Concealed Influence of Custom: Hume's Treatise from the Inside Out

Citation preview

The Concealed Influence of Custom

The Concealed Influence of Custom Hume’s Treatise from the Inside Out

vwv Jay L. Garfield

1

1 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2019 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. CIP data is on file at the Library of Congress ISBN 978–​0–​19–​093340–​1 1 3 5 7 9 8 6 4 2 Printed by Sheridan Books, Inc., United States of America

To the memory of Annette C. Baier Who taught us how to read Hume

Such is the influence of custom, that, where it is strongest, it not only covers our natural ignorance, but even conceals itself, and seems not to take place, merely because it is found in the highest degree. (EHU 4.24; SBN 28–​29)

CON T E N T S

Preface  ix PART I. Methodological Preliminaries 1. Introduction: Principles of Interpretation   3 2. Why the Treatise? Why Book II? Why Custom?   28 PART II. Book II: The Psychological Foundations 3. The Passions: Some Basic Distinctions   49 4. The Passions and Human Nature   71 5. Hume’s Moral Psychology   90 PART III. Books I and III: The Skeptical Framework Deployed 6. Epistemological Foundations   105 7. Causality   128 8. Skepticism with Regard to Reason   152 9. Of Skepticism with Regard to the Senses   172 10. Personal Identity and Philosophical Method   203 11. Ethics   225 PART IV. Living Carelessly 12. The Appendix: Second Thoughts about Second Thoughts   263 13. Persons as Customary Creatures   273 References  281 Index  289

( vii )

PR E FAC E

In 1972, the late Norman S. Care assigned me to read parts of the Treatise in my first philosophy class, in my first semester of college (a class I took by accident). After reading Hume’s discussion of personal identity, I declared a philosophy major. I am sure that I am not alone in having been lured into the study of philosophy by Hume. I have never stopped thinking about the Treatise, and this book is the distant effect of Norman Care’s teaching, but, like any effect, it has many causes and conditions. My interest in Hume was sustained as an undergraduate by an Oberlin colloquium on Hume. On that occasion, two eminent Hume scholars were extraordinarily generous and indulgent to a naive and probably overenthusiastic undergraduate. I still remember talking at length with Terence Penelhum about the structure of Hume’s ethics and the role of an ideal observer, and with Lewis White Beck about the degree to which Kant really diverged from Hume on causality. Neither of these giants of Hume scholarship had any obligation to talk to me, and I am very grateful that they indulged a novice philosopher. This is just one instance that demonstrates how important small acts of generosity to students can be, and just how much our lives depend upon chance and upon the kindness of others. I went from Oberlin to Pittsburgh, where I quickly fell under the spell of Annette Baier, who really taught me—​and so many others—​to read and to love Hume. Annette opened the Treatise as a systematic unity, and showed why it was so important to take Book II so seriously. I am always conscious of how much of my philosophical life I  owe to her patient and inspiring teaching. The reading of Hume I offer here is enormously indebted to her own. Indeed, virtually every specific idea I  advance derives in some way from her work. If there is anything I add here, it is simply to draw insights together that are scattered in her work, and to give a shape to them not explicit in her own corpus, but certainly suggested by it. So, to adopt the scholastic language appropriate to this philosophical genre, the present

( ix )

( x )  Preface

volume is perhaps a subcommentary on Annette Baier’s commentarial corpus as much as a direct commentary on the Treatise. The biggest difference the reader familiar with Baier’s exposition will note is the order in which I take the text, beginning with Book II and only then returning to Book I. That is not Baier’s approach, and she argues, especially in A Progress of Sentiments, that Book I is a necessary foundation for Book II, as Hume moves gradually from an individualistic to a social perspective. I disagree, and much of the burden of this book is to show that, as Kemp Smith (1941) argued, it is Book II that is the real foundation for both Books I and III (although I do not adopt Kemp Smith’s position that Book II was written prior to Book I, a view for which there is no evidence, and for which he provides no compelling defense). But I owe the reasons for taking Book II to be conceptually foundational entirely to Annette Baier’s insights into the Treatise. I suspect that she might approve of this small heresy. As a rather brilliant undergraduate student of mine once wrote as the first two sentences of a paper contrasting Descartes’s and Hume’s epistemological approaches, “Descartes writes in the first person singular; Hume writes in the first person plural. Everything else flows from that, because Hume, unlike Descartes, recognizes that justification is a social practice.” Annette Baier always emphasized in her reading of Hume his attention to our social and biological nature, and argued that that his social and biological perspective on human nature pervades the arguments in Book I as much as it does Books II and III. I am running with that insight. We will see how far it can be taken. The other difference between my approach and Baier’s is my account of the origin of Hume’s understanding of custom and of its role in his account of human behavior and of normativity. I am indebted to Baier for drawing attention to the many important ways this term figures in Hume’s philosophy. But I have only recently become aware of the roots of Hume’s thought about custom in legal theory, and those roots suggest a somewhat different reading of how custom figures in the Treatise from that Baier herself offers. The analysis I offer of Hume’s conception of custom, and of the ways in which that conception shapes the Treatise, may be the most distinctive feature of my reading. When I  read Hume, I  also note the resonances between Hume’s approach to philosophy and that of the Buddhist philosophers I also study, in particular Nāgārjuna and Candrakīrti. I am grateful for the opportunities I have had to study Buddhist philosophy, as it has given me perspectives on the Western tradition that would be otherwise unavailable to me. This study has also impressed on me two facts about the practice of philosophy and the study of its history. First, we always come to understand our own

Preface  ( xi )

tradition better by coming to understand another, and by having the opportunity to view our own traditions from the horizon constituted by another. Second, those of us in the West often take our own philosophical tradition as a kind of default Archimedean perspective from which philosophy is to be done and in terms of which interpretation is to proceed. I have found it useful to displace that prejudice by sometimes rereading the Western tradition from the standpoint of the Buddhist tradition. My hope is that by occasionally explaining what Hume is up to by putting a point as it is made in the Indo-​Tibetan Madhyamaka Buddhist tradition I can offer a perspective that illuminates aspects of his own position that might not otherwise be obvious. I  hope that in doing so I  am not doing violence to Hume’s text, and that I  thereby illuminate insights in each tradition through conversation with the other. I also hope that this encourages my Western-​educated colleagues to read more broadly in the Buddhist and in other traditions, both for their own sake, and as sources of insight into our own. This may also help to make Hume more accessible to my colleagues in Buddhist studies to whom I believe he could become a valuable resource. One of my purposes in writing this book is to invite these colleagues to take Hume seriously as a philosophical resource for their own work. So, there are four interpretative perspectives at work in this volume, that, while conceptually independent, interact with one another: one is a naturalistic skeptical interpretation of Hume’s Treatise defended through a reading of that text on its own terms; a second is the foregrounding of Book II of the Treatise as foundational for Books I and III. A third is the consideration of the Treatise in relation to Hume’s philosophical antecedents (particularly Sextus, Bayle, Hutcheson, Shaftesbury, and Mandeville), as well as eighteenth-​century debates about the status of customary law, with one eye on its sequelae in the work of Kant, in the later Wittgenstein, and in contemporary cognitive science. The fourth is the Buddhist tradition in which many of the ideas Hume develops are anticipated and articulated in somewhat different ways. I hope that these perspectives do not collide too violently. This way of reading of the Treatise may strike some as odd. I do not present Hume as a radical empiricist or as an atomist, and certainly not as a normative epistemologist or a metaphysician. I present him as a naturalist, as a skeptic, and as, above all, a communitarian, and on the reading I present, these are not only consistent orientations, but mutually supportive. I offer this reading as a plausible interpretation of the Treatise as a text. In offering this interpretation, I do not take myself to be divining what was in Hume’s mind as he composed it; but nor do I take myself to be offering

( xii )  Preface

ideas that simply occur to me as I  read it. Instead, I  hope that I  provide a compelling understanding of the text as a whole in the context of the literature to which it responded and in the context of the literature it inspired; I am locating it hermeneutically in the history of philosophy, and I am doing so in order to make it maximally informative and interesting to contemporary readers. I read the Treatise because I learn from it, and I hope that my reading can enable others to do so as well. In rereading the Treatise in this way, I  hope to exhibit it as a self-​ contained organic whole. Nonetheless, there are certain horizons external to the text that will play a role in my reading. First, there is the Humean horizon. I will sometimes advert to other texts by Hume in order to clarify or even to justify particular interpretative claims. Hume did not change his mind about all that much in his philosophical life, and often one can see what is going on in the Treatise better by looking at a subsequent formulation or even a retraction. So I sometimes advert to the History, and I will sometimes refer to the Enquiries, the Dissertation, and some of the essays and correspondence. Second, I will sometimes consider those who were influential on Hume to trace origins of ideas that might bring those ideas into sharper relief. Prominent among those I  take to be relevant in this context are Sextus Empiricus, Pierre Bayle, the Earl of Shaftesbury, Bernard Mandeville, and Frances Hutcheson. Hume read them all with care.1 While it would be wrong to say that he agrees with any of them on all points, often a Humean formulation or insight has a history, and we can sometimes appreciate his point better by reflecting on that history. And we have seen that this textual horizon includes the law as well. Third, there is a vast recent secondary literature on Hume, a mass of recent and contemporary Hume scholarship. I  am neither surveying nor responding to all of it. That would take the book too far afield. I will only refer to the literature that has most informed my own reading. Some recent commentators have developed the ideas I defend; with others I disagree. I will note the ways in which my reading engages with that literature, but without any pretense to comprehensiveness. Except where I  rely on them or respond directly to them with regard to the principal argument, I have kept discussions of others’ readings of Hume and related matters in the footnotes. Those who wish to remain aloof from these interpretative

1. Hume is also, of course, often responding directly to Locke. But enough has already been said about that response.

Preface  ( xiii )

debates can safely ignore those notes. In the end, this is a book about the Treatise and not about the literature on the Treatise. Over the years, I have had the opportunity to work with many colleagues who taught me more about Hume, and I have had the opportunity to teach Hume to many students, who also taught me more about Hume. I note in particular the late Chris Witherspoon, with whom I taught at Hampshire College; John Colman, my colleague in Tasmania; the late Barry Smith, with whom I taught at Smith College; Hsueh Qu, with whom I worked at the National University of Singapore; Cathay Liu, my colleague at Yale-​NUS College; and my longtime friend and colleague, stretching back to Oberlin and then to Pittsburgh days, Don Baxter. Among students, I especially wish to acknowledge that I have learned a lot from Angela Coventry in Tasmania; Kathryn Lindeman, Constance Kassor, and Lilly Frank in the Five Colleges; and Rachel Ong (to whom I owe those two wonderful sentences contrasting Hume and Descartes), Sai Ying Ng, and Elisabeth Tai in Singapore. I am very grateful to my friend and colleague Yasuo Deguchi, to the Graduate Faculty of Letters at the University of Kyoto, and to the Japan Society for the Promotion of Science for a JSPS visiting research professorship at Kyoto in 2017 that enabled me to teach a graduate seminar on the Treatise with a very capable and enthusiastic band of students. In that seminar, Yuki Koizumi, Yuki Kawasaki, Hu Chi-​chiang, James Fyfe, Taro Okamura, Kazunori Sawada, Riho Sato, Masumi Aoki, Shanshan Cao, Takuro Onishi, and Hsui-​mei Chen read and commented on an earlier draft of this book. Dr. Miguel Alvarez Ortega kindly sat in on this seminar and pushed me hard on questions in the history and philosophy of law, leading to much greater clarity about these matters than I had achieved before. His generosity and forcefulness contributed enormously to the seminar as a whole. This seminar was an experience of pure pleasure, and the participants’ comments and critique contributed a great deal to this book. That is what is so wonderful about teaching—​not what we have the chance to impart, but what we learn from those who think we are teaching them. This work has also benefited from the assistance of a trio of able and enthusiastic research assistants at Smith College. Thanks to Emma Taussig, Halley Haruta, and You Jeen Ha. It is hard to overstate the contributions these dedicated student assistants have made to this book; they have debated and clarified interpretations; they have made valuable suggestions regarding organization; they have helped with references, proofreading, and with so much more. I  could not have done this work without them. Thanks also to Katherine Fox, Riley Mayes, and Savitha Ravi for helping with the final editorial process.

( xiv )  Preface

I thank Rahul Govind of Delhi University for directing my attention to the literature on custom in the history of English law. I had not been aware of this important context for Hume’s thought before our conversations and before reading his book (Govind 2015), and now that I see that context, I also see that it is impossible to read Hume without it. We philosophers can learn a good deal by listening to our colleagues in history. And thanks to Christen Mucher for showing me just how pervasive this understanding of custom was in the context in which Hume was writing. Several colleagues generously read and commented on earlier drafts. In particular I thank the late Lynne Rudder Baker, whose collegiality, friendship, and contributions to my work over four decades I  will sorely miss. I also thank Don Baxter, John Connolly, Bill de Vries, Douglas Duckworth, Maria Heim, John Powers, Hsueh Qu, Kazunori Sawada, Tom Wartenberg, and Jan Westerhoff for critical comments. Don Baxter’s, John Connolly’s, Bill deVries’s, and Tom Wartenberg’s careful reading and probing questions pushed me to much greater clarity; Hsueh Qu challenged my readings and saved me from a number of errors. Kazunori Sawada’s relentless critique revealed important errors and omission in earlier drafts, and showed me how to sharpen some crucial points. Douglas Duckworth and Maria Heim offered sage advice on how to make the book more accessible to readers in Buddhist studies. Two anonymous readers for Oxford University Press offered many suggestions that have improved this book. While I  do not agree with any of these folks about all matters, each of them has changed my mind about some things, and forced me to better defense of controversial points. In sum, this book is much better for the assistance of my generous colleagues, and all defects that remain are my responsibility alone. Finally, I express my gratitude to my editor at Oxford, Peter Ohlin, for all of his help and support for this book as well as for his work on the others I have published with him.

PA R T  I

vwv Methodological Preliminaries

C H A P T E R  1

w

Introduction Principles of Interpretation

I

present in this book a reading of the Treatise of Human Nature. It is neither entirely original—​there is far too much Hume scholarship for anything to be entirely original—​nor obviously correct. I will diverge in many respects from some “mainstream” Hume scholarship, and in any case, the mainstream of Hume scholarship is a highly contested set of eddies, not a smoothly flowing current. I will follow many of those eddies, agreeing with those like Kemp Smith that Hume must be read as a kind of naturalist, but also with Popkin and with Fogelin that he must also be read as a skeptic (Kemp Smith 1905; Bayle 1965; Popkin 1980; Fogelin 1985, 2009). For the most part, however, I follow in the tradition of Annette Baier (1994, 1997, 2010a), who argues that Hume is first and foremost a communitarian (although she does not use this term). I will show that these are not competing interpretations of the Treatise, but rather are interdependent aspects of any cogent reading of the text as a whole, that by taking Book II as foundational, we come to a reading that reconciles Hume’s skepticism and his naturalism, and that the key to this reconciliation is his communitarianism. I make no claim for originality; I only present the Treatise as I read it now, in the hope that it is a reading of sufficient interest to move discussions of that text and its importance forward. This reading has evolved over about four decades of fascination with the Treatise and by my reading of the Pyrrhonian skeptical tradition and of Madhyamaka Buddhist philosophy, especially the work of Candrakīrti. Hume was certainly not a Buddhist; nor was he aware of Buddhist ideas.

( 3 )

( 4 )  Methodological Preliminaries

Nonetheless, Hume -​arrives at arguments and positions that are strikingly resonant with Buddhist arguments and positions, and this may be no accident given the probability of interaction between Greek and Indian skeptics.1 Some of the interpretations I present of particular arguments or theses might appear odd, but taken together, I believe they offer a cogent reading of the Treatise that is both historically plausible and philosophically sound. I do not pretend to vindicate every argument and thesis advanced in the Treatise; nor do I pretend to correct every error. Either would be a fool’s errand. But I do intend to defend what I take to be its central insights and method. There is a reason that the Treatise is still read, and that is because it has a lot to teach us. In this introductory chapter I set out five methodological principles that guide my reading of Hume. Each is defeasible, but each has a lot to say for itself, and together they inform my engagement with the text. They are as follows: . The Cover Principle 1 2. The Skeptical Inversion Principle 3. The Centrality of Custom Principle 4. The Pseudo-​idea Principle 5. The Principle of the Uniformity of Method

1.  THE COVER PRINCIPLE The Cover Principle is simple and, I think, obvious. It is even one Hume himself implicitly advises his reader to adopt, when he says that his aim is to “introduce experimental philosophy to moral subjects” [introduction 6, xvi]. Nonetheless, it is ignored repeatedly by eminent and otherwise sensible Hume scholars. The Cover Principle

If you are unsure about what Hume is doing, close the book and read the cover. The Cover Principle has three things going for it: first, it takes Hume at his word, and that is a good starting place in reading; second, it sets Hume’s text more plausibly in its historical context, reflecting his debts to Berkeley, Mandeville, and Malebranche, as well as his regard for Newton, and his 1. See Beckwith (2015), McEvilley (2012), and Kuzminski (2008) for more on the history between Greek and Indian philosophical interaction.

Introduction 

( 5 )

response to Locke; finally, if I am right, it will make the Treatise a much more interesting and less fallacious text than it is often taken to be. The proof, of course, as in the case of each of these principles of interpretation, is in the haggis. On the cover, you will find the words “A Treatise of Human Nature.” That tells you what the book is about. It is not about causality; it is not about whether there is a self; it is not about the metaphysics of morals or the existence of the external world. It is about human nature, that is, about psychology, broadly speaking. I will argue that Hume is (almost) always exploring the workings of the mind, including the place of the mind in culture, and the role of culture and language in shaping our cognitive and affective processes as well as our conceptual apparatus. If we follow the Cover Principle in our interpretation of the Treatise, we will find that when Hume approaches epistemological questions, he does so in the framework we today call “naturalized epistemology,” asking not what justification is, or whether we can know anything, or what the foundations of knowledge are, but rather how organisms like us come to be able to justify and to rely on claims, and what the social practices are that make such justification possible. We will find that when Hume approaches questions of metaphysics, he is almost always asking questions not about the fundamental nature of reality, but rather about our own conceptions about the nature of reality, and about the contents of the concepts we have, as well as those we falsely believe ourselves to have (more of this anon).2

2. This puts me at odds with Wright (1983), Pears (1990), and Strawson (1989), each of whom defends a realistic reading of Hume, and takes him to be pursuing a metaphysical and epistemological agenda, as opposed to a psychological program. I  will consider their views in more detail in ­chapters 7–​10. Baxter (2008) reads Hume as “a great metaphysician” (6). I agree with Baxter quite a lot in matters of detail, but not with regard to the big picture. I will be arguing that Hume is simply not a metaphysician at all. Part of our disagreement may stem from where we pick up the tangle of the Treatise. I start with the passions, and work my way out. Baxter starts with the account of time and identity and works from there. And we each give relatively short shrift to parts of the Treatise that the other takes as central. Nonetheless, I think that I can accommodate a reading even of those more apparently metaphysical sections as accounts of our own concepts, as opposed to metaphysical accounts of the nature of reality; Baxter has a harder time explaining why there is so much psychology in a treatise on metaphysics, and must explain away a lot of the structure of the arguments that I will explore. I agree with Schmitt (2014) when he says at the beginning of a masterful defense of a veritistic epistemological reading of Book I of the Treatise that “David Hume’s A Treatise of Human Nature manifests scepticism, empiricism, and naturalism in epistemology” (1). We each strive to reconcile these strands of Hume’s theory. I disagree with him, however, in seeing the Treatise as governed by an epistemological agenda. I think that Schmitt’s heroic defense of his reading of Hume would look different were he to

( 6 )  Methodological Preliminaries

We will find that when Hume addresses questions about ethics, he is not asking about the difference between right and wrong or about our duties or virtues as they are in themselves. We will find instead, if we keep the Cover Principle in mind, that Hume is asking about the nature of our moral reasoning and moral practices—​how, as J. L. Mackie was to put it a few centuries later, we “invent right and wrong.”3 In short, the Cover Principle invites us to take Hume at his word, and to read the Treatise as an early text in cognitive science, neither as an exercise in metaphysics such as Berkeley undertakes in the Principles and Dialogues, nor as an anticipation of Kant’s transcendental philosophy.4 Hume puts the point nicely in a way that also anticipates the interaction of this principle with his skepticism. He asserts near the end of Book I that his aim in the Treatise is “to establish a system of set of opinions, which if not true (for that, perhaps, is too much to be hop’d for) might at least be satisfactory to the human mind, and might stand the test of the most critical examination” [1.4.7.14, 272].5 Hume makes it plain here that his examination is not an a priori account of how the world or even the mind must be; instead, it is a fallible, revisable, but acceptable scientific theory that can move our understanding of our own nature along until the next theory (for which we might hope) comes along.6

attend to the Treatise as a whole, and not focus entirely on Book I. Kemp Smith (1905) offers perhaps the most compelling account of Hume’s naturalistic commitments. 3.  There may be one exception to this general principle, and this accounts for the “almost” in my explanation of the Cover Principle. As Baxter has urged (personal communication), 1.2.2 (Of the Infinite Divisibility of Space and Time) can plausibly be read as metaphysical. (And, I might add, it contains some of Hume’s most questionable claims, and those most at odds with his general naturalism about the mind.) But I take it that, like Hume’s missing shade of blue, “the instance is so particular and singular, that ’tis scarce worth our observing, and does not merit that for it alone we should alter our general maxim” [1.1.1, 6]. But even this section may be more an apparent than a real counterexample: it can also be read not so much as metaphysics as a cognitive psychology of our perception of external objects, as suggested by Hume’s remark at 1.2.5.26 that his account is meant to describe not the “real nature and operations” of external objects, but rather their “appearances.” 4. See also Hume (1978), Fogelin (1985, 1ff.), Biro (1993), Broughton (2004), and Loeb (2011a) for similar sentiments. Ainslie (2015) agrees, and also sees this naturalism as consistent with Hume’s skepticism. 5. All references to the Treatise are to the Selby-​Bigge and Nidditch second edition (Hume 1978) with pagination and reference to sections in square brackets referring directly to that edition; all references to the Enquiries are to Selby-​Bigge and Nidditch, third edition (Hume 1975, abbreviated EHU or EPM), which reprints the original 1777 edition. References to the Dissertation on the Passions are from Beauchamp’s edition (Hume 2007). 6.  In a letter to George Cheyne in 1734, Hume writes of his enthusiasm for this naturalism:

