, 91 Epicureanism 10 Epicurus 6 n., 150 n. evil 8, 145–9 as suffering 148–9 coherence problem of, see evil, logical problem of evidential problem of 8, 29, 145, 147, 168–78, 179, 210 inference problem of 145, 147, 158–67 logical problem of 8, 145, 147, 149–58 moral evil contrasted with natural evil 148–9, 199 see also argument from determinism and evil Feder, Johann Georg Heinrich 189 n. Feiring, Norman 92 n. F´enelon, Franc¸ois de Salignac de la Mothe 60 n.

index ﬁdeism 131 n., 46 n., 216–18 Fieser, James 16 n. Flage, Daniel E. 5 n. Flew, Anthony 110 n., 138 n. Fosl, Peter S. 14 n. Garrett, Don 23 n., 28 n., 66 n., 67 n., 105 n., 195, 198 n. Gaskin, J. C. A. xii, 3 n., 4 n., 7 n., 53, 138 n., 169 n., 199 n., 216 n. Gastrell, Francis 13 Gill, Michael 43 n., 216 n. Giovanni, George di 218 God as designer 6, 7 as ﬁrst cause 6, 7, 39, 186, 197–8 human ignorance of 111, 210 human tendency to anthropomorphize 132–4, 210–11 Hume’s use of the term ‘God’ 4–7, 38, 73 n. not a natural object of the human passions 49–50, 52–3, 59–65, 83–4, 86–7, 98, 107 question of the existence of 3 n. question of the necessary existence of 30–1 see also attributes of the deity Guyer, Paul 217 n. Haakonssen, Knud 43 n., 185 Hamann, Johann Georg 131 n., 217–18 Harris, James 90 n., 146 n. Harris, John 10 Hatﬁeld, Gary S. 67 n. Helm, Paul 110 n., 206 Herdt, Jennifer 82 n. Hick, John 158 n. Hobbes, Thomas 12, 76 n., 79 n., 86–7, 200, 205, 213 Hooker, Richard 130 Horne, George 189 n. Hume, David and moral atheism 4–9, 209–19 and ‘the science of man’ xi, 40, 214

243

attitude toward natural theology 25–7, 32, 40–1, 44–6, 73 n., 125–6, 141–3, 146, 169–71, 207–8, 210–11, 214–15 critique of natural theology 16 n., 22–5, 28–9, 45, 110–13, 137–41, 178, 210 critique of religious passions 74–83, 211 critique of revealed theology 16 n. interest in problem of evil 146–7 on artiﬁcial lives 59 n., 106 on general epistemology 21–2, 45–6, 110–11, 214–15; see also Hume, David, skepticism of on probabilistic reasoning 128 n., 140 n. on question of God’s existence 3 n. on the basis of aesthetic distinctions, see sentimentalism, in aesthetic theory on the basis of moral distinctions, see sentimentalism, in moral theory on the passions 54–9, 68–72 see also Hume, David, critique of religious passions on theological implications of evil, see evil skepticism of 19, 20, 27, 28–9, 72 see also skepticism, mitigated use of irony 15–17, 103, 175–8, 213 n. Hutcheson, Francis 35, 42, 43 n., 89, 117, 131, 132, 159, 185 iconography 63–4, 77–82 ideas abstract 70–1, 72–3 and images 67, 85, 87–8, 90–1, 92 see also imagination experience as origin of 67, 91 positive 5 relative 5, 78 role in guiding the passions 67–73, 85, 86, 90–1, 111