Introduction 

( 7 )

As Wright (1983) correctly notes, the particular scientific account of the mind that Hume adumbrates is hardly original with him. Hume’s broad conception of what science looks like owes a great deal to Newton, both with regard to Hume’s quest for a general set of laws descriptive of the psychological realm instead of an account of free agents acting on motives, and with regard to his insistence on the occult character of the true natures of the fundamental forces that activate the mind, just like the occult character Newton ascribed to the fundamental physical forces.7 Penelhum (2000c) puts the same point very nicely: To say that there can be a science of the mental, as Hume sees the matter, is to say that what we think, feel, or will can be explained as the effect of a cause and

Every one, who is acquainted either with the Philosophers or Critics, knows there is nothing yet establisht in either of these two Sciences, & that they contain little more than endless Disputes, even in the most fundamental Articles. Upon Examination of these, I found a certain Boldness of Temper, growing in me, which was not enclin’d to submit to any Authority in these Subjects, but led me to seek out some new Medium, by which Truth might be established. After much Study, & Reflection on this, at last, when I was about 18 Years of Age, there seem’d to be open’d up to me a new Scene of Thought, which transported me beyond Measure, & made me, with an Ardor natural to young men, throw up every other Pleasure or Business to apply entirely to it. (Letter number 3 to George Cheyne, March or April 1734, in Hume 1932, vol. 1, 13). (But note that Klibansky and Mossner [2017] take this letter to be addressed to Arbuthnot.) Whelan (1985) comments on Hume’s naturalism in terms that reflect these commitments: The term naturalism used in connection with Hume’s philosophy refers, in the first place, to the view of human nature as an integrated whole that Hume develops following his discovery that key problems of epistemology are insoluble within the limited framework of the theory of perceptions. Hume’s psychological naturalism is grounded in his conclusion that cognitive problems are overcome in practice through the operation of nonrational, natural mental processes whose interaction with rational ones is continuous, inescapable, and on the whole benign. Further, this naturalism involves a tendency to regard human nature so delineated as part of nature in a broader sense, in which all normal organisms are understood as being well suited, in their capacities, to their environments and coexistent in a general system of fundamental harmony. (67–​68) 7. Although as Wright notes (1983), it is to Copernicus that Hume explicitly compares himself in his exposition of the methodology of the discussion of the passions (2.1.3.6–​ 7, 282). Hume also makes the Copernican comparison in the first Enquiry (EHU 1.14, 14). But note that in this section he also argues, in distinctly Newtonian language, that a science of the mind should determine the laws governing its operation. And, we might also note, the idea of philosophy as an anatomy of the mind (as well as the centrality of the passions to its constitution) is also suggested by Shaftesbury: “The Parts and Proportions of the Mind, their mutual Relations and Dependency, the Connexion and Frame of those Passions which constitute the Soule or Temper, may easily be understood by any-​one who thinks it worth his while to study this inward Anatomy”

( 8 )  Methodological Preliminaries the instance of a natural law. Human minds are not strangers in nature, but are inextricably parts of it. Hume tries to demonstrate this in detail in the Treatise by showing how our beliefs and our emotive and conative commitments arise. (131)

Hume’s specific speculations about the form of psychological explanation, and the quasi-​hydraulic model he presumes regarding the operation of the animal spirits in the nervous system, on the other hand, follow Malebranche and Mandeville very closely. Indeed, Hume commends Mandeville for introducing the scientific method into the study of human thought.8 Nonetheless, Hume’s own naturalism is not entirely derivative of those of his predecessors.9 In particular, reacting against Descartes’s view that the mind knows itself immediately and with certainty—​a position tacitly accepted even by such critics of other aspects of the Cartesian philosophy as Mandeville and Malebranche—​Hume asserts in the introduction to the Treatise, [T]‌o me it seems evident, that the essence of mind being equally unknown to us with that of external bodies, it must be equally impossible to form any notion of its powers and qualities otherwise than from exact experiments, and the observation of those particular effects, which result from its different circumstances and situations. [Introduction 8, xvii]

(1964, 2:48). Hume also employs the anatomical metaphor in the Enquiry (EHU 1.8; SBN 10). 8. On the other hand, as Alanen (2005; 2006, 20) notes, the psychological states and processes to be explained and whose relations are to be characterized by the science of human nature Hume envisions are not merely brute feelings; they are intentional states whose content is relevant to the generalizations to be discovered. And we will see, as we proceed, that content is always relevant to Hume’s accounts. So, while the form of explanation Hume envisions is nomic, the descriptions under which cognitive states are to be explained is intentional. If we think of Hume as an intellectual ancestor of the twentieth-​century behaviorist movement, it is hence Tolman, not Hull, Skinner, or Watson, of whom we should think as his descendant, in virtue of his defense of a “cognitive behaviorism” that took mental content seriously as causally relevant in the explanation of behavior. We will return to this point in ­chapter 3. Despite his professed commitment to the experimental method, Hume does not—​ as do contemporary experimental philosophers—​actually perform or report on actual psychological experiments. Instead, he often relies on “thought experiments.” But this shouldn’t be seen as bad faith in any sense. There was no empirical science of psychology in the eighteenth century, and thought experiments are not foreign to empirical science—​indeed they are common in physics, and, as Hume argues in the introduction to the Treatise, experiments conducted on our own minds only reveal our minds as they are when we are conducting experiments, not as they are in their more natural state. 9.  Here I  use the term naturalism to indicate a commitment to the explanation of phenomena by natural science. This is not, for instance, the way Kemp Smith uses the

Introduction 

( 9 )

Hume, that is, like Tsongkhapa,10 takes seriously the opacity of the mind to itself, the possibility of cognitive illusion, and the fact that neither the operation nor the contents of the mind can be taken as given. This denial of privilege to our knowledge of the mind itself is what animates Hume’s particularly acute Pyrrhonian skepticism, to which we now turn.11

2.  THE SKEPTICAL INVERSION PRINCIPLE Qu notes in the first sentence of (2015b) that “One of the most important topics in Hume scholarship today is determining the relation of Hume’s naturalism to his scepticism” (1). I  agree, and while my understanding of that relation is not exactly Qu’s, I  agree with him that any satisfactory reading of the Treatise must reconcile these two trends in Hume’s thought. This second principle may strike most readers as less plausible than the Cover Principle, but in what follows we will see that it will illuminate a great deal of what is otherwise either obscure or implausible in the Treatise. The Skeptical Inversion Principle

Hume follows the Pyrrhonian skeptics in resolving skeptical dilemmas about the justification of convention by arguing that convention is foundational to ontology, not the other way around. This one requires some explanation, and reasonable people might disagree with me either regarding my reading of classical Pyrrhonian skepticism or about the degree to which Hume—​his own protestations to the contrary, and his own somewhat different reading of the Pyrrhonian tradition notwithstanding—​follows that tradition methodologically. Once again,

term. While many take Hume’s skepticism and naturalism to be in tension with one another, part of the burden of this book is to show that when each is properly understood, they are in fact mutually reinforcing, 10.  Tsongkhapa was a fourteenth-​to fifteenth-​century Tibetan philosopher who argued that our self-​knowledge is always conceptually mediated and therefore fallible. For an account of Tsongkhapa’s philosophy of mind, see Garfield (2012). 11. Hume’s recognition of the opacity of the mind is limited in its scope, applying principally to ideas, as opposed to impressions (although it may issue in false beliefs about the presence of impressions antecedent to ideas we take ourselves to have). In particular, as Qu (2015a) points out, Hume insists that the mind is, in general, transparent with respect to the qualitative character of our present impressions (though not always: at 2.3.3.8, 417 he reminds us that we can sometimes mistake calm passions for reason).

( 10 )  Methodological Preliminaries

I ask the reader to be patient and see whether this bears fruit. First a bit of explanation is in order.12 Here is how I  understand the Pyrrhonian tradition, following Sextus Empiricus, as embodying a very particular philosophical strategy, one that we also find in the Madhyamaka philosophical tradition in India: The skeptic confronts a dogmatic dispute, involving two extreme positions, regarding the degree to which a convention, a custom, or a mode of speech or thought is justified. Let us call one side the reificationist position and the other the nihilist. One party to such a debate—​the reificationist—​argues that the convention, custom, or mode of speech or thought is justified, because it is grounded appropriately in a convention-​independent reality. The nihilist, on the other hand, argues that because the convention, custom, or mode of speech or thought cannot be grounded in such extraconventional facts, it is not justified. So, for instance, we can imagine a debate between someone who believes that we are justified in saying that there is an external world and someone who believes that we are not. The reificationist in this case (for example, Reid) argues that since there is an external world, we are justified in talking about external objects, and our concepts and words are adequate to them. The nihilist (for example, Berkeley) argues that all that we ever experience are our inner experiences, and that we have no direct access to anything external. Therefore, he argues, we are never justified in talking about material objects. In the domain of ethics, to take a second example, we can imagine an argument between a reificationist (say Bentham or Mill) who argues that because there are extramoral facts (e.g., about utility) that ground it, our moral discourse about rights, duties, or right and wrong is justified. The nihilist (say Nietzsche) argues instead that since there are no such extramoral facts that by themselves make moral claims on us, our moral discourse is unjustified. In any such case, the skeptic responds with epochē, usually translated as suspension of judgment. There are, however, various ways to understand such epochē. In Garfield (1990) I  argued for a particular reading of this strategy, which I  simply offer here without additional argument. I begin by discussing what epochē is not. One might suppose that epochē

12.  For more detail regarding my reading of Pyrrhonism, see Garfield (1990). See also Baxter (2006) for a very strong argument for reading Hume as a Pyrrhonist. Although our respective accounts of Pyrrhonian method differ slightly, we agree (each following Mates (1996) and Popkin (1980)) that Hume—​despite his protestations to the contrary—​is a good Pyrrhonian, and not an Academic.

Introduction 

( 11 )

is an attempt to find a compromise, some middle ground between the two positions: maybe the external world exists in some attenuated sense; maybe some of our moral discourse is justifiable, but not other parts. This is not epochē in the Pyrrhonian sense.13 Alternatively, one might suggest that epochē consists in a kind of shrugging of the philosophical shoulders. There are, one might argue, good—​but perhaps nondemonstrative—​arguments on each side of the issue at stake in the dispute, and therefore equally good and equally bad reasons for each extreme position. So, on this account, one should say in such a circumstance that it is impossible to decide which to adopt. The skeptic, on this reading, simply refuses to endorse either of the two opposing positions, saying, “Maybe one, maybe the other.” This is a common interpretation of skepticism, and it gains some aid and comfort from certain remarks of Sextus Empiricus. Moreover, Hume himself sometimes talks about skepticism in this way, as do many of his commentators, including many who regard him as a skeptic as well as many who do not. But just as compromise is not the Pyrrhonian agogē (way of life), nor is shoulder-​shrugging.14 Instead, the true Pyrrhonian—​including, as I will argue, Hume—​adopts a more radical epochē than either of these. The Pyrrhonian rejects the entire debate between the reificationist and the nihilist as ill-​conceived. For, the Pyrrhonian observes, however much the two dogmatic opponents might appear to disagree, they agree about the only interesting thing, the biconditional presupposition that undergirds the debate. In the case of the debate about the existence of the external world, that is, the assertion that our discourse about, or conventions regarding, external objects are justified iff there is an independent external world; in the case of the debate about ethics, it is the assertion that our moral discourse or conventions are justified iff there are discourse-​or convention-​independent moral facts to ground them. Call biconditionals of this sort grounding biconditionals. The Pyrrhonian epochē consists in the denial of the grounding biconditional, and a consequent inversion of the direction of explanation. Instead of grounding 13.  Nor, for that matter, is this the “middle path” of Madhyamaka, as I  argue in Garfield (1995, 2015). 14. Here I take issue with Garrett (1997) when he takes Hume to understand by skeptical arguments any arguments that “in some way concern or tend to produce doubt and uncertainty” (208). Nonetheless, I agree with Garrett that in the end Hume’s skepticism “can, indeed, reconcile his aim for a positive system of the sciences based on human cognitive psychology with his use of skeptical arguments, and that he does so in a way that facilitates an improved, if also chastened, commitment to the historically developing products of human reason” (208). I think that this reconciliation emerges from Hume’s consistent Pyrrhonism.

( 12 )  Methodological Preliminaries

conventions in ontology, the skeptic places our conventions at bedrock, and argues that our ontology, morality, epistemology, etc., simply rest on conventions.15 This is the point of Sextus’s fourfold prescription, viz., that the skeptic lives by appearances, inclinations, the instructions of the arts, and the conventions of her society. Sextus puts the point this way in The Outlines of Pyrrhonism: Now, we say that the criterion of the Sceptic discipline is the appearance, and it is virtually the sense-​presentation to which we give this name, for this is dependent on feeling and involuntary affection and hence is not subject to question. . . . Now, we cannot be entirely inactive when it comes to the observances of everyday life. Therefore, while living undogmatically, we pay due regard to appearances. This observance of the requirements of daily life seems to be fourfold, with the following particular heads: the guidance of nature, the compulsion of the feelings, the tradition of laws and customs, and the instruction of the arts. . . . And it is by virtue of the instruction of the arts that we are not inactive in those arts which we employ. All the statements, however, we make without prejudice. (1985, 40)

This passage in the Outlines introduces several themes that will be important in Hume’s Treatise. First, Sextus argues that much of what drives our cognitive and behavioral life is that which is dependent on feeling and which is involuntary. He then argues that these involuntary feelings issue in our taking things for granted, things we do not question, not because we have reasons, but because we have no choice in the matter. In Hume’s hands, these feelings and affections will become the passions, and we will

15.  Compare Hallie’s discussion of a debate regarding the Eucharist in Sextus Empiricus (1964). As Mates (1996) puts it, “Pyrrhonism is not a doctrine. . . . Sextus calls it an agogē, a way of life, or perhaps better, a way of thinking and acting. . . . The Pyrrhonian skeptic, instead of basing his thoughts and actions on firm beliefs about how things really are in a mind-​independent external world, ‘goes by the appearances’ ” (7). Much has been made about the importance of the appearances (phainomena) in Pyrrhonism, but it is equally important to understand the role of convention and custom as a guide to action, an issue to which we turn shortly. Also see Baxter (2006, 200; 2008, 8–​14) and Popkin (1980b) for a nice discussion of Hume’s Pyrrhonism. Sawada (personal communication) points out that this account opens up a regress, with a metaskeptical position possible that suspends the debate between the skeptic and the dogmatists regarding the grounding biconditional. That is right, but that is not a vicious regress. The denial of the biconditional at any stage of the skeptical dialectic is consistent with and underwrites the suspension of the debate under consideration at that stage, even if it raises the possibility of a new skeptical debate regarding the biconditional in another skeptical context.

Introduction 

( 13 )

see that the passions become the unquestioned drivers of much of our cognitive and affective life. More importantly, we see Sextus here emphasizing not only the role of our biological or perceptual affections, but also the social dimension of our lives as skeptics. In this context, Sextus emphasizes the importance of laws, customs, the instructions of the arts, and in gen­ eral of the social practices and conventions that constitute our societies in determining the ways in which we not only behave but also reason. This positive side of the Pyrrhonian tradition, we will see, plays as great a role in Hume’s own skepticism as does the negative side. Hume was introduced to Pyrrhonian skepticism in part through his reading of Montaigne, but primarily through the article on Pyrrho in Bayle’s Historical and Critical Dictionary ([1696] 1965), an article that played a major role in the revival in France and England of interest in the classical skeptical tradition.16 And we can see a great deal of Bayle’s influence in Hume’s skeptical method. For instance, Hume’s confident simultaneous deployment of skepticism and the methods of science—​which might seem prima facie at odds with one another (as well as his devastating deployment of the skeptical method in his critique of religious dogma in the Dialogues)—​reflects Bayle’s remark that Pyrrhonism is dangerous in relation to this divine science, but it hardly seems so with regard to the natural sciences or to the state. It does not matter much if one says that the mind of man is too limited to discover anything concerning natural truths, concerning the causes producing heat, cold, the tides, and the like. It is enough for us that we employ ourselves in looking for probable hypotheses and collecting data. I am quite sure that there are very few good scientists of this century who are not convinced that nature is an impenetrable abyss and that its springs are known only to Him who made and directs them. (1965, 194–​95)

We see several ideas in this passage that animate Hume’s own project, beyond the distinction between the import of skepticism for religion and 16. Laird (1932) concurs that Hume faithfully follows the Pyrrhonians as they come down to him through Bayle and offers a detailed explanation of Bayle’s views and their reflection in Hume’s work. But I think that Fogelin (1985, 1) is wrong when he states that “the degree of Hume’s skepticism is variable, . . . His general posture is that of a moderate skeptic, recommending that we modestly restrict our inquiries to topics within our ken” (2). Instead I will argue that he is a thoroughgoing Pyrrhonian. Baxter (1993, 1) ably defends the thesis that Hume is a true Pyrrhonist, acknowledging that Hume himself disavows that label in virtue of a misrepresentation of the Pyrrhonian doctrine, and demonstrates convincingly how Hume’s skepticism frames and makes good sense of his much-​maligned account of our ideas of space and time, and that the account makes no sense whatsoever unless we read Hume as a Pyrrhonian.

( 14 )  Methodological Preliminaries

science. First, Bayle asserts—​and Hume will follow him in this—​that it is perfectly permissible for a Pyrrhonian skeptic to advance “probable hypotheses” and to collect data.17 This, I will argue, is exactly what Hume takes himself to be doing with regard to human psychology in the Treatise. But second, and perhaps more importantly, Bayle emphasizes that in doing so, one is not uncovering the “springs” of nature, but only advancing hypotheses with some descriptive or explanatory force. When Hume turns to topics such as causality, we will see this principle very much at work. Bayle then turns to the positive side of Pyrrhonism more specifically: Society has no reason to be afraid of skepticism; for skeptics do not deny that one should conform to the customs of one’s country, practice moral duties, and act upon matters on the basis of probabilities without waiting for certainty. They could suspend judgment on the question of whether such and such an obligation is naturally and absolutely legitimate; but they did not suspend judgment on the question of whether it ought to be fulfilled on such and such occasions. (1965, 195)

Hume takes this positive side, encoded in Sextus’s fourfold prescription, very seriously as well. Indeed, as I will be arguing throughout this book, his expansive understanding of custom comprises not only collective, social customs, but also the individual habits of mind and practice that a classical Pyrrhonian would have characterized as instinct or appetite.18

17. In this context, see Garrett (1997) for an excellent discussion of Hume’s willingness to advance inductive arguments in the Treatise and to advocate an empirical science of human nature despite his refutation of any rational or probable justification of induction. For connections between Hume’s Pyrrhonism and his atheism, see Russell (2008). 18.  And Hume retains this understanding of skepticism and his allegiance to it throughout his career. In the Enquiry, introducing his “sceptical solution” to the doubts he raises about theoretical and empirical reasoning, he writes (using the term “academic,” but really referring to Pyrrhonian skepticism): The academics always talk of doubt and suspense of judgment, . . . and of renouncing all speculations which lie not within the limits of common life and practice. Nothing, therefore, can be more contrary than such a philosophy to the supine indolence of the mind, its rash arrogance, its lofty pretensions, and its superstitious credulity. Every passion is mortified by it except for the love of truth.  .  .  . It is surprising, therefore, that this philosophy, which, in almost every instance, must be harmless and innocent, should be the subject of so much groundless reproach and obloquy. . . . Nor need we fear that this philosophy, while it endeavours to limit our enquiries to common life, should ever undermine the reasonings of common life, and carry its doubts so far as to destroy all action, as well as speculation. Nature will always maintain her rights, and prevail in the end over any abstract reasoning whatsoever. (EHU 5.1; SBN 41)

Introduction 

( 15 )

Hume, like the Pyrrhonians as (correctly) understood by Bayle, argues that custom constitutes the explanatory foundation of our lives, including our epistemic practices, verbal conventions, and moral judgments, not that custom-​independent certainty justifies our customs.19 It is for this reason that Pyrrhonian skepticism should neither be seen as a mere shrugging of the shoulders in ignorance or indecision, nor as the claim that we know nothing, but rather as a location of the conventional as the bedrock of our epistemic and moral lives. Finally, is worth noting that in the Outlines, Sextus closely links what we might call the biological dimensions of habit with the social dimensions. We will see Hume following him in this respect in his account of custom, which also straddles this boundary.20 While we must be careful about using the Enquiry to justify interpretations of the Treatise, here Hume provides us with clear evidence of his own understanding of skepticism, and of an understanding that is evidently at work in the Treatise. (See also the discussion of skepticism in EHU 4.21–​22 (SBN 36–​39), which recapitulates that of Treatise 1.4.3, providing further confirmation of the continuity of Hume’s thought in this regard.) It is also worth noting the careful distinction Hume draws between his own Pyrrhonian skepticism and the modern skepticism of Descartes, which he takes to be incurable if taken seriously (EHU 9.1; SBN 150). 19.  It is clear that Reid, for instance, misses this positive side of the Pyrrhonian method when he suggests that Hume’s positions on personal identity and the existence of the external world are inconsistent with his practice, saying: It seems to be a peculiar strain of humour in this author, to set out in his introduction, by promising, with a grave face, no less than a complete system of the sciences, upon a foundation entirely new—​to wit, that of human nature—​when the intention of the whole work is to shew, that there is neither human nature nor science in the world. It may perhaps be unreasonable to complain of this conduct in an author who neither believes his own existence nor that of his reader; and therefore could not mean to disappoint him, or to laugh at his credulity. Yet I cannot imagine that the author of the “Treatise of Human Nature” is so sceptical as to plead this apology. He believed, against his principles, that he should be read, and that he should retain his personal identity, till he reaped the honour and reputation justly due to his metaphysical acumen. Indeed, he ingenuously acknowledges, that it was only in solitude and retirement that he could yield any assent to his own philosophy; society, like daylight, dispelled the darkness and fogs of scepticism, and made him yield to the dominion of common sense. Nor did I ever hear him charged with doing anything, even in solitude, that argued such a degree of scepticism as his principles maintain. Surely if his friends apprehended this, they would have the charity never to leave him alone. (2000, 8–​9) 20. As will be apparent as I develop my account of Hume’s skepticism, I disagree with Garrett’s deployment of what he calls the “Title Principle” (not to be confused with my “Cover Principle”!): “Where reason is lively, and mixes itself with some propensity, it ought to be assented to. Where it does not, it can never have any title to operate upon us” (1.4.7.11, 270). The Title Principle, as I hope that my account will demonstrate, is not an autonomous epistemic principle; it is a mere summary of Hume’s broader Pyrrhonian commitments, which we will explore in greater detail throughout this study. Moreover, I will argue, Hume’s Pyrrhonism is at bottom psychological, not epistemological.

( 16 )  Methodological Preliminaries

Epochē in the classical Pyrrhonian sense is therefore a suspension of the debate between dogmatic positions in favor of a radically different kind of discourse. Skepticism of this kind, as Sextus emphasizes when he says that “we cannot be entirely inactive,” is not a purely negative project, but a positive project as well, made possible by the clearing of a certain kind of metaphysical underbrush. The positive project is the limning of the domain of human custom, of human nature.21 To be sure, there are passages in Sextus’s writings that suggest that he sees the Pyrrhonian position as a kind of epistemic nihilism, and that he endorses such a position. But it is hard to square that reading both with Sextus’s discussion of the fourfold prescription and with the reflexivity of skepticism that he emphasizes (an emphasis we will see echoed in Hume in his discussion of skepticism with regard to reason). It makes much more sense, I  think, to read those passages in which Sextus advocates a recusal from all belief, or denies that there is any knowl­ edge, as a rejection of dogmatic belief, or dogmatic knowledge, that is, of any doxastic attitude not tempered by the skeptical agogē. When Hume explicitly rejects skepticism, he rejects this nihilistic version; his own approach hews very close to Sextus’s actual program.22 Thus, I  will argue, Hume’s This emphasis on our social nature is a commitment Hume retains throughout his career, despite other changes in his views. In the first Enquiry, for instance, he remarks: Man is a reasonable being; and as such, receives from science his proper food and nourishment: But so narrow are the bounds of human understanding, that little satisfaction can be hoped for in this particular, either from the extent or security of his acquisitions. Man is a sociable, no less than a reasonable being: But neither can he always enjoy company agreeable and amusing, or preserve the proper relish for them. Man is also an active being; and from that disposition, as well as from the various necessities of human life, must submit to business and occupation: But the mind requires some relaxation, and cannot always support its bent to care and industry. It seems, then, that nature has pointed out a mixed kind of life as most suitable to human race. (EHU 1.6; SBN 8–​9) In this volume, while I will focus explicitly on the Treatise, it is Hume’s account of this “mixed kind of life” that will concern us. 21. This is not the only reading of Pyrrhonian skepticism. See Burnyeat (1983) for an alternative understanding. Baxter (1993) follows Frede (1997) in his treatment of Hume’s skepticism. My own reading is closer to those of Hallie (Sextus Empiricus 1985), Mates (1996), and Popkin (1980b) and I think explains Bayle’s and Hume’s use of skepticism better. While it is tempting to digress into a debate regarding interpretations of Sextus Empiricus, I leave that issue aside here, and merely note my own interpretative allegiance, while acknowledging that it is one among many. I hope that its merits are demonstrated by the reading of Hume it enables. 22. See Strawson (1985) and Kripke (1982) for insightful account not only of this method, but also of the connections between Hume and Wittgenstein, and Strawson in particular for an acute analysis of the connection between Hume’s skepticism and his naturalism. And read Qu (2015b) for the most searching examination of the relation of Hume’s naturalism and skepticism to his epistemology.

Introduction 

( 17 )

project is entirely continuous with the classical Pyrrhonian tradition as well as its later developments in the hands of Wittgenstein, whose ideas are so redolent of Hume’s.23 As we read the Treatise, we will see many examples of this inversion.24

3.  THE CENTRALITY OF CUSTOM PRINCIPLE All of this connects directly to the third principle: The Centrality of Custom Principle

When Hume turns to the explanation of cognitive capacities and achievements, the central explanans is custom, and this in two senses: individual customary behavior and social custom. The connection is clear: The second principle of interpretation leads us to see Hume as taking custom not as something to explain, but rather as that which explains; hence its central role in the Treatise. But its role is dual. Hume uses the term custom to designate what we would now regard as two phenomena, but which to Hume’s eyes appears to be a single phenomenon appearing in two distinct domains: custom as convention and custom as habit. First, there is social custom—​the way we do things, including conventions regarding language, individuation, explanation, praise and blame, etc. For Hume, custom in this sense is often explanatory bedrock. We will explore his understanding of custom in this sense, and its roots in English legal theory, in c­ hapter 2. Hume uses custom not only to denote social regularities, but also to denote individual psychological regularities, or our customary way of behaving. We might prefer to call this habit. So, for Hume, the fact that I customarily call to mind the color red when I think of apples is as much a matter of custom as the fact that we customarily set the table with the fork on the left and the 23. Some readers might be surprised by this remark given Wittgenstein’s oft-​quoted comment that he found reading Hume “a torture.” Maybe he did find Hume hard to read; but it is clear that he endured that “torture,” and, I think, clear that he learned a great deal from it. 24. See Frede (1997) for a similar account of classical skepticism (although he would not agree with my discussion of the “skeptical inversion”). Burnyeat, however, in a searching discussion of the structure of Pyrrhonism that focuses directly on Hume’s own explicit critique of that doctrine in the Treatise, takes a very different position, arguing that Pyrrhonism requires a complete recusal from all belief. In fact, Hume seems to share Burnyeat’s more radical reading of the Pyrrhonian position, which is why he refuses to take himself to be a Pyrrhonian.

( 18 )  Methodological Preliminaries

knife on the right. Custom, then, is regularity in behavior. Nonetheless, a close reading of the Treatise reveals that on Hume’s view, regularity in individual behavior is often grounded in social regularity (as one would expect given the legal background of this term), which in turn is norm-​constituting in virtue of the expectations it engenders, and lies at the basis of the explanatory power of nomic generalizations grounded in observations of regularity. This is why when we look to Hume on explanation, brute regularities, whether in individual or collective behavior, are often as deep as he takes us.25 Wittgenstein was to concur that in the citation of custom, in just this sense, “our spade is turned” (Wittgenstein 1953, ¶ 217). Hume returns to this idea in the Enquiry Concerning Human Understanding: Speaking of our natural tendencies to belief that constitute the positive side of the skeptical program, Hume writes, This principle is Custom or Habit. For wherever the repetition of any act or operation produces a propensity to renew the same act or operation, . . . we always say that this propensity is the effect of Custom. By employing that word, we pretend not to have given the ultimate reason of such a propensity. We only point out a principle of human nature. (EHU 5.5; SBN 43)

A bit later, Hume asserts, Custom, then, is the great guide of human life. It is that principle alone which renders our experience useful to us, and makes us expect, for the future, a similar train of events with those which have appeared in the past. Without the influence of custom, we should be entirely ignorant of every matter of fact beyond what is immediately present to the memory and senses. We should never know how to adjust means to ends, or to employ our natural powers in the production of any effect. There would be an end at once of all action, as well as of the chief part of speculation. (EHU 5.6; SBN 44–​45)

Hume’s skepticism, and his understanding of skepticism, may well have evolved between the time of the composition of the Treatise and that of the composition of the Enquiry. I will not be concerned here to trace that evolution. In any case, in these passages in the Enquiry, we see Hume still connecting his account of custom directly not only to his Pyrrhonian method

25. It is interesting to note that this same broad use of a single term to indicate both collective and individual regularities in behavior and thought is found in the use of the Sanskrit samvṛti (convention) and vyavahāra (ordinary, or everyday behavior), and enters Madhyamaka in a way prescient of Hume. See Whelan (1985).

Introduction 

( 19 )

generally, but also to his views about causality and action, connections that I will argue are already in place in the Treatise. Hume returns to this connection, and to the connection between custom and our brute, animal psychology—​to our cognitive instincts—​in EHU 5.11 (SBN 48) (as well as in EHU 9.5 [SBN 106]), when he argues that the customary nature of much of thought explains its speed, a speed we do not see in deliberate ratiocination (in EHU 5.22, 55).26 Baier has made this point with great force. Writing about the Enquiry Concerning Human Understanding, she says, “That principle [which enables inference] is said . . . to be custom or habit, and it is important to note that it is said to have authority, not simply to have causal influence” (1994, 80). Turning back to the Treatise, she observes that if anything is to have epistemic authority, if any step taken by the mind is to receive normative endorsement, it cannot fail to be some sort of instance of associative thinking. What gives it its authority will indeed be a special feature not found in any and every associative move. The rationalists, Hume believed, had misidentified that special feature. He is offering another way of understanding epistemic authority, one that allows us to give authority to some habits that are not habits of deductive argument and to establish them as rules. (80)

And elsewhere, commenting on the role of custom in its sense as social convention, she writes: There is no individualist bias to Hume’s epistemology within most of Book One—​ membership in a linguistic community, itself one of several such communities whose languages “nearly correspond” to each other, is essential for the sort of mental activity that he is analyzing in Part I of Book One, indeed in the whole of the Treatise. Only in Part IV of Book One, when scepticism and solipsism are worked though, is there any experiment made of separating the thinker from the normal human world. (1991, 32)

Baier is absolutely correct in observing that one of Hume’s most important contributions to philosophy is his articulation of the Pyrrhonian fourfold prescription in this way, seeing custom, in its guise both as mental habit and as social convention, not only as the cause of our cognitive practices, but as a source of the norms that govern them. 26. And, to press the legal analogy, the speed that allows us instantly to know whether a course of action accords with our customary practices, as opposed to the deliberate processes of a court of law.

( 20 )  Methodological Preliminaries

It is for this reason that Hume is the first serious naturalizer of epistemology. We will also see that this role of custom in the epistemic domain is entirely homologous with the role that Hume assigns it in the ethical and political domain. Hume, drawing in part on the Pyrrhonism he inherits from Bayle and Montaigne, in part on the analysis of our social nature that he inherits from Mandeville and Hutcheson, and in part on the rich traditions of British legal theory, develops the first complete Western philosophical system that takes custom seriously as a foundation of human life as an organic unity.27

27. As I argue elsewhere (Garfield and Priest 2002; Garfield 2002; Cowherds 2010; Garfield 2011; 2015, 20), Madhyamaka Buddhist philosophers such as Nāgārjuna and Candrakīrti made this move over a millennium earlier. The parallels between Hume’s use of custom or convention and that in Madhyamaka are fascinating. Gopnik (2009) has argued that Hume must have drawn explicitly on Buddhist ideas, on the grounds it could be no coincidence that these ideas are so similar, given that Hume was writing the Treatise at La Fleche at precisely the time that Ippolito Desideri and Charles François Dolu were in residence. Desideri had recently returned from Tibet, where he studied Buddhist philosophy and acquired a deep understanding of Madhyamaka. (Desideri 2010)  and Dolu had been part of a Jesuit embassy to Siam, where he had acquired some knowledge of Theravāda Buddhism. Gopnik has also pointed out (personal communication) that Bayle mentions Buddhist doctrines (as part of the “philosophy of the orient”) in a footnote to his essay on Spinoza, and Hume clearly read that essay. Hence, she concludes that it is likely that Hume knew about Buddhism and utilized Buddhist ideas in coming to his views about the self, causation, and other matters regarding which his views are strikingly similar to those of some Buddhist thinkers. I disagree with Gopnik’s interpretation of what is surely a fascinating historical coincidence (and given my own interest both in Buddhist philosophy and in the early interactions between European and Asian philosophers, I wish I could agree). First, is no positive evidence that Hume ever met either Desideri or Dolu, or read their work. And given Hume’s enormous antipathy to Christianity, if he had been able to extol the greater insight of a “pagan” tradition, he assuredly would have done so. Indeed, there is no evidence in any of Hume’s writings including not only the Treatise, but also the Natural History of Religion and the Dialogues Concerning Natural Religion, or in his correspondence that he had ever heard of Buddhism. And Hume was willing to write about his conversations with his Jesuit colleagues, and to remark on their curious ways. In a letter to George Campbell, he writes: It may perhaps amuse you to learn the first hint, which suggested to me that argument which you have so strenuously attacked. I was walking in the cloisters of the Jesuits’ College of La Flèche, a town in which I passed two years of my youth, and engaged in a conversation with a Jesuit of some parts and learning, who was relating to me, and urging some nonsensical miracle performed in their convent, when I was tempted to dispute against him; and as my head was full of the topics of my Treatise of Human Nature, which I was at the time composing, this argument immediately occurred to me, and I thought it very much graveled my companion; but at last he observed to me, that it was impossible for that argument to have any solidity, because it operated equally against the Gospel as the Catholic miracles;—​ which observation I thought proper to admit as a sufficient answer. I believe you will allow, that the freedom at least of this reasoning makes it somewhat extraordinary to have been the produce of a convent of Jesuits, tho perhaps you may think

Introduction 

( 21 )

Attention to this central role of custom, along with appreciation of Hume’s Pyrrhonism, explains why Laird is wrong to say that “Hume set out to lay the foundations of pure sensory phenomenalism” (1932, 25).28 Part of the burden of this book, as I indicated out the outset, is to show that Hume is not pursuing a primarily epistemological project (at least in any sense other than that of naturalized epistemology), and certainly not an ontological project in the Treatise. But it is equally important to note that when he does turn to the question of the nature and source of our knowledge, whether or not founded in sensation, it is always modulated by custom. Sensation might provide the starting point for cognition, but the normative dimensions of knowledge derive not from its sensory origins, but from our psychological and social epistemic customs. In short, Hume is too much an externalist to be any kind of phenomenalist. One might wonder why it is that custom plays such a central role in Hume’s philosophical program. As we will see in the next chapter, the origins of Hume’s thought about custom lie in debates about custom in the law, and about the relationship between customary and common law in England that occupied British legal theory in the eighteenth century. Hume’s understanding of normativity and necessity as grounded in custom directly reflects his understanding of the foundational role of custom in the law. the sophistry of it savours plainly of the place of its birth. (Letter 194 to George Campbell, 1762, in Hume 1932, 361). If he had met Desideri and heard tales of Tibet, or if arguments from exotic Tibet struck him as undermining Catholic orthodoxy, he would surely have remarked on it. Moreover, even though Hume did read Bayle’s entry on Spinoza, nothing resembling the material in the footnote on Buddhism appears in the Treatise. Hume clearly makes use of Bayle’s discussion of Spinoza in his own treatment of the Spinozistic philosophy [1.4.5.18–​26, 240–​44], but makes no use of the material on Buddhism, suggesting that if he read it, he did not take it to be important. I conclude that—​although Gopnik is correct to assert that Hume had the opportunity to learn about Buddhism at La Fleche, both from colleagues and from texts available in its library—​there is in fact no evidence that he actually knew anything of the Buddhist tradition, either Theravāda or Mahāyāna. But this does not mean that there is no connection, only that Gopnik has not found the right one. As McEvilley (1981, 2002)  and Beckwith (2015) have argued, the Pyrrhonian tradition and Madhyamaka probably evolved in conversation with one another, partly through Pyrrho’s documented visit to India in the entourage of Alexander, and partly through the mediation of the intellectual milieu in the Persian court, in which Indian and Greek philosophers each participated. Hume is picking up on the fruits of this mediation not through conversations with recently returned missionaries, but through his reading of Bayle and Montaigne. Sidelle (1989), although he mentions neither Hume nor the Madhyamaka tradition in his study, presents a spirited defense of conventionalism that is very much in a Humean or Mādhyamika vein. 28. Also see Passmore (1952) for a concurring opinion.

( 22 )  Methodological Preliminaries

4.  THE NOMINALIST ANALYSIS OF PSEUDO-​I DEAS One of the more remarkable features of Hume’s philosophical program, often overlooked, is that he often provides not an analysis of a concept he sets out to investigate, but rather the verdict that, despite our conviction to the contrary, we have no such concept, in virtue of having no corresponding impression. So, the fourth principle of interpretation is this: The Pseudo-​idea Principle

Assume that for most philosophically important terms, Hume will not provide an analysis of an idea, but a demonstration that no idea corresponds to a term that appears to make sense. So, for instance, rather than ask what Hume’s analysis is of the concept of a self, of causation, or of external existence, we will ask how Hume shows that we have no such ideas at all. We will then ask how Hume understands our use of the terms that appear to denote such ideas. This will take us back to the previous two principles of interpretation. Often it will turn out that while it appears that a custom for using a term is justified by the fact that to that term corresponds an idea, Hume argues that there is in fact no such idea. Nonetheless, he often argues that it does not follow that the term itself cannot be used meaningfully.29 That is, in cases such as this, Hume shows that our linguistic customs, while they may “bewitch” us, to use Wittgenstein’s term—​leading us to suppose ideas where there are none—​may nonetheless be perfectly in order, so long as we are not deceived by them.30 This, we will see, leads Hume to a very radical kind of nominalism—​a genuine precursor of Wittgenstein’s use theory of meaning—​according to which meaning often consists purely in verbal customs. Hume can advance this semantic nominalism (Sellars

29. Green (1968) observes that with regard to the idea of a persistent object, “Such an idea being according to Hume’s principles impossible, the appearance of our having it was the fiction he had to account for; and he accounts for it, as we find, by a ‘habit of mind’ which already presupposes it. His procedure here is just the same as in dealing with the idea of vacuum. In that case . . . having to account for the appearance of there being the idea of pure space, he does so by showing, that having ‘an idea of distance not filled with any coloured or tangible object,’ we mistake this for an idea of extension” (256). Note here the connection that Green observes between this strategy of Hume’s and his reliance on custom (a habit of mind). See also Dauer (2011) for an argument that Hume adopts this attitude toward necessary connection. 30. The affinities to Candrakīrti’s remarks that mundane practices and conventions are just fine, so long as we do not subject them to analysis, are obvious (Cowherds 2010; Garfield 2015).

Introduction 

( 23 )

was to call this position “psychological nominalism” a few centuries later) precisely because of the normative force he assigns to custom: as I noted previously, his naturalism, his skepticism, his commitment to custom as an explanatory device, and his nominalism are hence intimately connected, and this in virtue of the normative force of custom. Hume hence also distinguishes between believing something about the world and believing a sentence to be true. We might believe that the sun rises in the east, and, in virtue of that, believe that the sentence “The sun rises in the east” is true. But we also might believe that a sentence is true while having no idea what it means (even if we think we do). Many of us probably believe things like E = mc2 in just such a fashion. Or we might believe that when someone we trust asserts sincerely a sentence in a language we do not speak that the sentence she utters is true, whatever it means.31 Hume compares the illusion that we have a belief or concept that we could not possibly possess to phantom limb illusions among amputees. They take themselves to be feeling a sensation in a limb that they no longer have, just as we might take ourselves to have ideas or beliefs that we cannot possibly have [1.3.9.17, 117]. Another important application of this principle occurs in the context of Hume’s discussion of agent freedom. In this context he argues that we have a “false sensation or experience  .  .  .  of the liberty of indifference” [3.2.2.2, 408]. That is, we believe ourselves to have an impression of a free will, which we do not in fact have, and consequently believe ourselves to have an idea of the freedom of the will, which idea we do not even have.32 Hume also argues [1.4.3.9–​10, 222–​23] that the ancients thought that they had an idea of substance, an idea that, he argues, they could not have had. Henry James Richter ([1797] 2000) accuses Hume of incoherence in this regard, arguing that in order even to deny that we have an idea, we must know the idea we deny ourselves to have, and so must have it. In the context of Hume’s denial that we have any idea of necessary connection (to which we turn in c­ hapter 6), Richter writes: Those who resolve to abide by the principle [that any idea depends upon a precedent impression] . . . should at least be as candid as Mr. Hume has been, 31. Hume worries that those who attend Latin mass become so inured to believing that utterances in a language they do not understand are true that they come to believe the truth of theological sentences in English, and thereby take themselves to come to have beliefs about the world that they cannot possibly have, simply because no ideas attach to the words those sentences contain [1.3.8.4, 99–​100]. 32. See Baier (1991) and Garrett (1997) for excellent discussions of this passage.

( 24 )  Methodological Preliminaries and first, carefully examine whether there be not some impression from which it might be derived; after that, they may, if they please, deny its existence, as a dogmatical shoemaker might swear you have no feet, because his shoes will not fit them. But it may be said, according to Mr. Hume’s system, an idea is in fact no idea, unless it be derived from some impression; nor till he has discovered that impression, does he speak of it positively as such, he calls it only a supposed idea. What an excess of refinement is this! . . . [H]‌ere is an idea which exists only in idea. I shall not, however, attempt to prove the existence of this idea [of necessary connection] as to those who have not the idea, it would be impossible, and to those who have, superfluous. Yet it may not be amiss to apprize those who deny its existence, of the dilemma to which they are reduced. Either they must acknowledge they have the idea, whose existence they deny; or confess they have no idea of what they deny. (Richter [1797] 2000, 25–​26)

This objection would indeed have force were Hume’s analysis of pseudo-​ ideas not grounded in his nominalism. For then he could fairly be charged with forming the idea of ideas that we cannot have. But once we appreciate Hume’s account of the relation between language and universals, the objection can be dismissed. Hume’s argument can now be read as having as its conclusion simply that we falsely believe that certain words refer to ideas. There is surely no incoherence in claiming that some terms we take to refer in fact fail to do so. Hume, that is, is not engaged in a metaphysical deconstruction of real concepts, so much as in showing that our verbal behavior is often best explained not by assigning a concept to every predicate. Hume is also often charged with a kind of crude metaphysical or epistemological analysis leading to the denial that we have any knowledge of the external world,33 or that we can make any sense of personal identity or causal explanation. He is then treated as a mere whipping boy for Kant. I will argue that instead he is engaged in a principled refusal of the kind of metaphysical analysis that Kant thought was necessary, replacing it with an empirical account of verbal practices that require no metaphysical ground; he is a psychologist and a naturalist. Hume on this reading will turn out to be more a postmodern than an early modern thinker, and so a much more interesting philosopher than he is often made out to be.