244 index imagination 65–6, 90–1, 111 and the fancy 67 encompasses the understanding 66–7, 107 inﬁnity 156 see also attributes of the deity, inﬁnitude Jacobi, Friedrich Heinrich 217–18 Jacquette, Dale 156 n., 174 n. Jansenism 190 n., 200 Joan of Arc 77 n. Jones, Peter 14 n. Jonson, Ben 85 n. Jossua, John-Pierre 146 n. Jurieu, Pierre 131 n., 201 justice 121 n., 165 Kail, P. J. E. 42 n., 132 n. Kant, Immanuel 24, 111, 184 n., 217, 218 Kemp Smith, Norman xii, 14 n., 117 n. Kenny, Anthony 32 n. King, William 57 n., 146, 154, 157 Kingsley, Charles 4 n. Knox, R. A. 74 n. Korsgaard, Christine 97 n. Kraay, Klaas J. 149 n., 153 n., 173 n., 174., 175 n., 177 Lactantius, Lucius Caelius Firmianius 150 n. Laud, William 81 n. Lee, San Hyun 91 n. Leechman, William 17–18 n., 54 n. Leibniz, Gottried Wilhelm 60 n., 154, 157, 200, 213 n. Lennon, Thomas M. 146 n. Lipsius, Justus 60 n. Locke, John 74 n., 184, 185, 191 love 86, 97 Loyola, St Ignatius 106, 110 Lucretius 11 Luther, Martin 213 Lutheranism 190 n., 200 n.

Mackie, John 158 n. Malebranche, Nicolas 88, 92, 134 n., 185, 191 n. Manicheanism 147, 169, 173, 174, 200–1 Martinich, A. P. 12 n. McCormick, Miriam 28 n. McIntyre, Jane L. 51 n., 55 n., 73 n. McLean, G. R. 189 n. Merrill, Kenneth R. 188 n. Mill, John Stuart 158 n. Milton, John 10, 111, 200 n. Montaigne, Michel de 184, 218 Moore, A. W. 156 n. Moore, James 43 n. More, Henry 74 n. Mori, Gianluca 146 n. Mossner, E. C. 7 n., 18 n., 54 n., 146 n., 183 n. motivation depends on passions 119–20, 209 to act morally 35, 115, 119–20, 209 Mure, William 15, 17, 52, 54 n., 63, 65–6, 68, 69, 72, 75, 77 n., 90, 98, 99 n., 113 natural religion, see natural theology natural theology 23–5, 40–1 core 20, 28–9, 44, 45, 46, 135, 140–1, 207 irrelevant to morality and human life 216 n., 218–9 liminal 20, 28, 45–6, 134–7, 141–2, 157, 171, 207–8 see also Hume, David, attitude toward natural theology see also Hume, David, critique of natural theology necessity doctrine of, see determinism negative theology, see via negativa Neiman, Susan 146 n. Nicole, Pierre 133 n. Nielsen, Kai 110 n. Norton, David Fate 23 n., 43 n., 122 n. Nussbaum, Martha 73 n.

index O’Connor, David 125 n., 149 n., 153 n., 170 n., 171 n., 175 n. Origen 157 Ovid (Publius Ovidius Navo) 121 Pascal, Blaise 59 n., 88, 106, 110, 126, 131 n. passions 34–5, 43, 49 abnormal passions 58–9, 104, 105–6, 110 see also Hume, David, on artiﬁcial lives and emotional displacement 76–7, 82 calm 98, 99 n. natural objects of the 43–4, 49, 54–60, 99, 101, 109 not naturally directed toward God 49–50, 52–3, 59–65, 66, 73, 83–4, 91, 99–100, 107, 109 parochial character of essential to human survival 72 proper objects of the 57–8, 59 n., 109 psychological mechanisms controlling the 55–9, 65–6, 67–73, 78, 85, 86, 90–1 religious 74, 84 n. early modern theories of 84–93 Hume’s error theory of 75–83 sexual 77 n. violent 98, 99 n see also attributes of the deity, passions Passmore, John 216 n. Penelhum, Terence 138 n., 162 n. Philo (character in Hume’s Dialogues) acts as Hume’s spokesman xi-xii, 5, 23, 212 n. Pike, Nelson 158 n., 175 n. Plantinga, Alvin 158 n., 179 n. Plutarch 13 n. Pope, Alexander 158 n. Popkin, Richard H. 146 n., 218 prayer 53, 78 n., 84