33. As both Moore and Stroud understand him (Stroud 1984, 106ff.).

Introduction 

( 25 )

5.  THE PRINCIPLE OF THE UNIFORMITY OF METHOD The final methodological principle that drives my reading is a conviction that when we see the Treatise aright, we see that, despite its often apparently conflicting remarks and youthful exuberances, it is methodologically remarkably homogeneous. Hume has a vision, both of how to do philosophy, and of what human nature and a science of human nature look like, and it is when we lose sight of that vision and the method to which it gives rise that we misread Hume. The final methodological principle that drives my reading then is this: The Principle of the Uniformity of Method

Assume that Hume’s arguments and analyses are uniform in structure. Like all of the preceding principles, this principle is heuristic and defeasible. Hume, like any philosopher, is entitled to employ arguments and analyses of various forms. Nonetheless, I will argue in what follows, he almost always deploys exactly the same form of argument: a Pyrrhonian dissolution of an apparently irreconcilable duality, a demonstration that a custom we thought required grounding is in fact itself the ground, and an exploration of how we think that is aimed at providing a theory of human nature. And Hume’s theory of human nature is one according to which we are creatures of custom, that we can never be more than this, and also that this is no cause for despair. He therefore seeks explanations of our capacities not in any transcendental metaphysics or epistemology, but in empirical social, developmental, and cognitive psychology. This same form of argument, this same pattern of analysis, will be present whether we are talking about the self, causation, justification, ethics, or even theology. If I  am right, demonstrating this consistency in method and vision will be the most important contribution of this study. Kemp Smith (1941) famously defends a similar principle on the grounds of Hutcheson’s thoroughgoing influence on Hume. He writes: Hutcheson’s teaching leads up to [“Hume’s New Scene of Thought”], though by steps which Hutcheson certainly never contemplated. . . . [I]‌f the fundamental judgments of morals, as of aesthetics, rest on feeling, not on reason; and if in matters of moral conduct Nature has been thus careful in providing us, independently of all calculation and reflexion, with these “immediate monitors,” may it not be so likewise in the professedly theoretical field? May not our so-​ called judgments of knowledge in regard to matters of fact and existence be

( 26 )  Methodological Preliminaries really acts of belief, not of knowledge—​belief being a passion and not a form of insight, and therefore, like all passions, fixed and predetermined by the de facto frame and constitution of our human nature? (1941, 43–​44)

Kemp Smith argues that Hutcheson’s influence on the Treatise is broad, extending to Book I as well as to Books II and III, where it may be more apparent. Here he argues that the role Hutcheson assigns to the passions and moral sense in determining moral judgment is taken up by Hume and extended to the domain of ordinary empirical knowledge and reasoning, and hence that Hume’s commitment to belief as more a matter of the “sensitive than of the cogitative part of the soul” [183] derives indirectly from Hutcheson.34 Kemp Smith extends this account of Hutcheson’s impact on the uniformity of method to note an intriguing analogy between Hutcheson’s treatment of the passions and Hume’s treatment of the idea of necessary connection. He points out that Hutcheson (as would Hume) takes the passions all to be impressions of reflection, and that moral qualities are not directly observed in nature, but are projected based upon the observation of the relevant impressions of reflection. (We will return to this point in ­chapter 3.) Just so, according to Hume, in the case of causation: no necessary connection is observed between events in nature; instead, an impression of reflection gives rise to a projection of a connection between events. Unlike Kemp Smith, I do not believe that all of the uniformity in Hume’s method is explained by the impact of Hutcheson; nevertheless, I  am inclined to agree with him that this much is well explained by that influence.35

34.  Note that here I  follow Kemp Smith only in the insistence on a uniformity of method in the Treatise. We disagree about what that method is: Kemp Smith argues that Hume uniformly takes feeling to be the basis of our epistemic and moral life; I emphasize the role of custom. 35. Baier also gestures toward this uniformity (and its relationship to Hume’s naturalistic understanding of custom) when she writes: Our ideas copy our impressions not merely in their “simple” content, but in their perceived relationships, in their keeping company, and their ganging up. It is perceived relations and associations between persons that provide the models for the Humean mental relationships, and his idea-​associates can be “kin.” Our social relationships are the outcome of our biological mammalian nature, of our friendships, and of the social artifices and inventions we have made, and all three affect Locke’s, Hume’s and our perception of our mental and feeling states and their relationships. (1991, 29) But see Norton (1982) for a dissenting view. Norton argues (53ff.) that Hume’s approach to ethics is diametrically opposed to his approach to epistemology, that he is a realist and a naturalist in the ethical domain, but not in the epistemological domain.

Introduction 

( 27 )

As I noted at the top of this chapter, I do not take the considerations I offer here in favor of these principles of interpretation to be decisive, only motivating. The only convincing argument for them is the reading of Hume that emerges from them. I  set them out here so that the reader can see what I am doing and that I have nothing up my sleeve. We now turn to a second set of preliminary methodological questions:  Why, when writing about Hume, am I focusing on the Treatise, and why, within the Treatise, am I so focused on Book II? And why on custom?

Norton also argues that Kemp Smith exaggerates the impact of Hutcheson on Hume (and indeed that he misunderstands Hutcheson). I disagree, and find Kemp Smith’s observations that Hume adopts so much not only of Hutcheson’s framework, but also his idiosyncratic vocabulary (e.g., perceptions in lieu of Locke’s ideas and the distinction between direct and indirect passions) very convincing in this regard (Kemp Smith 1941).

C H A P T E R  2

w

Why the Treatise? Why Book II? Why Custom?

I

n this chapter, I explain why I focus directly on the Treatise, as opposed, for example, to what might be thought to be Hume’s more mature thought in the Enquiries; why within the Treatise I take Book II to be foundational; and how Hume understands custom, an explanatory principle that does so much work in the Treatise, and whose role, I believe, is often misunderstood or underappreciated. I am writing a book about the Treatise, not about Hume’s thought in gen­ eral, and, as I note in the preface, the interpretation I offer differs from many others in two principal respects. First, I  will argue—​as noted above, following both Kemp Smith and Baier—​that it is Book II, not Book I, that is the conceptual foundation of the Treatise. Second, I will argue that we cannot understand Hume’s project without an appreciation of his own understanding of custom, and in particular, without an appreciation of the grounding of his thought about custom in the legal theory and debates of his time. Custom is the source of Hume’s thoughts about normativity, not only in ethics and in political theory, but also in epistemological, linguistic, and scientific practice, and is the source of his insight that our psychological and social natures are inextricably linked. The centrality of custom and the link between the psychological and the social are closely connected, because it is in the account of the passions in Book II that Hume directly links his thought about custom to human psychology. I now take each of these points in turn, and will conclude this chapter with some remarks on the textual horizons that contextualize my reading. But first, a few remarks on why I am choosing to write about the Treatise, per se.

( 28 )

W H Y T H E T R E AT I S E ?  

( 29 )

1.  WHY THE TREATISE? Annette Baier begins her masterpiece A Progress of Sentiments (1991) by asking the question, “Why another book on Hume?” (vii) I agree with her answer: Because he is a philosopher who invited “the latest posterity” to reinterpret him, because my own interpretations differ in some ways from others, including those to which I am very much indebted, and because my own interpretations of the Treatise would “loosen and fall of themselves, when unsupported by the approbation of others . . .” [1.3.7.2, 264–​65] (1991, vii)

Hume does invite posterity to reinterpret him, and I, like Baier, offer an interpretation that I take to be somewhat different from those of others, including “those to which I  am very much indebted,” including even, in some details, Baier’s own. I also agree with Baier’s remark in the next paragraph that the Treatise must be read as one book. On the other hand, Baier never asks, “Why the Treatise, as opposed, say, to the Enquiries?” After all, Hume regarded the Enquiry Concerning the Principles of Morals as his finest work, and many, Hume included, have taken the Enquiries to be the more mature, careful articulation of Hume’s views. Hume himself in a letter to Hutcheson (September 17, 1739, in Hume 1932, vol. 1, 32–​36) expressed embarrassment over both the tone of the Treatise and what he saw as errors in that work. (“I must own, my Proofs were not distinct enough & must be alter’d” [36].)1 1.  Strawson argues that one should only read the Treatise account of causality in the context of the discussion of causality and induction in the Enquiry Concerning the Human Understanding. He writes: The main case [for his account of Hume’s analysis of causality] rests on the Enquiry. Some have claimed that the Enquiry is somehow of less importance than the earlier Treatise, but there is every reason to take it as more representative of Hume’s considered views on causation than the Treatise—​even if one thinks . . . that one can simply take no notice of Hume’s partial repudiation of the Treatise in the Advertisement which he prefaced to the Enquiry in the last year of his life. Hume was at the height of his powers when he wrote the Enquiry. He had not gone soft in the head. He had had more time to think. He was trying to make his position as clear as possible. He meant what he said and said it beautifully. (1989, 8) One need not deny the clarity and beauty of the Enquiry to disagree with Strawson’s hermeneutic policy here. The Enquiry Concerning the Human Understanding, whatever its virtues, has a very different structure than the Treatise of Human Nature, and in particular, devotes much less attention to the role of the passions and to that of custom in human nature than does the Treatise. Even if one thinks that Hume got things better in the Enquiries than in the Treatise (and I do not), the Treatise demands attention on its own, in virtue of the enormous influence it has had, and because it represents the only comprehensive exposition of Hume’s system.

( 30 )  Methodological Preliminaries

Perhaps Baier, like so many who have recently written about Hume, saw the answer to this question as too obvious to merit posing it. There is a reason that the vast majority of recent books and articles on Hume have focused on the Treatise. Despite its youthful enthusiasm, and despite Hume’s own later dissatisfaction with it, the Treatise remains Hume’s philosophical masterpiece. It draws together all of the elements of Hume’s philosophical vision in an organic unity. It explores the difficulties Hume himself found with his own views with disarming candor. It follows all of the byways. But more than that, while the two subsequent Enquiries revisit much of the ground covered by Books I and III of the Treatise, neither attends with care to Book II, and they are composed in such a way that the links between them, and in particular their relation to the issues explored in Book II, are obscured. Hume did return to the passions in the Dissertation on the Passions at the end of his life. But once again, that text is disconnected from each of the Enquiries. Moreover, it recapitulates, in very much abbreviated form, the material explored in greater detail and with more grace in Book II of the Treatise. The organic unity of the Treatise and its enormous detail account for its continuing fascination. That’s why I am writing about the Treatise.

2.  WHY BOOK II? My reading of the Treatise will take Book II—​easily the least addressed book of the Treatise—​as its starting point, an approach suggested by Baier’s own work. I start there not because Book II is so often ignored or given short shrift in treatments of the Treatise, although that neglect might on its own justify this approach. I start there because it is the conceptual foundation of the Treatise. In A Progress of Sentiments, Baier writes: Philosophical reading habits still seem to be resistant to Hume’s attempt to link the theses of Book One with those of Books Two and Three, indeed resistant to treating the Treatise as a treatise, as distinct from a series of self-​contained section-​long philosophical essays. (1991, 158)

This reflection remains true today. Many treatments of the Treatise focus only on Book I, or, if they go beyond Book I, skip blithely over Book II to the Also see Taylor (2002) for an argument for the superiority of the moral theory in the Enquiry Concerning the Principals of Morals to that of the Treatise. There is also room for debate here, and I think that the treatment of ethics in the Treatise gains particular force because of the context of the system as a whole.

W H Y T H E T R E AT I S E ?  

( 31 )

treatment of morals in Book III. As a result, not only do they miss out on the treatment of the passions, per se; they also miss the structure of Hume’s enterprise, as well as its connection to the literature to which he is responding. The great exception to this rule is Kemp Smith (1941). In part III of his Detailed Consideration of the Central Doctrines, taken in what may be Presumed to have been the Order of their First Discovery, he argues that the doctrines of sympathy and of the operations of the passions adumbrated in Book II must have been settled before the doctrines developed in Book I: “Hume’s doctrine of belief is, I shall argue, modelled throughout upon his doctrine of sympathy, which must have been antecedently arrived at” (161). Baier agrees with the conceptual (but not the compositional) primacy of the Book II account when she writes, “The whole of Hume’s epistemology, in Book One of the Treatise, is in the service of his philosophy of passion and action in Books Two and Three” (1994, 78). While in general I agree with Baier, in this study I defend a stronger interpretative claim, closer to Kemp Smith’s position (once again, with regard to conceptual, not compositional, priority): the epistemology is not only in the service of the account of the passions; it is also grounded in that account; and the ethics of Book III is also grounded in Book II. Therefore, unlike Baier, I will follow Kemp Smith’s suggestion and start with Book II, and then return to Book I before moving on to Book III. Hume states early on [1.1.2.1, 8] that his principal topic is the passions.2 The Treatise is, after all, a treatise of human nature, a book about how the mind works. Hume’s avowed purpose is to “introduce the experimental method into philosophy,” that is, to create a psychology. Hume’s approach to human psychology is to eschew an account of free agency in favor of a causal account of how the mind works in which agency is naturalistically reconstructed; to eschew an account of spontaneous reasoning in favor of an account of blind, unconscious mental processes that come to constitute the reasoning on which we in fact rely. In this regard, Hume is indebted to Mandeville.3 In the introduction to the Fable of the Bees, Mandeville writes: One of the greatest Reasons why so few People understand themselves is, that most Writers are always teaching Men what they should be, and hardly ever trouble their heads with telling them what they really are. As for my 2. Alanen (2006) agrees. 3. Malebranche and Hobbes are also important intellectual antecedents in this regard. Pears (1990) finds this naturalism hard to swallow. Commenting on Hume’s remark that “all probable reasoning is nothing but a species of sensation. . . . Objects have no discoverable connexion together; nor is it from any other principle but custom operating on the imagination, that we can draw any inference from the appearance of one to the existence of another” [1.3.8.11, 103], Pears writes, “This is one of many passages in which he presents his theory on a popular stage in paint and powder” (96).

( 32 )  Methodological Preliminaries part  .  .  .  I  believe Man (besides Skin, Flesh, Bones, &c. that are obvious to the Eye) to be a Compound of various Passions, that all of them, as they are provoked and come uppermost, govern him by turns, whether he will or no. (1989, 77)

Hume concurs with many of the radical theses enunciated in this brief passage of a very brief introduction. Self-​understanding, Hume will argue, is limited: we are not transparent to ourselves, and we tend to conceive our nature more in terms of how we would imagine ourselves to be than in terms of how the best science tells us we actually are. (Not much has changed in that regard.) That science, Hume will argue, tells us that our thought and behavior are determined by psychological forces, not by the free action of the autonomous agents we may take ourselves to be, and those governing forces comprise the passions. It is for that reason that Book II is so essential to the project. Moreover, as Mandeville argues in the Fable, morality is concerned primarily with the way those passions interact in the context of collective action, and with how they are thereby modulated in the context of social interaction and collective agreement. Perhaps most radically of all, Hume’s approach to human psychology, like that of Mandeville, is not only causal, but also social, as opposed to individualistic. Hume recognizes the fundamental role that upbringing, language, and collectively constituted conventions for behavior and thought have in determining our mode of engagement with the world. As Baier (2011) notes, Hume’s account of justice as an extension of natural sympathy, enabled by the recursive character of our collective customs, is grounded in his sense that we are naturally social creatures, and that any account of human nature must be an account of our social nature.

Pears argues instead that Hume is after a normative epistemology, an account of epistemic warrant. He writes that “Hume’s theory of mind is intended to help us towards answers to two distinct questions: ‘What ideas may we legitimately have?’ and ‘What may we legitimately believe?’ ” (3). Part of the burden of this study is to show that Hume is not after a normative account of epistemology; he is after a psychological account. This error runs through Pears’s treatment of Hume’s epistemology. Unlike Pears, I take Hume at his word. He is offering a thoroughgoing naturalistic account of reasoning as a causal process mediated by natural cognitive processes in the imagination; he is not offering an account of epistemic warrant; nor is he defending reason as a probative faculty. (See also Morris 2006, 89–​91, for an argument that Hume, in his discussion of causal reasoning, is providing a normative epistemology, as opposed to a naturalistic account of causal reasoning.) But Morris adds an important wrinkle, locating the normativity in the actual norms that we use, that is, in our customary behavior. With this account of normativity grounded in custom, and with the understanding of custom in turn grounded in the law, the naturalist need have no quarrel.

W H Y T H E T R E AT I S E ?  

( 33 )

For this reason, the foundation of Hume’s account of psychology lies in his account of custom. In this he follows not only Mandeville, but also Shaftesbury, who so eloquently argued for the essentially social nature of human psychology. In particular, Hume is sympathetic to Mandeville’s relativism about morality (Baier 2001, 332–​37), grounded in his view that our ethical sensibilities, like our aesthetic sensibilities, are shaped by the cultures we inhabit; although, as we will see, he also insists that that variability is limited by our natural psychological constitution. While Hume’s attention to custom, and more particularly to the passions, owes a great deal to Mandeville, Hutcheson, and Shaftesbury, it is not simply derivative of their views, and indeed it could not be, given the differences among those philosophers themselves. Whereas Hutcheson, in his essay on the passions, mentions both social conventions and the individual passions, he never joins the two. Whereas Mandeville sees us as driven by our passions, and sees our passions as shaped by our social milieu, he does not see these forces as species of the same genus. It is Hume who first takes the broader category of custom he borrows from the law to comprise in a seamless way our biological and our social behavioral determinants, and this insight drives much of the Treatise.4 To understand Books I and III, one must understand Hume’s account of custom, and that account is developed in Book II. So, while contemporary readers often take Hume to mean by the word “custom” social convention, the meaning the term most naturally has today, as we have seen, his usage is much broader than that, connoting what we customarily do, how we customarily think, and always with a view to its norm-​constituting power. That is, custom comprises not only social convention, but also biologically determined or otherwise fundamentally cognitive patterns of thought or behavior, as well as acquired habits, including the habits and practices of reinforcing our customs. The natural kind then is also nicely captured by the term passion, the nominal topic of Book II, again following Shaftesbury and Hutcheson’s usage. A passion is something that happens to us, not something we do, that with respect to which we are passive, not active. The central insight animating the Treatise is that to explain our behavior, including epistemic, metaphysical, and moral behavior, including both our solitary cognitive behavior and our social behavior, is to characterize the causes of that behavior—​to identify the structures and event types with respect to which we are passive, and to understand the regularities that connect them to

4. John Wright (1983) sees this as well.

( 34 )  Methodological Preliminaries

their cognitive, linguistic, and behavioral effects. And we are passive in this sense not only with respect to the innate and the learned, but also with respect to the social. Hume’s model for the explanation of behavior is the physics of his time (as it was to be, mutatis mutandis for certain of his behaviorist descendants, such as J. B. Watson and Clark Hull). When he distinguishes between impressions and ideas on the basis of their respective force and liveliness (5), his model for force is Newtonian. Impressions impart more force in their interactions with other perceptions than do ideas; they are more closely related to their distal causes than are ideas. His focus on the passions—​the topic that most concerns me in this volume—​reflects his view that the proper explanation of human behavior is one that takes us to be passive in relation to the psychological forces that act on us, not independent actors imbued with a spark of divine freedom, whose psychology is to be explained in terms of reasons. Hume’s naturalism is, like contemporary naturalism, a commitment to understanding human cognition using the methods of the natural sciences.5 Book II thus forms the foundation of the Treatise. Only by appreciating Hume’s account of custom and of the nature of the passions and their role in our cognitive life can we understand the epistemology and psychology he develops in Book I, as well as the account of ethics and the moral sentiments in Book III. Much of what I will in what follows argue is erroneous in contemporary readings of Book I stems simply from a failure to appreciate the importance of Book II, in which Hume is concerned to explain our cognitive and social capacities, that is, our customary thought and behavior—​including both that which is customary in the habitual sense and that which is customary in the sense that it is grounded in social convention. This is the ground of Hume’s naturalistic explanation of the normative dimensions of human life. In order to understand the connection, and the way that custom as mediated by the passions can play this role, it is essential that we understand the origins of Hume’s thinking about custom. And that origin lies in eighteenth-​century debates about customary law, a topic to which we now turn.

5.  Pears writes, “Where reason fails us, nature supports us. That is the message of Hume’s naturalism” (1990, 79). This is wrong. Hume’s naturalism is more radical and thoroughgoing than this. Nature isn’t there as a safety net into which we fall when reason cannot support us; reason itself is a natural phenomenon, the operations and the limitations of which are to be explained by a science of human nature. As Owen (1999) correctly notes, Hume sees reason as a function of the imagination, a faculty that operates via complicated associative principles. Reason as it would have been imagined by Descartes or by Leibniz is simply not on the cards for Hume, to succeed or to fail.

W H Y T H E T R E AT I S E ?  

( 35 )

3.  “THE STRENGTH OF CUSTOM” Hume studied the law and had a deep interest in history, an interest that led later in his life to literary fame earned by his History of England (Hume 1983–​85). His legal and historical studies would have acquainted him with the debates concerning the status of customary law in the eighteenth century, in which both the scope of customary law—​as opposed to common law—​and the role of custom as a foundation of the normative force of law were at stake.6 Custom, in this discourse, was considered a principal source of nomicity and normativity. Indeed, Francis Bacon had remarked that custom is the “principal Magistrate of Man’s Life” (Thompson 1991, 2–​3; Bacon 1909, 104).7 It is therefore essential to read Hume’s use of the word custom, so frequent in the Treatise, and his appeals to regularity in the context both of natural law and of ethics in the context of this legal history, a context all too often ignored in philosophical literature on Hume, but well studied by historians of his period. Pocock (2016) writes: In agrarian societies which are highly decentralized and traditional, but which a professionally organized class of literate bureaucrats, obedient to a central direction, is trying to bring under control, it is common—​at least in the West—​to find a distinction between unwritten custom, usage, or tradition, recognized by the king’s servants but recognized as being already established by the spontaneous and traditional adoption of society itself, and the written commands,

6. Rorty (1982, 169–​70) also notes Hume’s indebtedness to history and the law for his view regarding the normative force of custom. 7. The passage in which this remark occurs is worth quoting in full, as it reflects the thought about custom with which Hume was familiar: Therefore, since custom is the principal magistrate of man’s life, let men by all means endeavor to obtain good customs. Certainly custom is most perfect when it beginneth in young years: this we call education; which is, in effect, but an early custom. So we see, in languages the tongue is more pliant to all expressions and sounds, the joints are more supple to all feats of activity and motions, in youth than afterwards. For it is true that late learners cannot so well take the ply; except it be in some minds that have not suffered themselves to fix, but have kept themselves open and prepared to receive continual amendment, which is exceeding rare. But if the force of custom simple and separate be great, the force of custom copulate and conjoined and collegiate is far greater. For there example teacheth, company comforteth, emulation quickeneth, glory raiseth: so as in such places the force of custom is in his exaltation. Certainly the great multiplication of virtues upon human nature resteth upon societies well ordained and disciplined. For commonwealths and good governments do nourish virtue grown, but do not much mend the seeds. But the misery is, that the most effectual means are now applied to the ends least to be desired. (Bacon 1909)

( 36 )  Methodological Preliminaries edicts, ukases, or statutes as Fortescue calls them, imposed upon society by order of the king and his literati, whether or not those claim to be digesting or modifying what was previously unwritten tradition. . . . English lawyers sometimes attempted to distinguish upon this basis between unwritten law or lex non scripta, which might be written down, but which claimed no authority but that of custom and tradition, and written law, lex scripta or statute, whose authority was that of the author of the writing. (12–​13)

Pocock argues that in the domain of custom it was standard practice in Hume’s time to take long usage itself as legitimating. He quotes Fortesque, whom Hume also read: The kingdom of England was first inhabited by Britons, then ruled by Romans, again by Britons, then possessed by Saxons, who changed its name from Britain to England. Then for a short time the kingdom was conquered by Danes, and again by Saxons, and finally by Normans, whose posterity hold the realm at the present time. And throughout the period of these nations and their kings, the realm has been continuously ruled by the same customs as it is now, customs which, if they had not been the best, some of those kings would have changed. . . . Indeed, neither the civil laws of the Romans, so deeply rooted by the usage of so many ages, nor the laws of the Venetians, which are renowned above all others for their antiquity—​though their island was uninhabited, and Rome unbuilt, at the time of the origin of the Britons—​nor the laws of any Christian kingdom, are so rooted in antiquity. Hence there is no gainsaying nor legitimate doubt but that the customs of the English are not only good but the best. (quoted at 13–​14, from De Laudibus, chap. 16)

None of this would have been news to Hume. Indeed, as Pocock also notes, Hume’s younger contemporary, Edmund Burke, affirmed that “because a custom or a particular institution had a ‘prescriptive’ claim—​i.e. was already established—​there was a ‘presumption’ in its favor” (quoted at 15). And Govind (2015) makes the connection to Hume explicit when he observes: In the early eighteenth century, with Hutcheson’s elaboration of moral sense and his intense engagement with Mandeville, the place of the moral within social action and its relationship to public interest became the key subject of investigation in England. . . . [T]‌he “historical jurisprudence” of Montesquieu and his attempt at relating positive law to climate and custom, formed the other great dimension of the landscape. Hume and Rousseau in different ways try and account for the moral at a more fundamentally human level than Montesquieu, and do so by radicalizing

W H Y T H E T R E AT I S E ?  

( 37 )

the “customary” dimension.  .  .  . The moral is now articulated within a specifically temporal framework through which a man and the nature of his species-​ being . . . come to define and explain the foundation of positive law. (188–​89)

That is, British moral theorists, Hume among them, understood customary law—​the law that emerged from local patterns of behavior and traditions—​ as constituting the foundation of the legitimation of positive law. (See, for instance, Hale 1713.) On this view, regular patterns of conduct establish the expectations of one another that come first to be enforceable by local magistrates who recognize traditional claims, and then come to be codified in written law, which in turn might be extended in scope to become the common law of the kingdom. This view of the normative force of custom in the law and in other domains of daily life was, moreover, not confined to legal scholars. Instead, it was a common understanding of normativity in Great Britain, not only in Hume’s time, but for a long time before then. Elsky (2014) argues persuasively that in the early seventeenth century, legal rights were broadly understood to be grounded in custom. She relates the tale of Lady Anne Clifford’s battle for her inheritance and shows that that case depended on Lady Anne’s maintaining the custom of “Causeing the Boundaries to be ridden” (521), or regularly perambulating her land, as a means of demarcating the boundaries and establishing proprietorship, a practice not recorded in lex scripta, but enshrined in customary practice. Lady Anne Clifford’s suit balanced this customary practice against written law. Elsky writes, “The force of custom, of doing something over and again through time, lies at the heart of common law, the legal system that governed the holding and devolution of land. Indeed, English common law was often referred to as ‘the common custom of the realm’ ” (523–​24).8 In characterizing the arguments of Lady Anne’s lawyers, Elsky continues, “They employed the logic of common law [that is, customary law]: promulgations in writing do not bring something new into existence but rather codify an already existing custom” (529). The late seventeenth-​century to early eighteenth-​ century legal theorist Matthew Hale, she points out, argues that Authoritative and Original Institutions are not set down in Writing . . . with that Authority that Acts of Parliament are, but they are grown into Use, and have acquired their binding Power and the Force of Laws by a long and immemorial 8. Elsky here and elsewhere (2013a, 2013b) uses the term “common law” to denote what is more often referred to as “customary law,” using “common law” in its present sense rather than that in which it was used at that time, referring to the law common in the realm.

( 38 )  Methodological Preliminaries Usage, and by the Strength of Custom and reception in this Kingdom. The Matters indeed, and the Substance of those Laws, are in Writing, but the formal and obliging Force and Power of them grows by long Custom and Use. (Quoted in Elsky 2014, 531)

The source of normativity, on this account, is the legitimate expectation of the continuation of traditional forms of conduct as well as traditional rights and obligations.9 And so, despite the growing importance of written common law in English jurisprudence, in Hume’s time, customary law was universally recognized as binding, and as binding because it was customary.10 Carter’s Lex Customaria, a text Hume would have studied, states:

9. On the other hand, as Govind points out in another context, “What is broadly seen as ‘custom’ cannot be understood outside the constituting ambit of sovereignty, that is, that by which laws were ultimately authorized” (2017, 1). By the eighteenth century, as the power of the Crown was increasingly asserted, we find an essential interplay in British legal theory between customary law and common law, in which origins and initial legitimacy may be grounded in custom, but custom requires eventual certification by common law. Govind puts it this way: It [is common law] rather than any putative native custom or tradition which . . . [is] constitutive.  .  .  .[I]‌n fact, according to 18th-​century jurisprudence, custom itself had to be reasonable, that is, “of artificial and legal reason, warranted by authority of law.” (6, quoting Blackstone) This point is also important for understanding the background of Hume’s use of this term. There were two issues at stake in eighteenth-​century debates about the relationship between customary and common law. One concerned their relative scope, and the legitimacy of the gradual replacement of customary by common law as the Crown grew in strength—​a development that Hume supported. The other—​more theoretically interesting debate—​concerned their relative position in the order of legitimation. Some—​including contractarians—​held that common law provided the normative underpinnings of custom. Part of Hume’s critique of the social contract theory in Treatise 3.2.7 and 3.2.8 and in his essay “Of the Original Contract” is that contracts or promises have no binding force until there are customs establishing them, and so that common law gets its normative force from the customs and customary law that precede and undergird it. Hume’s moral realism and naturalism consists in his appeal to facts about custom as constitutive of normative facts. 10.  Pocock (2016) comments that the influential English jurist and chief justice Edward Coke (fifteenth century) repeated Fortesque’s argument to King James I (17). And he quotes John Pym (seventeenth century), arguing that the very legitimacy of Crown or common law rests on a foundation of custom: There are plain footsteps of those laws in the government of the Saxons; they were of that vigour and force as to overlive the Conquest, nay, to give bounds and limits to the Conqueror, whose victory gave him first hope. But the assurance and possession of the Crown he obtained by composition, in which he bound himself to observe these and the other ancient laws and liberties of the kingdom, which afterwards he likewise confirmed by oath at his coronation. From him the said obligation descended to his successors. It is true they have been often broken, they have been often confirmed by charters of kings, by acts of parliaments, but the

W H Y T H E T R E AT I S E ?  

( 39 )

For a Custom taketh beginning and groweth to perfection in this manner. When a reasonable Act once done is found to be good, and beneficial to the People, and agreeable to their nature and disposition, then do they use it and practise it again and again, and so by often iteration and multiplication of the Act, it becomes a Custom; and being continued without interruption time out of mind, it obtaineth the force of a Law. (Quoted in Thompson 1991, 97)11

Pocock sums this up:  “The Englishman who saw his realm as a fabric of custom, and himself as a custom-​generating animal, saw proprietor, litigant, judge, counselor, and prince as engaged in a constant activity, one of preserving, refining and transmitting the usages and customs that made him and England what they were” (341). And Pocock also notes that Hume endorsed this fundamental role of custom in preserving society and in grounding law (495). We should also bear in mind that while for Hume and his contemporaries custom in this sense is local, unwritten, and informal, it is, for all that, legitimating. Once again, this point has clear ancestry in British legal theory. Thompson reminds us that law was derived from the customs, or habitual usages, of the country; usages which might be reduced to rule and precedents, which in some circumstances were codified and might be enforceable at law. This was the case, above all, with lex loci, the local customs of the manor. These customs, whose record was sometimes only preserved in the memories of the aged, had legal effect, unless directly voided by statute law. (4)

Thompson provides numerous examples of court actions brought in the eighteenth and early nineteenth centuries in Britain on the grounds not that common statute law had been violated, but that local customary law had been violated, where the violation typically involved simply the abrogation of a traditional practice. Some involve labor practices; others access to orchards or grazing areas. In each case, the unwritten custom was held to have the force of law. Moreover, he notes, despite that fact that custom is local, perhaps pertaining only to a single manor, a pasture, or a stretch of river or shoreline, it may supersede common law in that locality (97–​98),

petitions of the subjects upon which those charters and acts were founded were ever petitions of right, demanding their ancient and due liberties, not suing for any new. (Quoted at 359) 11. As Elsky (2014, 531) points out, this passage in fact originates in Davies (1615, 3).

( 40 )  Methodological Preliminaries

indicating once again the powerful normative force custom was taken to have at Hume’s time.12 A reader of the History of England will note that this sense of the centrality of custom—​as a pattern of uncodified habits that generate expectations and that both govern our actual behavior and ground normative claims—​never leaves his thought. In fact, he returns to this theme in his account of the evolution and moral foundation of English law at several points in the History. In the second appendix to volume 1, Hume presents an extended account of the evolution of English law from its origins in Norman feudal law (overlaid and infused by Saxon legal practices as well). He begins by noting that “the feudal law is the chief foundation, both of the political government and of the jurisprudence, established by the Normans in England” (1983–​85, 1:455). The kind of foundation Hume has in mind is not only historical, but justificatory. But, as we will see, Hume’s understanding of that justificatory force is not originalist in the sense that term is used by some contemporary conservative American constitutional theorists. Instead, Hume takes the Norman conquest to have instituted a sequence of customs that culminate in the legal practices of his day. And he takes those practices to be justified not in virtue of being specifically prescribed by early Norman law, but in virtue of having been initiated by the customs it enshrined. He begins with a discussion of the origins of property rights: The attachment, naturally formed with a fixed portion of land, gradually begets the idea of something like property, and makes the possessor forget his dependant situation, and the condition which was at first annexed to the grant. It seemed equitable, that one who had cultivated and sowed a field, should reap the harvest: Hence fiefs, which were at first entirely precarious, were soon made

12. Pocock (2016) makes a similar point. He notes that petitioners often appealed successfully to customary law in order to challenge common law (25–​26) and that manorial customary law was bound strictly to precedent (199–​200). Customary law—​ not common law—​also governed such things as the obligations of innkeepers to their guests (277–​288). See also Bentham (1825). Elsky (2013a, 197–​99) calls our attention to the fact that in Utopia, Thomas More argues that one of the respects in which Utopia is superior to other nations is that its inhabitants are ruled entirely by custom, as opposed to by written law, thus only by laws that have immediate warrant and force. And Elsky (2013b) demonstrates that Spenser and Harvey in the sixteenth century each appealed directly to custom contesting the appropriateness of English or French pronunciation in poetics. Spenser, for instance, argues that “[the English pronunciation] is to be wonne with Custome, and rough words must be subdued with use” (quoted at 4). This shows that an appreciation of the normative force of custom in British thought goes far beyond the courts and the legal theorists, and was part of the intellectual air that someone like Hume would have breathed.

W H Y T H E T R E AT I S E ?  

( 41 )

annual. A  man, who had employed his money in building, planting, or other improvements, expected to reap the fruits of his labour or expence: Hence they were next granted during a term of years. . . . It was found, that a man would more willingly expose himself in battle, if assured, that his family should inherit his possessions. . . . Hence fiefs were made hereditary. . . . The idea of property stole in gradually upon that of military pay; and each century made some sensible addition to the stability of fiefs and tenures. (1983–​85, 1:458)

In this discussion, we see an analysis of the rights of property as emerging not from abstract principles, but from a network of customs, each adopted not as a matter of moral principle or judicial imperative, but because of the utility of those customs in a social matrix. For the early Normans, even networks of authority, Hume emphasizes here, are grounded in informal, not in formal, or fundamentally juridical institutions: In all these successive acquisitions, the chief was supported by his vassals; who, having originally a strong connexion with him, augmented by the constant intercourse of good offices, and by the friendship arising from vicinity and dependance, were inclined to follow their leader against all his enemies, and voluntarily, in his private quarrels, paid them the same obedience, to which by their tenure they were bound in foreign wars. (1983–​85, 1:458)

And he makes this point explicit when he concludes this discussion: Law, in its commencement, was not an intricate science, and was more governed by maxims of equity, which seem obvious to common sense, than by numerous and subtile principles, applied to a variety of cases by profound reasonings from analogy. An officer, though he had passed his life in the field, was able to determine all legal controversies which could occur within the district commanded to his charge; and his decisions were likely to meet with a prompt and ready obedience, from men who respected his person, and were accustomed to act under his command. (1983–​85, 1:460)

Note that Hume begins this passage with a clear distinction between lex scripta—​common law—​and lex non scripta—​customary law. He points out that at the beginning all law was customary, grounded in the practice of local officeholders. But even the authority of those local officeholders is derived in turn from the fact that those subordinate them were “accustomed to act under [their] command.” Authority, that is, is customary all the way down: a network of habits of obedience and social structures established by those habits empowers

( 42 )  Methodological Preliminaries

local officers; local officers rule on matters of law relying on local customs and “maxims of equity, which seem obvious to common sense.” The “subtile principles, applied . . . by profound reasoning” that constitute common law come along later. And, Hume then emphasizes, these threads of custom are woven into a fabric that structures society, law, and the practices of government: After this manner, the vast fabric of feudal subordination became quite solid and comprehensive; it formed everywhere an essential part of the political constitution; and the Norman and other barons, who followed the fortunes of William, were so accustomed to it, that they could scarcely form an idea of any other species of civil government. (1983–​85, 1:460)

Moreover, this entire fabric is not, ab initio, woven on a weft of justification, but rather on the basis of consent, that is, of agreement, or custom: The northern nations had no idea, that any man, trained up to honour, and enured to arms, was ever to be governed, without his own consent, by the absolute will of another; or that the administration of justice was ever to be exercised by the private opinion of any one magistrate, without the concurrence of some other persons, whose interest might induce them to check his arbitrary and iniquitous decisions. The king, therefore, when he found it necessary to demand any service of his barons or chief tenants, beyond what was due by their tenures, was obliged to assemble them, in order to obtain their consent. (1983–​85, 1:461)

Only later, Hume argues (1:461–​66), in the interest of greater military and economic efficiency, does greater legislative, executive, and judicial authority gradually become concentrated in the Crown, with Parliament as an ally. That concentration leads to the establishment of common law. Nonetheless, even when common law is well established, lower courts remain in practice the locus of juridical decision-​making in local matters, and they often rely on customary law. Regardless of the relative scope of customary and common law, Hume’s point is that while the eventual sovereignty of common law is both rational and inevitable, that common law has its origins in, and receives its legitimation from, customary law. Hume sums it up thus: Such an accumulation of powers [by the king’s court over time] was itself a great source of authority, and rendered the jurisdiction of the court formidable to all the subjects. . . . Law now became a science. (1983–​85, 1:473)

W H Y T H E T R E AT I S E ?  

( 43 )

It might appear that custom in this sense would be an entirely conservative, even reactionary, force, cementing practice in place and making it impossible to legitimate change. But this would be to miss the fact that, just as positive law includes the laws that govern the change in laws, among the customs that a people adopts are customs for assessing and revising customs. It is the recursivity of custom that saves a legal theory that takes custom so seriously from rank conservatism and relativism and that explains how it is possible to examine and to revise our practices and values using those very practices and values as bootstraps.13 13. Loeb (2005) argues for a very different account of justification, giving stability a central role. For instance, Loeb (3) cites Hume’s comment in the context of his discussion of the abstraction from one’s particular interests in moral judgment, “In order, therefore, to prevent those continual contradictions, and arrive at a more stable judgment of things, we fix on some steady and general points of view” [3.3.1.15. 581–​82] to support the idea that in moral judgment stability is what justifies our confidence. But this passage does not support this view. The “stability” that Hume is after here is really defensibility, as opposed to mere personal preference, and what justifies the moral judgment, as we will see in ­chapter 11, is in the end a set of customs. A bit later (10) Loeb, in a discussion of 1.4.2 (“Of the Sceptical and Other Systems of Philosophy”) refers to the “uneasiness” Hume notes arises from the contrary implications of principles in tension with one another in skeptical arguments (e.g., 1.4.2.34, 205) to argue that similar considerations apply to Hume’s epistemology. Once again, the text cited does not support the conclusion. Here Hume is merely indicating that we cannot rationally assent to inconsistent positions, not that a stable assent to one position or the other would be ipso facto justificatory. There is much in his book with which to agree, but I cannot agree that Hume takes stability, per se, to be the basis of epistemic or moral justification. Hume appeals to custom much more explicitly and much more regularly than he does to stability. And, while, as the present discussion indicates, custom is not always stable, it does have the kind of conservative ballast that generates a kind of broad stability in our practices. So, as Loeb sees, our justificatory practices are stable. Nonetheless, that stability is not a justificatory primitive, but an artifact of custom. Custom, not stability, is doing the heavy justificatory lifting. Korsgaard (1996, 60–​61) defends a similar view regarding normativity. She writes, “A faculty’s verdicts are normative if the faculty meets the following test: when the faculty takes itself and its own operations for its object, it gives a positive verdict” (60). She argues that Hume’s despair at the end of Book I reflects the failure of the understanding to meet this test, and his satisfaction at the end of Book III to reflect the success of the moral faculty in meeting this test. While she is right that the understanding is self-​ undermining in a way that the moral sentiments are not, this does not show that this is the source of normativity. Hume is quite clear that that origin is custom. Custom is just tied more tightly to sympathy than it is to reason. I take the plausibility of the account of the Treatise I present in this volume to demonstrate this point. To rebut Loeb’s carefully and extensively argued reading in detail would require a book of its own. Here I only can say that I believe that when taken as a whole, the account I  present that sets custom at the basis of normativity and justification in Hume’s work provides a better systematic account of Hume’s project than does Loeb’s, and moreover, that the role that considerations of stability play in Hume’s account are in fact explained better by this account than by one that gives those considerations a foundational role.

( 44 )  Methodological Preliminaries

This recursivity of custom allows it to develop, and permits a kind of progressiveness as law and morality develop in a society. Elsky notes that Spenser in his debate with Harvey over the future of meter in English poetry has this recursivity in mind. She points out that Spenser “views custom as the mechanism that accommodates change as it is introduced by conquest” (Elsky 2013b, 6). She argues that this position regarding English poetry has an explicit political dimension: For Spenser, custom will not further cement the practice of rhyme, but enable its displacement. He thus reimagines the stalwart “custom” as a flexible mechanism for innovation . . . . . . [C]‌ustom was usually identified with indigenous practices as a means of resisting invasion and foreign encroachment. Yet Spenser employs custom in precisely the opposite way, viewing it as an agent of conquest, as part of the process by which the native accommodates itself to foreign rule. (Elsky 2013b, 7)

Hume, like Spenser, takes custom to have this progressive use as well, and to be capable of legitimating innovation as well as the preservation of a practice. Hume’s legal understanding of custom also affords him the resources to understand conceptual, social, political, and moral change.14 At the close of volume 2 of the History, Hume draws these ideas together, grounding the argument for the justificatory primacy of custom over common law in his view of the history of the evolution of English law: The ancient Saxons  .  .  .  seem to have admitted a considerable mixture of democracy into their form of government.  .  .  . After this tribe was settled in England . . . the balance seems to have inclined to the side of aristocracy. The Norman conquest threw more authority into the hands of the sovereign, which, however, admitted of great controul. . . . The establishment of the Great Charter exalted still higher the Aristocracy . . . and gradually introduced some mixture of Democracy into the constitution. (1983–​85, 2:524–​25) In each of these successive alterations, the only rule of government, which is intelligible or carries any authority with it, is the established practice of the age, and the maxims of administration, which are at that time prevalent, and universally assented to. Those who, from a pretended respect to antiquity, appeal at every turn to an original plan of the constitution, only cover their turbulent spirit and their private ambition under the appearance of venerable forms. (1983–​85,  2:525)

14. See also McArthur (2007).

W H Y T H E T R E AT I S E ?  