245

predestinarianism, see Church history and predestinarianism see also argument from determinism against duties to God see also argument from determinism and evil Priestley, Joseph 199 n. Prinz, Jesse 102 n. probabilistic reasoning 128 n. Pufendorf, Samuel 13 n., 184, 185 Pyrrhonism 72 n. Quintilian (Marcus Fabius Quintilianus) 85 n. Radcliffe, Elizabeth 43 n. Ramsay, Andrew Michael 60 n., 133 n. rationalism and the proper objects of the passions 130 in moral theory 101, 130–1, 213 n. Redmond, M. 217 n. Reid, Jasper 90 n., 93 n. Reid, Thomas 185 religion 84 and moral practice 1–2, 215–16, 218–19 and pious panegyric 155–6, 211 and self-deception 75–7, 82–3 and toleration 83 n. see also Church history see also iconography see also passions, religious see also prayer revealed theology 16 n. Rivers, Isabel 12 n., 16 n., 89 n. Russell, Paul 3 n., 12 n., 21 n., 28 n., 51 n., 97 n., 149 n., 153 n., 161 n., 195 n., 198 n. Schneewind, Jerome B. 185 n., 213 n. secondary qualities 42, 50, 102 Seneca 84 n. sentimentalism in aesthetic theory 84, 104, 105

246 index sentimentalism (cont.) in moral theory 41–3, 44, 49–50, 96–7, 98, 99, 101, 104–5, 107, 131, 194–5, 213 n. role in case for moral atheism, see argument from sentimentalism sentiments as basis of moral distinctions 41–2, 96–7, 99, 104–5, 107, 131, 194–5, 213 n. see also sentimentalism, in moral theory calm 98 moral 41–2, 44, 49–51, 96–8, 104–5, 194 violent 98 see also passions Sessions, William Lad 149 n., 173 n. Shaftesbury, Lord (Anthony Ashley Cooper) 42, 43 n., 131, 132, 212 Siebert, Donald T. 82 n. Simmons, A. J. 122 n., 191 n. skepticism 72 mitigated 21–3, 26, 29, 47, 170–1, 215–16 Skinner, Quentin 85 n. Smart, Ninian 84n. Smith, Adam 57 n. Smith, John 87–8, 92 ‘spectres of false divinity’ 76, 134 Spence, Joseph 77n. Spinoza, Benedict de 12, 134 n., 212, 213 Spinozism 10 St John, Henry, see Bolingbroke Stewart, Dugald 212 n. Stewart, M. A. 54 n., 131 n., 146 n., 147 Stoicism 60, 103 Strahan, William 183 n. Strawson, Galen 5 n. Stroud, Barry 193–4 Stump, Eleonore 158 n.

suicide 183–4, 188–9 Sumner, L. W. 186 n. Taylor, Jacqueline 97 n., 105 n. teleological argument, see argument from design Tillotson, John 1, 3, 12, 215 n., 216 n. Tillyard, E. M. W. 86 n. Tindal, Matthew 212 toleration 83 n. Traherne, Thomas 85 n. Tweyman, Stanley 169 n., 171 n. understanding, see imagination, encompasses the understanding unique cause objections, see causal reasoning and the uniqueness of the universe via negativa 2, 32 virtues (and vices) 96–7, 185, 195 n., 204 artiﬁcial 121 n., 165 Voltaire (Franc¸ois-Marie Arouet) 212 voluntarism 213 n. Walker, D. P. 191 n., 200 Warburton, William 1–2, 3, 12, 13–14 Watts, Isaac 89 Webster, John 211 Wesley, John 90 n. Whichcote, Benjamin 130, 215 n. Williams, Bernard 117 n., 135 n. Windstrup, George 191 n. Winkler, Kenneth P. 42 n. Wootton, David 12 n., 146 n. Wykstra, Stephen 125 n. Yandell, Keith E. 138 n., 170 n., 173 n., 189 n. Zakai, Avihu 92 n.