( 45 )

The point could not be any plainer. The only rule of government, which is intelligible or carries any authority with it, is the established practice of the age. Justification proceeds from custom and grounds common law. It does not, as those who “appeal at every turn to an original plan” would have it, go the other way: lex scripta derives its authority from custom, and cannot therefore confer authority on custom. At the beginning of volume 5 he writes, “Habits, more than reason, we find in everything to be the governing principle of mankind” (1983–​85, 5:4). And it is this conviction that custom lies at the base of our psychology, of our social practices, of the norms that govern our reasoning, of our moral judgments, and even of our faith in the reality of the world around us that animates the positive side of Hume’s skepticism in the Treatise.15 Custom is the ligature that binds Hume’s naturalism, his skepticism, and the nominalism that we will encounter in the Treatise. It constitutes the framework that will allow him to move from the observation of behavioral and social regularities to the characterization of moral and epistemic practices with genuine normative content; it gives rich explanatory content to the positive side of his Pyrrhonism; and it will allow a theory of meaning that will explain how we can successfully use words that fail to correspond to any ideas. And Hume would never have had to argue for this understanding of the normative force of custom. It would have been obvious to any of his contemporary readers, even if it is not to us. That is,

15. Whelan makes this point as well: [Hume] fails to discover logically necessary or self-​evident principles of right . . . and he is unable to find perceptible qualities of things or relations that determine their virtuous or vicious character. In this sense nature indeed provides no standards. We nevertheless make evaluative judgments, just as we draw causal inferences, and the lack of rational grounding does not normally—​even in the case of the skeptic—​ prevent our making them with conviction; genuine and total suspension of belief in morals, just as with respect to causation, is impossible in life. Faced with this state of affairs, Hume, in a manner common to many skeptics, turns his attention from epistemology to the psychology of belief, and he finds that moral judgments arise from the distinctive feelings of approval or disapproval that normal people experience upon considering certain objects or circumstances. Moral like causal judgments, that is, depend on some deeply rooted and nonrational propensity of our nature, and indeed it is this origin that gives the most basic of them their usual strength, rendering them immune, in practice, to the unanswerable doubts that philosophers may propose. Hume likewise grants the importance of the role of custom—​in the sense both of individual mental habits and of social tradition—​in the realm of ethics that the earlier skeptics emphasized; but here too his doctrine is more complex: some moral qualities—​the “natural virtues”—​he finds to be the largely invariable and direct expressions of natural impulses, whereas others—​the “artificial virtues”—​depend in large degree on social conventions and tradition. (1985,  20–​21)

( 46 )  Methodological Preliminaries

it would have been something Hume would expect us to take for granted in all of our reasonings. This is the understanding of custom that frames the Treatise. I urge the reader to keep this understanding in view, along with the fact that Hume sees custom as habit and custom as convention as components of a single phenomenon. If we do so, we will see that the Treatise is a highly consistent, unified text, systematically exploring human psychology as conditioned by custom, and explaining our epistemic, linguistic, moral, and political practice in terms of custom so understood. We will also see that this treatment of custom is what binds Hume’s naturalism and his skepticism.

PA R T  I I

vwv Book II The Psychological Foundations

C H A P T E R  3

w

The Passions Some Basic Distinctions

1.  INTRODUCTION: THREE DISTINCTIONS THAT STRUCTURE THE DISCUSSION I claimed earlier that the Treatise, true to its name, is a book about human nature, or, to be more specific, a book committed to empirical claims about how the mind works–​a text in speculative cognitive science–​and therefore often more about what we would now call psychology than philosophy. And as I noted in the previous chapter, Book II is where the foundation of Hume’s psychology can be found, and we will see that Book I—​where many of the better-​known Humean theses are defended—​is much better understood with Book II in the background.1 So, we begin in Book II. Book II is structured like a sandwich, with the more general psychological work done in parts I and III, and the moral psychology in part II, give or take a bit. For the purposes of this exposition, I will begin with the bread, treating the general psychological framework in the present chapter, and 1. Hume quite clearly did not think that he should start with Book II, or he would have done so. He was more concerned to establish the methodological foundations of his science than his psychological foundations at the outset, reflecting legitimate worries about the justification of the entire enterprise. This may or may not have been the right strategy for his readership. But I hope that the present order of march will be more illuminating for contemporary readers and that it will bring aspects of the Hume’s psychology to center stage early in the story in ways that make the entire drama more intelligible than it is when they are left for so long in the wings. ( 49 )

( 50 )  Book II

turning to the moral psychology in the next chapter. Throughout, we will read Hume closely, paying careful attention to his language and specific formulations, which are far from casual. Hume’s account is structured around a set of distinctions. These are in broad brush inherited from Hutcheson, who in turn develops them on the basis of Shaftesbury’s account.2 But the details are often original with Hume. The first is the distinction between the direct and the indirect passions; the former are those that are reflexive, immediate responses to experiences; the latter require some additional objective content and the operation of the imagination. Here is how Hutcheson draws the distinction: We may consider what universal Distinction can be assigned between these Modifications, and the several Sensations above-​mentioned; and we shall scarce find any other than this, that we call, “the direct immediate Perception of Pleasure or Pain from the present Object or Event, the Sensation.” But we denote by the Affection or Passion some other “Perceptions of Pleasure or Pain, not directly raised by the Presence or Operation of the Event or Object, but by our Reflection upon, or Apprehension of their present or future Existence; so that we expect or judge that the Object or Event will raise the direct Sensations in us.” In beholding a regular Building we have the Sensation of Beauty; but upon our apprehending ourselves possessed of it, or that we can procure this pleasant Sensation when we please, we feel the Affection of Joy. (1969, 27–​28)

Hutcheson’s distinction is between a sensation and passion, and he reserves the latter term for those states to which Hume will refer as indirect passions. Hume’s distinction is more psychologically complex and nuanced than is Hutcheson’s, which is drawn simply on the basis of the difference between present immediate experience and anticipated future experience. So, we shall see, while Hume inherits Hutcheson’s general framework (including Hutcheson’s commitment to the view that it is the passions that are the proximate causes of behavior), the important details that are set in the framework are Hume’s own.3

2. Laird (1932) points out that Hume also owes a great debt to Malebranche for the general architecture of the theory, and in particular the idea that there are overarching universal principles that sort the passions into natural categories, with the distinction between pride and humility as the governing distinction. 3. It is often important to keep in mind what Hume is inheriting in order to appreciate how he frames a question or why he adopts the language he does. Hume is often innovating, but is almost always doing so within the context of a discourse he takes for granted.

The Pa s s i o n s  

( 51 )

The second major distinction, also inherited from Hutcheson (who, once again, refines it from one inherited from Shaftesbury) is that between the calm and the violent passions.4 Hutcheson allies this distinction closely to that between the direct and indirect passions, suggesting that the direct are often violent in virtue of the immediate relation of the cause of the passion to the passion itself, whereas the indirect are calm because they are caused by mere imaginings, as opposed to sensations. A very simple-​ minded reading of Hume, grounded in a cursory reading of Book I  and the distinction between impressions and ideas, might lead one to suspect that Hume would be on board with this extensional collapse of these two distinctions. But Hume is more nuanced and more complex here as well. Once again, his innovation is not in the distinction being drawn, but in the psychological account in terms of which he draws it. As we shall see when we turn our attention to Book I, this will give us cause to read the distinctions in that Book with more care as well. A third distinction, that is original to Hume, and that animates his account, is that between the cause and the object of a passion. In drawing this distinction, Hume introduces the idea of the intentionality of the passions and distinguishes intentionality from causality, opening the possibility for a psychology that is both naturalistic and cognitive. While Hume argues that the passions are the original springs that animate action, including cognitive action, and the most fundamental explanatory principles of our psychology, and while he characterizes that animation causally, on the model of the physical sciences, Hume limns the structure of the passions intentionally. For this reason his account of the passions is also suitable to form the basis of a theory of cognitive content. Hume, therefore, may be a proto-​behaviorist in terms of his associationism, his commitment to causal explanation, and his reliance on that aspect of custom we would call habit. Nonetheless, to the degree that he is, he is an ancestor not of the crude behaviorism of Watson or Skinner, which has no place for any explanatory role for content or for behavior described intentionally, taking it to be entirely replaced by causality. Instead, as I suggested earlier, he is the proper ancestor of Tolman, who insisted on a cognitive behaviorism in which images and cognitive maps, as well as verbal schemata, play crucial roles in explaining habitual associations, memory, affect, and learning. This is not to say that Hume anticipated all of the developments of twentieth-​century psychology, only to say that the trajectory on which he sets the philosophy of mind is one that may be more

4. See Hutcheson (1969, 60ff.).

( 52 )  Book II

interesting than it might appear, and to say that he develops the initial conceptual resources that make that trajectory possible, a trajectory that takes us to the behaviorist movement in its most sophisticated form.

2.  THE PASSIONS: CALM AND VIOLENT; DIRECT AND INDIRECT Hume opens his account of the passions by classifying them in terms of the distinction between impressions and ideas drawn in Book I. Impressions, there, are defined as the immediate effects of stimuli upon the mind, and ideas as their faded copies, at least as a first pass. Impressions then, are things that happen to us when we encounter the world, regardless of whether that world is external or internal; impressions are the immediate deliverances of sense, whether that sense is outer or inner.5 Importantly, impressions are things that happen to us, with respect to which we are passive. It is hence natural that the passions are treated in terms of impressions. Hume introduces the point as follows: Secondary, or reflective impressions are such as proceed from some of these original ones, either immediately or by the interposition of its idea. Of the first kind are all the impressions of the senses, and all bodily pains and pleasures: Of the second are the passions, and other emotions resembling them. [2.1.1.1, 275]

The passions are impressions of reflection. Passions therefore are things that happen to us, but passions—​unlike those that proceed from original impressions—​happen to us not (as both Shaftesbury and Hutcheson would have it)6 consequent immediately upon our interaction with external objects. Instead, Hume asserts, they proceed from the “interposition of its idea.” This distinction is not at all obvious, so let us imagine an example. I  now see my copy of the Treatise. It is green and produces, on Hume’s understanding of the perceptual process, a green impression—​a sensory

5. It is worth bearing in mind that Hume, in his introductions of impressions of reflection, is treating our experience of our own cognitive states on a sensory model, a treatment not unlike that common in Buddhist traditions that distinguish six, not five, sensory faculties, comprising not only the external senses, but an introspective sense faculty whose direct objects are our cognitive states (manas-​vijñāna). 6. Although the language of “original” and “secondary” impressions and the application of this language derives from Hutcheson. This is one more case in which we must note not just the origins of Hume’s language, and so his reference points, but also the very different ways that he uses the language he borrows from his predecessors.

The Pa s s i o n s  

( 53 )

impression. But I can, as I am doing now, introspect and become aware of that impression; in doing so, I experience an impression of reflection, an impression in inner sense of an impression conveyed to me by vision. That is the first kind of impression. While it is something that happens to me,7 it is not a passion. How do I interpose the idea, and what comes of that? Well, first I need the idea of the book, that representation I use in reasoning or in memory, which in some sense resembles the impression of the book, but lacks some of its force and vivacity—​that is, it is not by itself so efficacious in producing new cognitions. But I can introspect and become aware of that idea. And then I can form a new impression of reflection consequent upon that—​an introspective impression of an idea. Hume claims that passions are this kind of thing, impressions of ideas. So, when I see my book and experience joy—​a passion—​that joy is not, pace Hutcheson, directed immediately at the book. Instead, while it is directed on the book, it is mediated by the idea of the book. A passion, for Hume, is always therefore, in a specific sense, cognitive, not a mere reaction to a stimulus, as would be a mere sensation, even a sensation of pleasure, but a state that involves, but is not constituted by, a kind of judgment. This claim might appear to be in conflict with Hume’s statement that “a passion is an original existence, or, if you will, modification of existence, and contains not any representative quality, which renders it a copy of any other existence or modification” [2.3.3.5, 415]. But it is not. This is clear from the conclusion he immediately draws: “’Tis impossible, therefore, that this passion can be oppos’d by, or be contradictory to truth and reason” [2.3.3.5, 415]. Hume here emphasizes the character of the passion as an affective response, and not simply a judgment. He makes it clear in the subsequent discussion [2.3.3.6–​7, 415–​17] that passions also involve judgments, and it is those judgments that can be assessed rationally. This is why neither the book itself, nor the impression of the book, nor the impression of the impression could count as a passion; the passion must be an impression of the idea, and so is necessarily cognitive in that sense as well. Moreover, the judgments that are involved in the passion necessarily take their objects conceptually. That is, we are joyful, angry, afraid, or envious for reasons, and reasons are expressed in terms that evoke universals,

7. It is not an idea that I form, for instance, although I might also have an idea of my copy of the Treatise, and consult it in memory when I  am looking for the book, having misplaced it. So, it is not a faded copy of that impression produced by the imagination, but rather a vivid awareness of the inner experience produced in inner sense [1.3.8.15–​16, 106].

( 54 )  Book II

not only particulars. Later, we will encounter Hume’s attack on the opposition of reason and passion; here we have a different reason for worrying about that duality: passions are responsive to reasons, and it is the interposition of ideas that makes that the case. This doctrine is not the denial that impressions of reflection unmediated by ideas are irrelevant to the passions. That would be odd. Hume remarks: Bodily pains and pleasures are the source of many passions, both when felt and consider’d by the mind; but arise originally in the soul, or in the body, whichever you please to call it, without any preceding thought or perception. [2.1.1.2, 276]

These events may cause passions, but they are not passions themselves: we must distinguish the pain that causes my anger from the anger; the pleasure that causes my joy from the joy. When my face smarts at your slap, the anger is because the idea of the slap is an idea of a harm; when I am joyful at finding my copy of the Treatise, the joy is because the Treatise is the kind of thing that gives me pleasure. The explanation of the passion thus necessarily adverts to an idea, even if the initial cause is an impression (and this is one important respect in which the laws of human nature that Hume envisions are necessarily intentional in character, as Alanen [2005] points out). And this is true whether or not we are introspectively aware of the role of the idea in the genesis and structure of the passion. Hume’s point here is psychological, not phenomenological. Hume turns next to Hutcheson’s distinction between the calm and violent passions. First note that while he uses Hutcheson’s language, he reconstructs Hutcheson’s distinction—​a distinction drawn merely in terms of the intensity of the passion—​constructing his own distinction not as one of degree, but as a distinction of kind. Here is a place where words are important. Violent connotes violation, a deviation from the normal course of things. Hume is directing us to the fact that our own trajectory through life is not generally perturbed by those passions directed at phenomena that do not concern us directly, but that trajectory is often changed by those passions directed at phenomena that intimately concern us. This concern for psychological dynamics, rather than simply degree, sets Hume apart from his predecessors in this tradition. The reflective impressions may be divided into two kinds, viz., the calm and the violent. Of the first kind is the sense of beauty and deformity in action, composition, and external objects. Of the second are the passions of love and hatred, grief and joy, pride and humility. [2.1.1.3, 276]

The Pa s s i o n s  

( 55 )

The calm passions (still described as reflective impressions) for Hume are those that involve an aesthetic or ethical assessment of external phenomena; the violent ones are those that are more immediately concerned with ourselves, and with the assessment of the impact of the cause of the passion on us. So, when I judge that someone is kind and feel admiration for her character, or that she is cruel, and develop contempt, these are calm passions, and may not stir me to uncharacteristic action; this may be a standing, disinterested judgment, and having made it, I go about my business interacting with her consistently with that judgment. No matter how intensely I feel it, it does not significantly alter the arc of my life. When I judge that someone has insulted or harmed me, on the other hand, I experience anger, a violent passion, or when someone has assisted me, I experience affection, another violent passion, likely to stir me to action. These cause me to stop what I am doing and to take unplanned action, to deviate from my current behavioral trajectory. Ardal (1966) and Loeb (1977), despite their differences regarding the classification of the passions, agree that, according to Hume, the difference between calm and violent passions is only one of degree, and not of kind. I think that this is a mistake. First of all, Hume quite explicitly says that these impressions “may be divided into two kinds.” He could have talked in terms of degrees, which would have been the most natural way to distinguish between the calm and the violent, and which would have followed Hutcheson more closely; he did not. Second, there is a very clear distinction in kind between the two lists he provides as examples. The first—​the calm passions—​are standing modes of affection; they endure for long periods of our lives. Moreover, they are passions that arise from the contemplations of things whose connection to us may be indirect, and which reflect our sense of how they affect others. That is, they are nonegocentric, directed at beauty or virtue, considered in a disinterested way. Hume will flesh this account out further in his discussion of moral virtue in Book III. The second—​the violent passions—​are those that directly involve us, that are essentially egocentric in character.8 8. So Loeb compounds his erroneous understanding of the structure of the passions when he writes, On Hume’s principles, no passion, whether indirect or direct, can generally be calm . . . [A]‌lthough calm emotions are defined by Hume as those generally of low emotional intensity, he gives an account of the psychological (associationist) principles producing indirect and direct passions such that these prove to be violent. (1977, 398) Hume is in fact, as we have seen, absolutely explicit in asserting that on his view there are calm as well as violent passions, and he defines the calm passions not in

( 56 )  Book II

This is a difference in kind, not in degree, and this classification is both original to Hume and essential to his moral theory, which will concern primarily the cultivation of the calm passions.9 This brings us to the second division among the passions, that between the direct and the indirect: When we take a survey of the passions, there occurs a division of them into direct and indirect. By direct passions I  understand such as arise immediately from good or evil, from pain or pleasure. By indirect such as proceed from the same principles, but by the conjunction of other qualities. . . . under the indirect passions I comprehend pride, humility, ambition, vanity, love, hatred, envy, pity, malice, generosity, with their dependants. And under the direct passions, desire, aversion, grief, joy, hope, fear, despair and security. [2.1.1.4, 276–​77]

The definition of direct passions must also be handled with care, for it could be taken simply to be an echo of Hutcheson, and to omit the mediation of the idea that, as we have seen, Hume claimed was the hallmark of a passion. We see that this cannot be correct, not only in virtue of the obvious inconsistency it would introduce within the span of a few pages, but also in the contrast with the indirect passions defined in the next sentence. The key to the distinction is that while the two kinds of passions “proceed from the same principles,” those noted above, the indirect passions proceed “by the conjunction of other qualities.” Hume does not tell us in this context what these other qualities might be. But in section III he makes it plain: But tho’ the causes of pride and humility be plainly natural, we shall find upon examination, that they are not original, and that ’tis utterly impossible they

terms of “low emotional intensity” but rather in terms of their causes and their relative stability. They are not violent precisely because of that relative stability and their relatively disinterested character; that is, for those reasons, they do not cause us to violate the normal trajectory of our lives, as do the violent passions. It is curious that Loeb adopts this position, since when he turns to the moral sentiments directly, he effectively defines them in terms of their mediation by sympathy, and their tendency to arise through contemplation and reflection (1977). The error is further compounded when Loeb argues that Hume takes the moral sentiments not to be passions at all (passions all being violent, while moral sentiments are calm on Loeb’s view), hence effectively rupturing the link between Books II and Books III (402). 9. The idea that specifically moral responses are passionate, while at the same time in an important sense disinterested or nonegocentric, is another one that Hume shares with philosophers in the Buddhist tradition. One is put in mind of the brahmavihāras, or divine states, so central to Buddhist moral theory, and in particular those of metta (love), muditā (sympathetic joy) and upekkhā (nonegocentricity) that are taken as the hallmarks of a moral point of view (Garfield 2015).

The Pa s s i o n s  

( 57 )

shou’d each of them be adapted to these passions by a particular provision, and primary constitution of nature. Beside their prodigious number, many of them are the effects of art, and arise partly from the industry, partly from the caprice, and partly from the good fortune of men. Industry produces houses, furniture, cloaths. Caprice determines their particular kinds and qualities. And good fortune frequently contributes to all this, by discovering the effects that result from the different mixtures and combinations of bodies. ’Tis absurd, therefore, to imagine, that each of these was foreseen and provided for by nature, and that every new production of art, which causes pride or humility; instead of adapting itself to the passion by partaking of some general quality, that naturally operates on the mind; is itself the object of an original principle, which till then lay conceal’d in the soul, and is only by accident at last brought to light. [2.1.3.5, 281]

The relevant distinction is between those causes that are original and those that are not, although all of these causes are natural. This distinction, and its relationship to the property of being natural, tells us quite a lot about Hume’s account of human nature and approach to the topic. To be original in this sense is to be innate in the sense in which that term is used in contemporary cognitive science—​to be determined by the genetic endowment of the organism.10 Note that this is a narrower property than that of being natural. Events or objects can be natural causes of pride or humility without being original causes thereof, and that is because Hume recognizes that we are naturally social organisms, and that some of the causes of pride and humility (and in general of indirect passions, for this is the class he is discussing) are socially constructed, through industry, or even by caprice.11 Laird claims that Hume’s account of pride and humility is simply a repetition of Mandeville’s (Laird 1932, 195, 211). But while Hume concurs with Mandeville regarding the naturalness of these passions, and in the general account of the judgments that motivate them, Mandeville’s theory is in the

10. There is a clear antecedent for this distinction and for the account of the relation of custom to the passions in Shaftesbury: “That which is of original and pure Nature, nothing besides contrary Habit and Custom (a second Nature) is able to displace” (1964, 2:25). 11.  Hume emphasizes his understanding of what it is to be natural in a letter to Hutcheson: I cannot agree to your Sense of Natural. Tis founded on final Cause; which is a Consideration, that appears to me pretty uncertain & unphilosophical. For pray, what is the End of Man? Is he created for Happiness or for Virtue? For this Life or for the next? For himself or for his Maker? Your Definition of Natural depends upon solving these Questions, which are endless, & quite wide of my Purpose. I  have never call’d Justice unnatural, but only artificial. (Letter 3, to Francis Hutcheson, in Hume 1932, 33).

( 58 )  Book II

end individualistic, and Hume’s is social. That is, while Mandeville allows that the individual passions are modulated by social interaction, Hume takes many of them to be constituted by social context. The social dimension, as Norton (1982, 39) notes correctly, derives in part from Shaftesbury. Shaftesbury writes, “In short, if generation be natural, if natural affection and the care and nurture of the offspring be natural, things standing as they do with man, and the creature being of that form and constitution he now is, it follows that society must also be natural to him” (1964, I, 78). Hume joins Mandeville’s insight that the passions lie at the foundation of human nature with Shaftesbury’s insight that our social nature is as much a part of our nature as is anything innate. This union leads to the social account of the passions that is Hume’s greatest contribution to the moral psychology of the eighteenth century. To miss that advance is to miss the heart of the Humean project.12 The fact that certain passions arise from social context does not, in his view, make them less natural. Social laws and social explanation are as natural as biological laws and biological explanation. A crucial insight of the Treatise is that the distinction between nature and artifice does not track the distinction between the biological and the social; instead, human nature arises at the intersection of the psychological and the sociological.13 This insight comes back in the discussion of justice in Book III, as we shall see in ­chapter 11. The direct passions, then, are those we can explain by original principles alone, those for which the links between the cause and the passion is innately determined;14 the indirect are those for which that causal link is in part socially determined. This yields a two-​dimensional matrix for classifying the passions: they may be calm or violent; they may be direct or indirect. Each of these dimensions marks a distinction in the

12. See Alanen (2006) for a nice discussion of the social dimensions of the passions and moral sentiments and Cohon (2008) for more on the relationship between Hume’s account and those of Hobbes and Mandeville on the one hand and Shaftesbury and Locke on the other. 13. But we should be careful with this metaphor, as for Hume the social and the psychological are not independent dimensions, but rather are completely enmeshed with one another. As Ainslie (2005) points out, one of the important psychological forces that mediates the connection between the original and the social is natural sympathy; it is our original tendency to sympathize with others that allows the coordination of affective reaction that in turn enables the collective constitution of value and of the social structures that enable it. See also Cohon (2006) for an astute discussion of the role of sympathy in the artificial virtues and the stability of society. 14.  At least under the right description. Desire may be direct, but its immediate objects may be socially constructed (e.g., money or status). But these are desired as goods, and, as Hume makes clear [2.1.1.4], the connection between desire and the good, like that between aversion and evil (in whatever form that might take), is original.

The Pa s s i o n s  

( 59 )

structure of the passion. All are mediated by ideas; all are impressions of reflection; all are explained causally, but nonetheless under intentional descriptions.

3.  OBJECTS AND CAUSES: PASSIONS AND THE SELF We now come to one of Hume’s great insights about the structure of the passions, the distinction between the cause and the object of a passion, a distinction that leads us directly into Hume’s most important and detailed conception of the self. That account occurs in the context of this crucial distinction, and it is important to note—​and this will be especially important when we come to discuss in c­ hapter 8 the relation between this account and that of Book I—​that the self here appears on the side of the object, not the cause. Hume introduces the account of the distinction as follows: ’Tis evident, that pride and humility, tho’ directly contrary, have yet the same object. This object is self, or that succession of related ideas and impressions, of which we have an intimate memory and consciousness. . . . Whatever other objects may be comprehended by the mind, they are always consider’d with a view to ourselves; otherwise they wou’d never be able either to excite these passions, or produce the smallest encrease or diminution of them. When self enters not into the consideration, there is no room either for pride or humility. [2.1.2.2, 277]

The causes of pride are those things we take to be good; those of humility are those we take to be ill.15 So, clearly, these passions differ with respect 15. Hume is clearly having a bit of fun with the church here as well, regarding pride as a positive force in our moral lives, and humility as something to be avoided, thus inverting their valence in many a Christian sermon. But he intends a clear point about moral psychology:  pride—​understood as a positive self-​regarding attitude—​can inspire and reflect virtue; humility—​as a negative self-​regarding attitude—​can reflect and discourage vice. is A bit later in the Treatise he remarks: There may, perhaps, be some, who being accustom’d to the style of the schools and pulpit, and having never consider’d human nature in any other light, than that in which they place it, may here be surpriz’d to hear me talk of virtue as exciting pride, which they look upon as a vice; and of vice as producing humility, which they have been taught to consider as a virtue. But not to dispute about words, I observe, that by pride I understand that agreeable impression, which arises in the mind, when the view either of our virtue, beauty, riches or power makes us satisfy’d with ourselves: And that by humility I mean the opposite impression. ’Tis evident the former impression is not always vicious, nor the latter virtuous. (2.1.7.8, 297–​98) See also Whelan (1985, 150) for a similar observation.

( 60 )  Book II

to their causes. But pride and humility are about the self.16 When I  am proud of my dog’s behavior, I  am proud because he is my dog. My dog’s achievements may cause the pride, but the pride is in myself as his owner. I might admire your dog, but I can’t be proud of him. When I am humbled by my cat’s murder of a finch, the humility has to do with the fact that she is my cat. I may be horrified by the predations of your cat, but they cannot humble me, for she is your cat. Humility and pride are fundamental to Hume’s account of the passions because they form a psychological bedrock of self-​regarding attitudes that take us from mere judgment to passion—​to an impression of reflection that can stir us to action. In general, although not all passions take the self as their object (consider love, hatred, anger, and gratitude, for example), Hume takes it that for a passion—​as opposed to a disinterested judgment—​to arise, a relation to self is necessary. So, when I am angry that your ox gored the postman, that anger, despite the fact that the ox is yours and the postman is a relative stranger to me, reflects the fact that I am offended by the act.17 I might be generally disapproving of the wanton behavior of cattle, believing it to be a disadvantage to any community in which they roam, but without cultivating anger. When I form this judgment, I see their behavior as having nothing to do with me, personally. But when a postman in my community, to whom I have a relation, is injured, I, as well as he, suffer. Without that sense of personal offense, I might condemn the act, but it could not arouse the passion of anger.18

16. Hume never abandons this model of the passions. In the Dissertation he writes, “In these two sets of passion, there is an obvious distinction to be made between the object of the passion and its cause. The object of pride and humility is the self:  The cause of the passion is some excellence in the former case; some fault in the latter. The object of love and hatred is some other person: The causes, in like manner, are either excellencies or faults” (2007, 2:2). And we should note that although “the object of love or hatred is some other person,” that person must have some relation to ourselves in order for a judgment about them to be converted to a passion. John Connolly (personal communication) notes that had Hume thought of generalizing this point beyond the passions, he might have moved closer to Kant on the transcendental unity of apperception. 17. We could change the details and get a different result. I might hate the postman, or feel affection for him, for instance, and these relations might modulate the response. But the point here is just that for a passion to arise at all, some relation to oneself is essential. 18.  It is interesting here to compare the account of anger and its relation to the positing of a self and its centrality in Śāntideva’s How to Lead an Awakened Life (Bodhicāryāvatāra). Śāntideva also argues that when anger arises, it immediately leads to a strong sense of a personal self and of personal offence, as well as a strong sense of the selfhood of the object of the anger; and he, like Hume, distinguishes the cause from the object of the anger, although he does not use that vocabulary. Moreover, he argues that there is no real entity that corresponds to the self posited in reaction to the passions.

The Pa s s i o n s  

( 61 )

This is one of the reasons that in order to arouse passion in an audience for what is happening remotely (say rage at one nation attacking another on a distant continent, or joy in the achievement of a hitherto unknown athlete in Olympic competition), propagandists and other advertisers are concerned to generate a sense of identification with the victims, using such devices as photographs and biographical details, not simply relying on a disinterested judgment about the kinds of acts they have performed. In generating this commitment, while they need not lead us to take ourselves as the objects of the passions—​although when we become proud or ashamed of the acts of our compatriots that may be the case—​they must lead us to have some sense of relation to the objects of those passions. I remarked previously that Hume’s Book II account of the passions gives them a central role in the construction of the idea of ourselves. But note that nothing has been said so far to indicate the content of the idea of the self, or the reality of that to which it is supposed to correspond. Those who charge Hume with inconsistency for maintaining the reality of the self in Book II while denying it in Book I, or, more subtly, with maintaining in Book I that we have no idea of a self, while supposing in Book II that we do, should attend more carefully to how gingerly Hume handles this question in this context. First, there is the question of causation: But tho’ that connected succession of perceptions, which we call self, be always the object of these two passions, ’tis impossible it can be their cause, or be sufficient alone to excite them. For, as these passions are directly contrary, and have the same object in common; were their object also their cause; it cou’d never produce any degree of the one passion, but at the same time it must excite an equal degree of the other; which opposition and contrariety must destroy both. [2.1.2.3, 277–​78]

First, note what Hume says that we call self:  a connected succession of perceptions. That is all that is the object of the passions, not a soul or a substance in which those perceptions inhere. And that succession is causally insufficient to give rise to the passions. Hume offers one clear reason for that: if it were the cause of one passion, it could not be the cause of a contrary passion. Hume could have gone further and noted that successions are the wrong kinds of things to do the causation. Perceptions cause other perceptions, but successions are extended collections, not events. And if there is no self beyond that succession, there is no other sense in which the self could be causal.19 But the passions do not only take the self as an object 19. One might compare Hume’s point here to that Dignāga makes in Investigation of the Percept (Duckworth et al. 2016). Dignāga argues there that external objects cannot

( 62 )  Book II

after the fact; Hume assigns them a crucial role in the constitution of the conventional identity we call self. That is: We must, therefore, make a distinction between the cause and the object of these passions; betwixt that idea, which excites them, and that to which they direct their view, when excited. Pride and humility, being once rais’d, immediately turn our attention to ourself, and regard that as their ultimate and final object; but there is something farther requisite in order to raise them: Something, which is peculiar to one of the passions, and produces not both in the very same degree. The first idea, that is presented to the mind, is that of the cause or productive principle. This excites the passion, connected with it; and that passion, when excited, turns our view to another idea, which is that of self. Here then is a passion plac’d betwixt two ideas, of which the one produces it, and the other is produc’d by it. The first idea, therefore, represents the cause, the second the object of the passion. [2.1.2.4, 278]

The psychological account here is delicate, and it is important to note that the passions have a role in constituting, not only referring to, the self. Something stimulates a passion; perhaps the sight of a rope on the ground that I take to be a snake excites fear. The fear, however, even though it is caused by the rope, is not of the rope, but of an imagined snake. And it is not even of that imagined snake simpliciter; it is of what that snake might do to me. The fear, that is, directs my attention to my vulnerability. That idea—​ the idea of myself as a vulnerable being—​is created by the passion, which we can then understand as placed “betwixt” the idea of a nonexistent snake and that of a nonexistent self, an idea caused by impressions of passions and confused with that of a substratrum of experience. As I foreshadowed in c­ hapter 1, Hume often reminds us that we think we have ideas that we in fact cannot have, as no impression could ever be what we perceive, because a percept must be both the cause and the apparent object in perception; but all of the causality in any object is exhausted by the momentary tropes in which it consists, and they are not what appears in perception. Hume is not making that point, and indeed his account of causation is more sophisticated than Dignāga’s, allowing for causal explanations linking phenomena at much higher levels of description, as when our wealth or power is a cause of pride. While Dignāga would agree that wealth or power can cause pride, his own account of causation as a relation exclusively among momentary partless events at the bottom of the ontological hierarchy makes it hard for him to make sense of that; Hume, by seeing causation as constituted by explanatory regularities, is freed from this ontological straitjacket. Nonetheless, if a succession were to be the initial cause of the passion, it would have already to be regarded as the kind of thing of which we could have a perception, in which we could take a concern, and so it would have already to be regarded as a natural whole; but that would beg the question in this context.

The Pa s s i o n s  

( 63 )

correspond to them. A sense of the opacity of the mind—​of our inability to know the contents of our own thoughts—​lies at the bedrock of his system, and itself sets the need for the “application of experimental philosophy to moral subjects” [introduction 8, xvi]. But it is then incumbent upon him to explain not only the origins of the ideas we have, but also the possibility of thinking that we have ideas that we do not have. Indeed, the latter seems to be manifestly absurd, as Descartes or Berkeley would have argued: to believe that you have an idea is to have that idea. Hume, however, wisely dissents. He agrees with his predecessors and contemporaries (perhaps unwisely) that conceivability is a guide to possibility. In Book I he argues: Whatever can be conceiv’d by a clear and distinct idea necessarily implies the possibility of existence; and he who pretends to prove the impossibility of its existence by any argument deriv’d from the clear idea, in reality asserts, that we have no clear idea of it, because we have a clear idea. ’Tis in vain to search for a contradiction in any thing that is distinctly conceiv’d by the mind. Did it imply any contradiction, ’tis impossible it cou’d ever be conceiv’d. [1.2.4.11, 43]

Nonetheless, many a mathematician claims to have had the idea of a proof of the parallel postulate or of a squared circle. And Hume is certainly aware of this. Moreover, in order even to deny that they had such ideas despite their insistence to the contrary, it would appear that we would have to have the idea of the thing we denied them. If not, how could we understand either their position or ours? Hume’s reply to this is grounded in his nominalism, a commitment that is for this very reason essential to his system. When we deny that an early mathematician has an idea of a squared circle, or a proof of the parallel postulate, Hume must say, we deny not of a particular idea that they have it, but rather that there is any idea corresponding to those words; when the deluded mathematician believes that she has such an idea, she really believes falsely that the words she speaks make sense. She falsely believes that the ideas she has of the words in fact constitute the idea of an abstract object. The origin of the confusion then is semantic: it is the confusion of words with their putative referents. Hume’s commitment to the link between conceivability and possibility may be one of the deepest inconsistencies in Hume’s system, and we must be alert for illicit appeals to the conceivability criterion of possibility in the Treatise. Let us be clear about why Hume is not entitled to this principle by his own rights, however obviously true he takes it to be. Hume, as we have seen, is committed not only to the possibility, but to the pervasiveness, of cognitive illusion (the nominalist pseudo-​idea principle).

( 64 )  Book II

We are often in error not only about the external world, and about the principles that actuate our own cognition, but even about the ideas we take ourselves to have. We often confuse, Hume argues, the customary use of a word with the possession of an idea corresponding to it. He deploys this principle to great effect in his discussion of our putative ideas of external objects, the self, causation, and substance, to name but a few salient instances. And he insists that our false self-​attribution of ideas in general reflects the fact that we can find no impression from which the ideas we take ourselves to have could derive. These principles—​at the heart of Hume’s project—​give two reasons to reject the conceivability-​possibility link Hume so blithely adopts.20 First, the possibility of cognitive illusion renders the conceivability-​possibility principle, even if cogent, empty in practice. For, since we can never be certain that we can conceive that which we take ourselves to conceive, any inference to possibility is at best dubious. But even worse, the principle must, by Hume’s own lights, be nonsense. For if possibility is to mean anything beyond conceivability, we must have some impression of possibility distinct from that of what we conceive as the source of the idea, and have account of how these two ideas are connected. And Hume must concede that there can be no such impression, since on his view our only guide to possibility is conceivability. Therefore, on Hume’s own view, there can be no two distinct ideas of conceivability and possibility, one of which is a guide to the other. With this understanding of what it could be to think that we have an idea that we do not have, let us now return to the confusion of ideas that leads us falsely to believe that we have an idea of a self: Hume is not arguing here that the passions give us the idea of a succession of impressions. We get that idea from the impression of that succession. Nor do they give us the idea of a substantial self that is the continuing substratum of that succession; he argues in Book I that there is no such idea. Nonetheless, the passions do cause us to confuse the idea that we do have (that of the succession of distinct ideas) with the idea that we do not (that of the substantial substrate), and hence explain the otherwise mysterious mistake with which we are all charged in Book I [1.4.3.3–​4, 220]. My fear causes me to posit an object for that fear, and a succession is an unfit object. I therefore take there to be a referent of I beyond that succession,

20.  One he inherits from Descartes and Berkeley, and one that is also uncritically adopted in much contemporary metaphysics—​it is, for instance, the foundation of the “Canberra plan.”

The Pa s s i o n s  

( 65 )

and take it as the intentional object of my fear.21 But that is a semantic confusion; a confusion of my mastery of the pronoun with an idea of an imaginary referent. Without the passions, Hume suggests, we might never have any sense that we have an idea of a self. The idea that we do have, with which we are apt to confuse it, is, however, complex and friable, as we soon discover.22 In order to understand this idea of myself and Hume’s account of its origin, we will need to take a few detours. First, consider this remarkable passage a bit later in Book II: ’Tis evident, that the idea, or rather impression of ourselves is always intimately present with us, and that our consciousness gives us so lively a conception of our own person, that ’tis not possible to imagine, that any thing can in this particular go beyond it. [2.1.11.4, 317]

Our analysis will require us to square this with what Hume says fewer than seventy pages before: There are some philosophers, who imagine we are every moment intimately conscious of what we call our self; that we feel its existence and its continuance in existence.  .  .  . The strongest sensation, the most violent passion, say they, instead of distracting us from this view, only fix it the more intensely, and make us consider their influence on self either by their pain or pleasure. . . . Unluckily all these positive assertions are contrary to that very experience, which is pleaded for them, nor have we any idea of self, after the manner it is here explain’d. For from what impression cou’d this idea be deriv’d? This question ’tis impossible to answer without a manifest contradiction and absurdity. . . . It must be some one impression, that gives rise to every real idea. But self or person is not any one impression, but that to which our several impressions and ideas are suppos’d to have a reference. [1.4.6.1–​2, 251]

21. And of course it can be an intentional object despite its nonexistence, like the snake, and even despite its impossibility, like the proof of the parallel postulate I hope to discover. Intentional objects need not be things of which we have genuine ideas, if we are good nominalists. 22. See Qu (2017) for a very different reading both of Hume’s Book II account of the self, and of the nature of the illusion that we have an idea of the self. Qu argues that the Book II self is a “bundle of dispositions” (644). My principal reason for disagreeing with Qu’s analysis is that he argues that the passions are too transient to give rise to the requisite idea, failing what he calls the “durability constraint.” I  think that this misses the role that the standing, calm passions have in this account.

( 66 )  Book II

This juxtaposition poses interpretative challenges. In Book II, Hume tells us that we have an idea and an impression of ourselves that is always with us and that it is the most vivid impression we can possibly imagine; in Book I, he argues that we have no idea of a self, because it would be a “manifest contradiction and absurdity” to have any impression that could correspond to it. In Book I, he ridicules philosophers who argue that “the most violent passion[s]‌, . . . instead of distracting us from this view, only fix it more intensely, and make us consider their influence on self either by their pain or pleasure,” and in Book II, he appeals directly to the passions as giving rise to the idea. If we are to understand Hume’s account of the passions and their relations to our sense of ourselves, we must explain how these two passages fit together, for each is essential to Hume’s project. In order to do so, I will appeal to a distinction at which Hume hints, but which he does not draw consistently, and the way this distinction is deployed by Candrakīrti, who also explicitly rejects the idea of a substantial self, but defends the reality of a conventional person, as well as upon Hume’s metaphor of the self as a theater (although one “that must not mislead us” [1.4.6.4, 253]).23 Let us, following Candrakīrti, distinguish between a self and a person.24 The self is the substantial, persistent entity—​the agent, the subject—​with which Hume suspects “some philosophers” to be so intimately acquainted, and the very idea that he attacks in Book I. Let us use person—​mindful of its etymology in the theater as a mask or a persona that one dons in assuming a role—​to refer to that which the passions enable us to construct in its stead in Book II. Armed with this distinction, and with attention to the very different contexts in which these two discussions occur, we can sort out both the relationship between these two discussions and the way that the passions give rise both to an impression and an idea of ourselves. Note that Hobbes had already remarked on this etymology, and that Locke had argued that the term person is forensic, not metaphysical; so neither this distinction nor a conventionalist understanding of personhood is original to Hume. In the Book I  discussion (to which we will return in more detail in ­chapter 8) Hume is keen to show that we have no “idea of self, after the manner it is here explain’d” [1.4.6.2,  251] (emphasis mine). For there he is talking about a self, and that self is described in highly individualistic terms, as having an existence independent of and prior to its attributes,

23. For the details of Candrakīrti’s account see Garfield (2015). 24. Candrakīrti uses ātman and pudgala to draw this distinction.

The Pa s s i o n s  

( 67 )

experiences, and the customs and relations into which it enters. And it is regarded as a substance that persists with an intrinsic identity independent of its attributes. Hume, as we will see in ­chapters 6–​9, will argue that we cannot so much as conceive of such an entity; that when we use these words, no ideas correspond to them.25 In this context, he points out that there is nothing of which we have introspective or perceptual experience that could answer to this description. Hume himself notes here that in order to understand the idea we confuse with that of a substantial self, we must “distinguish betwixt personal identity, as it regards our thought or imagination, and as it regards our passions or the concern we take in ourselves” [1.4.6.5, 253]. He thus indicates from the outset that these two discussions are related and that they involve contrasting ideas. Book I is concerned with the idea of the self. In Book II, by contrast—​ the context of the present discussion—​ Hume is discussing the person, and the origins of our respective ideas of the persons we are. In that context, we are discussing our passions, and our relationships to one another in the context of which those passions arise. Love and hatred require relations to others; even the self-​regarding passions of pride and humility most often take as their causes social goods (riches, power, houses, accomplishments, etc.) and presume the survey—​ at least in our imagination—​of others. My pride in my dog requires that I imagine your admiration; my humility regarding my cat involves my anticipation of your disapproval. Experiencing these passions and entering into these relationships constitutes us as persons—​as dramatis personae—​in the theater of our community. They make us who we are and define our characters. Awareness of ourselves—​of the masks we and our fellows carve for us in our joint affective lives and customary interactions—​creates an impression—​an impression of reflection—​of our constructed, collectively constituted identities, an impression that “is always intimately present with us, and that . . . gives us so lively a conception of our own person.” This is not the impression of a substance, of an unchanging self that is prior to or independent of our experiences and relations, but of one that is constituted by them. There is

25. Don Baxter (personal communication) points out that while no consistent ideas can correspond to such words, “confounded ideas” might. But this seems inconsistent with the conclusion that Hume draws when he says, “In order to reconcile which contradiction [that between the identity we ascribe to a substance and the diversity of the ideas we have] the imagination is apt to feign something unknown and invisible, which it supposes to continue the same under all these variations; and this unintelligible something it calls substance” [1.3.4.4. 220]. This unintelligible something is not an idea, but a feigned idea.

( 68 )  Book II

hence no contradiction between the rejection of an idea of a self in Book I on the grounds that there can be no impression of one and the affirmation of an idea—​and indeed an impression—​of a person in Book II. That person is the object of the passions of pride and humility.26 Hume points out immediately that the things of which we can be proud (or humble) are endless in their variety: A man may be proud of his beauty, strength, agility, good mein, address in dancing, riding, fencing, and of his dexterity in any manual business or manufacture. But this is not all. The passions looking farther, comprehend whatever objects are in the least ally’d or related to us. Our country, family, children, relations, riches, houses, gardens, horses, dogs, cloaths; any of these may become a cause either of pride or of humility. [2.1.2.5, 279] 26.  Ainslie emphasizes Hume’s psychological account of this role of the passions in constituting our identities as persons in a social context when he argues that “our minds are configured” for “recognizing existential connections,” and that the indirect passions are the mechanisms for this attunement (1999, 480). He also draws our attention in this context to the deep analogy between Hume’s treatment of the mechanisms that enable us to engage in causal inference and those that mediate our responsiveness to the properties that our constitutive of our identities as individuals in society with others. Ainslie shows that we are tuned to respond to certain properties of ourselves and others, and to certain relationships, and those attunements—​mediated by the passions, not the deliverances of reason—​constitute and reveal those properties we take to be essential to our personae and those of others, just as we are psychologically tuned to respond to certain kinds of regularities by attributing necessity to them, constituting and revealing causal relations not through reason, but through innate psychological dispositions (1999, 487ff.). This reading of the sense of self constituted by the passions in Book II as essentially social, as opposed to the individualistic sense of self dismissed in Book I, gains additional support if we consider this remark Hume makes in the course of his discussion of love and hatred: Those, who take a pleasure in declaiming against human nature, have observ’d, that man is altogether insufficient to support himself; and that when you loosen all the holds, which he has of external objects, he immediately drops down into the deepest melancholy and despair. . . . To this method of thinking I so far agree, that I own the mind to be insufficient, of itself, to its own entertainment, and that it naturally seeks after foreign objects, which may produce a lively sensation, and agitate the spirits. . . . Hence company is naturally so rejoicing, as presenting the liveliest of all objects, viz., a rational and thinking Being like ourselves, who communicates to us all the actions of his mind; makes us privy to his inmost sentiments and affections; and lets us see, in the very instant of their production, all the emotions, which are caus’d by any object. [2.2.4.4, 352–​53] Here Hume emphasizes that the role of company and social intercourse is not simply amusement, but also the presentation of the passions that allow us to construct ourselves in company. See also Rorty (2006) for a similar account of the mechanisms by which pride and other passions constitute the self in a manner entirely consistent with the rejection of the idea of self in Book I, and for illuminating reflections on the relationship between this project and Hume’s subversion of Christianity.

The Pa s s i o n s  

( 69 )

When I am proud of my dog, but also humiliated by my address in fencing, I represent myself as a character related both to dog and to sabre. There is nothing more to the idea of myself constituted by these passions than these relations, though; I  am a character defined by them, not a subject existing independent of them. On the other hand, while I have ideas of the achievements of my dog and my failure with my sword, and while I have impressions of the passions they engender, if I take myself to be an independent substance to which they are related, that is something of which I have no idea at all. The explanation of how I may come to believe that I am such a self despite having no such idea will wait until we return to the discussion in Book I in ­chapter 8; demonstrating how I construct the idea I do have of myself as a person is the central task of this section of Book II. The mechanism for that construction, we have seen, is provided by the passions, and the object constructed is a person suitable to be their object. In this section, Hume’s task has simply been to provide this naturalistic, a psychological, account of the origin of that idea, which, we will see, forms the foundation of the Book I account of the illusion that we have an idea of a self, and of the illusion that we are selves. Hume argues that the relation of the passions to the self is far from accidental: ’Tis evident in the first place, that these passions are determin’d to have the self for their object, not only by a natural but also by an original property. No one can doubt but that 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. [2.1.3.2, 280]

In virtue of the constitution of the person by the passions, and the tendency to conflate that person with a self, we are designed by nature to suffer from a double illusion—​the illusion both that we have an idea of a self and that we are that to which that idea is adequate—​just as we are designed by nature to suffer from the Müller-​Lyer illusion. Its universality and resistance to philosophy demonstrates that. Moreover, Hume argues, it is original—​that is, innate. We do not learn to posit a self through social interaction, although that might be essential to becoming a person.27 27.  The similarity of this view to the Buddhist analysis of the innateness of self-​ grasping—​the tendency to take oneself to be a substantial, independent entity—​and of the relation of that self-​grasping to affective life is striking.

( 70 )  Book II

Hume points out that whether we are talking about the idea of the person or the merely apparent idea of the self, part of the psychological story must advert to original principles in the mind, or what we would call innate psychological mechanisms. He argues that we must presuppose that the mind is endowed with such original qualities in any case: That this proceeds from an original quality or primary impulse, will likewise appear evident, if we consider that ’tis the distinguishing characteristic of these passions. Unless nature had given some original qualities to the mind, it cou’d never have any secondary ones; because in that case it wou’d have no foundation for action, nor cou’d ever begin to exert itself. Now these qualities, which we must consider as original, are such as are most inseparable from the soul, and can be resolv’d into no other: And such is the quality, which determines the object of pride and humility. [2.1.3.3, 280]

The fact that the passions take the self as their object, and so give rise to this robust social persona, derives from an original, or innate, quality of the mind. This must be so despite the fact that the sense of self that results is necessarily artificial. Without an originally existing object of which we could have introspective knowledge, the only way that we could produce one so reliably would be to be wired to do so. That is the foundation of our social nature: just as we naturally acquire language when immersed in a language community, we naturally construct a personal identity when we interact affectively with others. We now turn from consideration of the classification of the passions and their object to a discussion of their causes, a discussion that will add considerable detail to the picture we now have before us.

C H A P T E R  4

w

The Passions and Human Nature

1.  THE CAUSES OF THE PASSIONS So much for the object of the passions. It is both natural and original, even if it is, strictly speaking, nonexistent. We cannot help but generate the self-​ illusion when we experience passions, and this tendency is innate. This is a matter of individual psychology. But what of the causes of our passions? Are the causal relations that give rise to our passions also natural and original? In this chapter, we consider Hume’s account of the structure of the passions, and of their role in human psychology. We will begin with his account of the distinction between the causes and the objects of passions before considering the relation of the passions to reason and the imagination, as we build an account of their centrality to Hume’s explanation of the springs of human behavior. We will conclude with attention to the way that Hume takes this account to dissolve the problem of understanding the freedom of the will. Throughout, we will see the central role that custom—​in both the individual and the collective sense—​plays in Hume’s naturalistic program. We will also see the importance of Hume’s insistence that our social nature is not accidental to us, but rather is essential: we are not, on his view, originally individuals who happen to choose to live together in societies, but rather, like bees, naturally social, and incapable of achieving personhood outside of a social context. Hume begins his discussion by asking whether our passions and their causes are natural to us: We may, perhaps, make it a greater question, whether the causes, that produce the passion, be as natural as the object, to which it is directed, and

( 71 )

( 72 )  Book II whether all that vast variety proceeds from caprice or from the constitution of the mind. This doubt we shall soon remove, if we cast our eye upon human nature, and consider that in all nations and ages, the same objects still give rise to pride and humility; and that upon the view even of a stranger, we can know pretty nearly, what will either encrease or diminish his passions of this kind. [2.1.3.4, 280–​81]

Hume here infers from the fact that the causes of passions are universal to the claim that they are natural. It could not be an accident, he thinks, that insults universally cause anger, and that benevolent actions give rise to affection. Hume, though a relatively cosmopolitan Scot, was no anthropologist, and never traveled outside of Europe. It is unclear whether his view would have survived a compete survey of the world’s civilizations. On the other hand, it is likely that a core claim would survive. The work of psychologists such as Paul Ekman (Ekman and Davidson 1994)  suggests that there are levels of descriptions at which certain basic emotions have universal kinds of causes and expression. In any case, he certainly did not think that the particular causes of the passions are original, and in this he was surely correct: But tho’ the causes of pride and humility be plainly natural, we shall find upon examination, that they are not original, and that ’tis utterly impossible they shou’d each of them be adapted to these passions by a particular provision, and primary constitution of nature. Beside their prodigious number, many of them are the effects of art, and arise partly from the industry, partly from the caprice, and partly from the good fortune of men. Industry produces houses, furniture, cloaths. Caprice determines their particular kinds and qualities. And good fortune frequently contributes to all this, by discovering the effects that result from the different mixtures and combinations of bodies. ’Tis absurd, therefore, to imagine, that each of these was foreseen and provided for by nature, and that every new production of art, which causes pride or humility; instead of adapting itself to the passion by partaking of some general quality, that naturally operates on the mind; is itself the object of an original principle, which till then lay conceal’d in the soul, and is only by accident at last brought to light. [2.1.3.5, 281]

This point is important, as it reminds us of the importance to Hume of the distinction between the natural and the original. The latter is but a small subset of the former, and the fact that we come by some of our proclivities socially does not make them any less natural, for it is natural for us to develop some psychological traits—​for instance, our capacity to

The Pa s s i o n s a n d H u m a n  N at u r e 

( 73 )

speak a language—​through social interaction. Our language faculty may well be original in us, but no natural language is original in this sense. Nonetheless, that original faculty, when left to operate on the natural environment in which we develop—​human society—​transforms us into speakers of the language spoken in that community. That transformation, and our speaking of our native language, is therefore as natural as anything about us. Hume now draws another surprising conclusion regarding how we must conceive ourselves: the persons we are extend well beyond our bodies and what we might naturally think of as our minds. In contemporary terms, we as subjects are broadly supervenient beings; our identity as persons is determined by more than our own minds and bodies: Again, in considering the subjects, to which these qualities adhere, I make a new supposition, which also appears probable from many obvious instances, viz. that these subjects are either parts of ourselves, or something nearly related to us. [2.1.5.2, 285] First, I find, that the peculiar object of pride and humility is determin’d by an original and natural instinct, and that ’tis absolutely impossible, from the primary constitution of the mind, that these passions shou’d ever look beyond self, or that individual person, of whose actions and sentiments each of us is intimately conscious. [2.1.5.3, 286]

While Hume, unlike Kant, has no theory of synthesis, we see him here supposing that the person we take ourselves to be is the product of a kind of natural synthesis, not a mere discovery of something already existent. The passions are not only natural, but original Nonetheless, in order to count as passions—​in order to move us—​they must be directed only to things that are intimately involved with our identity. While there might be considerable variation among persons and cultures regarding what we appropriate as constitutive of our identity, we each instinctively identify ourselves deeply with things outside of our biological being, whether those be family members, possessions, sports teams, or religious communities.1 The construction of the person then, Hume argues, while perhaps variable in detail, is both natural and social. We are bundles of perceptions, but we are bundles intimately connected to the world around us. Though Hume does not put it this way, this multiplicity of connections and the indefinite nature of bundles, must inevitably blur our boundaries, with 1. See Ainslie (1999) for a similar account.

( 74 )  Book II

the consequence that our identity is underdetermined by any facts independent of custom.2 Hume then draws all of this together in a single, dense explanation of the structure of a passion: If I compare, therefore, these two establish’d principles of the passions, viz. their object, which is self, and their sensation, which is either pleasant or painful, to the two suppos’d properties of the causes, viz. their relation to self, and their tendency to produce a pain or pleasure, independent of the passion; I immediately find, that taking these suppositions to be just, the true system breaks in upon me with an irresistible evidence. That cause, which excites the passion, is related to the object, which nature has attributed to the passion; the sensation, which the cause separately produces, is related to the sensation of the passion: From this double relation of ideas and impressions, the passion is deriv’d. The one idea is easily converted into its co-​relative; and the one impression into that, which resembles and corresponds to it:  With how much greater facility must this transition be made, where these movements mutually assist each other, and the mind receives a double impulse from the relations both of its impressions and ideas? [2.1.5.5, 286–​87]

Let us work through this passage with care. It is not immediately clear what is being explained, or what the explanation is. The object of a passion is always the self; but that self is a merely intentional object, and so not a cause of the passion; in fact, it will be the effect of the passions. The cause is some real object taken to be associated closely with the self (or, more properly, the person), and it excites a passion when it gives rise to some kind of pleasure or pain. The sensation of the passion—​its hedonic tone, whether positive or negative—​covaries with the hedonic tone of the awareness of the cause. Painful passions such as humility or anger are caused by things we experience as painful whether in their own right, or as regarded by others; pleasant ones such as pride or affection from things we experience as pleasant, either in themselves or in the gaze of others. The unity of the passion, Hume contends, arises from this “double relation,” viz., the fact that the hedonic tone of the cause of the passion is in harmony with that of the passion itself. But what is the point of all of this? I claimed in c­ hapter 1 that we should try as much as possible to adhere to the Cover Principle, viz., the principle 2. Once again, I note the affinity to Buddhist conceptions of the person as bundles of bundles (skandhas) whose identity is conventionally constituted—​particularly those defended by philosophers such as Candrakīrti and Śāntideva.

The Pa s s i o n s a n d H u m a n  N at u r e 

( 75 )

that Hume is first and foremost articulating an empirical theory of human nature, of how the mind works. Hume is not concerned primarily with metaphysics, epistemology, or ethics, but rather with our psychology. His leading idea is that an explanation of human thought is an explanation of its causes, and of the mechanisms by which those causes operate. As part of this project, he is also concerned to explain how passions can at the same time be experienced as pleasant or painful impressions, and be intentional—​how they can comprise judgments or assessments while still having a qualitative character—​and how these aspects are unified in a single state, and not simply realized in a pair of accompanying states. In this passage, he is explaining how a passion—​an impression of reflection—​can simultaneously be directed upon a merely intentional object which is not its cause, and which can convey no impression sufficient to give that passion any force or vivacity, and yet itself have sufficient force and vivacity to appear to us as passions do, and to cause behavior. We must, that is, explain causally the fact that a passion is felt intensely and has sufficient motive force to cause behavior, and the fact that it is directed upon the merely intentionally existent self, not upon its cause, as well as the fact that it is a single impression of reflection, and not two associated impressions or ideas. The complexity of the account as involving this double relation arises from the need to explain this unity. My idea of the cause of the passion is closely associated with my idea of myself; my perception of the cause involves an intense hedonic tone, either pleasant or painful; the impression of reflection that arises from the awareness of that perception acquires that tone, and is directed upon the cause of the initial perception. Nonetheless, since the idea of that cause is closely associated with my idea of myself, an impression of reflection immediately arises of myself, and that impression carries the hedonic tone of the immediately prior impression of reflection. This latter impression of reflection is the passion, an impression the intentional object of which is me, but the cause of which is the perception of some other object associated with me. This subsequent impression of reflection is imbued with pleasure or pain derived from the perception of the cause, and that pleasure or pain in turn gives it sufficient force to be experienced as a passion, and to give rise to behavior. Consider for example, my pride in my dog’s Frisbee-​catching ability. The cause is the graceful leaping catch executed at high speed. Witnessing the catch gives me pleasure. My idea of my dog, however, is closely associated with my idea of myself—​he is, after all, my dog, and I trained him to do this. The pleasure caused by watching him catch the disk gives rise to an impression of reflection—​my awareness of that pleasure as pleasure in his

( 76 )  Book II

catch, which is pleasant itself. But that awareness, in virtue of the association of the idea of the dog with my idea of myself, gives rise to an impression of reflection of myself as the master of the accomplished dog, which retains that pleasant hedonic tone while being intentionally directed to myself. This impression constitutes pride; that pride in turn is sufficiently powerful to generate my bragging about my dog’s prowess in a proud way. My humility at my poor dancing is caused by my awareness of the pain that my dancing causes others as well as the pain I experience when I contemplate my address on the dance floor; the awareness of that pain gives rise to a painful sensation that is similarly associated with me, the poor dancer, as its object. The impression of reflection that pain occasions, in harmony with the pain caused by the dancing itself, constitutes the passion of humility. The account is causal all the way down, and brings a double relation to bear to explain a single impression of reflection that satisfies the phenomenological, intentional, and causal descriptions of a passion. The principal causes of our behavior, on Hume’s account, are comprised under the general rubric of custom, and as we have seen, custom comprises both our original psychological constitution and the natural impact of social intercourse. The account of the passions is an account of how all of this works. I need both the original tendency to associate my dog with myself and to take pleasure in certain kinds of things; I also need the social structure that makes dog ownership possible and a shared awareness of certain canine accomplishments as suitable objects of pleasure. Otherwise, pride in the accomplishments of dogs would make no sense. Hume immediately emphasizes this point, as well as its close relation to the Book I account of the relation between impressions and ideas as follows: This being fully comprehended, it may now be ask’d, Whether nature produces the passion immediately, of herself; or whether she must be assisted by the co-​operation of other causes? [2.1.5.7, 287] In a word, nature has bestow’d a kind of attraction on certain impressions and ideas, by which one of them, upon its appearance, naturally introduces its correlative. If these two attractions or associations of impressions and ideas concur on the same object, they mutually assist each other, and the transition of the affections and of the imagination is made with the greatest ease and facility. [2.1.5.10, 289]

That is, there is a natural and, indeed, to use Hume’s language, an original tendency in the mind to associate certain kinds of impressions and ideas, for certain kinds of passions to arise in certain kinds of circumstances. That much is original. But how that original tendency plays out in particular

The Pa s s i o n s a n d H u m a n  N at u r e 

( 77 )

cases depends very much upon accident and culture. This entire picture, Hume argues, is of a piece with the general account of the relation between impressions and ideas, according to which impressions give rise to resembling ideas, and to impressions of reflection, and that chains of association follow relations of similarity and causality. And in the end, we constitute ourselves as affective centers of gravity in social contexts. This uniformity of method, he argues, is therefore mutually reinforcing. The success of explanation in each case adds explanatory force to the other: To illustrate this hypothesis, we may compare it to that, by which I have already explain’d the belief attending the judgments, which we form from causation. I have observ’d, that in all judgments of this kind, there is always a present impression, and a related idea; and that the present impression gives a vivacity to the fancy, and the relation conveys this vivacity, by an easy transition, to the related idea. Without the present impression, the attention is not fix’d, nor the spirits excited. Without the relation, this attention rests on its first object, and has no farther consequence. There is evidently a great analogy betwixt that hypothesis, and our present one of an impression and idea, that transfuse themselves into another impression and idea by means of their double relation: Which analogy must be allow’d to be no despicable proof of both hypotheses. [2.1.5.11, 289–​90]

The relevant analogy goes like this: when we form a belief—​an idea with enough force and vivacity to give rise to action—​the belief gets its vivacity from a contiguous perception. For an impression of reflection to constitute a passion, it requires sufficient force and vivacity to generate action; any impression of reflection insufficient in force and vivacity to cause action might be an awareness of our current cognitive state, but would not amount to a passion. That force and vivacity, Hume claims, is derived from the present impression that gives rise to the passion. So, he argues, since we see the same mechanism accounting for the power of beliefs and passions to cause action, principles of theoretical parsimony suggest that we are on the right track in each case.3 Hume then offers what he calls five “limitations of this system,” or, as we might put it in more modern terms, boundary conditions within which the theory is meant to operate. The first two concern proximity. The passions are aroused only by things that are taken to be closely related to ourselves. That is, in the Newtonian terms in which Hume thinks, there is a kind of

3. Ainslie (1999) makes a similar point.

( 78 )  Book II

inverse square law governing the ability of a stimulus to excite a passion. Now this proximity, he notes, need not be physical proximity; it could well be social, or even merely perceived proximity [2.1.6.2, 290–​91]. Second, we only find pride and humility excited by those things that we take to be “peculiar to ourselves, or at least common to us with a few persons” [2.1.6.4, 291]. The point is that these passions in particular (which Hume sees as among our most basic in the construction of our person, or our sense of self) are self-​regarding. No matter how near something is to us, if we do not see it as associated with us in particular, while it might excite some response, it could not elicit these passions, and so will not be important in our self-​construction.4 Third, there is an essential social dimension to the stimuli of pride and humility: we must see the sources of pleasure or pain to us as also sources of pleasure and pain to those around us. If I am proud of my cooking, it must be because I see good cooking as something others would value and admire; I can’t be proud of the dirt in my house, not because of anything intrinsic to the dirt, or because it has no connection to me, but because I cannot see that dirt as something that others would admire. If I am humbled by my address at dancing, it must be because I think that others would be dismayed to see me dance; I cannot be humbled by my dog’s skill at catching Frisbees because I cannot imagine others disapproving of it.5 Fourth, these passions so essential to self-​construction are excited only by things whose connections we see as relatively stable, not accidental and momentary, or at least are proportional to the degree we see them as reflective of stable states. A one-​off is less the subject of pride and humility than a long-​standing trait or achievement. And finally, what Hume calls general rules [2.1.6.6,  293], or our prejudices about kinds of persons, social classes, etc., determine the kinds of passions we experience. He puts it thus: “Custom readily carries us beyond the just bounds in our passions, as well as in our reasonings” [2.1.6.6, 293]. Once again, while the principles that determine our passions, and so the ways in which we understand ourselves, may be natural, they are not entirely original.

4.  There is a connection here to the legal framework from which Hume draws the notion of custom. Customary law was always understood as local law, as opposed to common law that bound all in a realm. In order to be bound by customary law, one needed to belong to the community to which that law pertained. 5.  If, however, I  were raised in one of those cultures that regarded dogs and any human association with dogs as disgusting, I might well be humbled at the thought of having spent so much time training a dog to do this; the norms that govern the passions are local, like customary laws.

The Pa s s i o n s a n d H u m a n  N at u r e 

( 79 )

All of these “limitations,” or boundary conditions, on the account of pride and humility, and so of our self-​construction, remind us just how much Hume’s account of custom—​that concealed influence on our thought and behavior that the Treatise is meant to reveal—​is an account of our second nature, or as Bacon put it, “the magistrate of man’s life” (1909, 104). While the fact of its operation and the innate principles of psychology are one side of custom, they are at most the foundation for the vast and complex effects of our social environment on the shape of our passions, our conception of self, and on our psychology more generally. To understand human nature requires us to understand human psychology; but to understand human psychology requires us to understand human society. Hume ends this section with a remarkable observation: I shall close this subject with a reflection deriv’d from these five limitations. This reflection is, that the persons, who are the proudest, and who in the eye of the world have most reason for their pride, are not always the happiest; nor the most humble always the most miserable, as may at first sight be imagin’d from this system. An evil may be real, tho’ its cause has no relation to us: It may be real, without being peculiar: It may be real, without shewing itself to others: It may be real, without being constant: And it may be real, without falling under the general rules. Such evils as these will not fail to render us miserable, tho’ they have little tendency to diminish pride:  And perhaps the most real and the most solid evils of life will be found of this nature. [2.1.6.10, 294]

Hume has called our attention to two important facts about our passions: first, the general fact of the hiddenness of the springs of our psychology; and second, the more specific fact that the passions are second nature, and so are subject to the vicissitudes of social context and of our own perceptions of relations we might bear to others and to the world around us, as opposed to how those relations stand themselves. A consequence of this is that pride and humility, let alone other passions such as anger or love, cannot be taken to track reality, but only our perception thereof. They might appear to be guides to our own happiness or misery, but they are inconstant guides. What makes us happy or miserable may often be hidden from us.6

6. Once again, the affinity to so much Buddhist moral psychology is intriguing. We also see in that context repeated arguments that what we see as sources of happiness are often sources of suffering, and vice versa. See Garfield (2015), esp. c­ hapters 1 and 2.

( 80 )  Book II

2.  PASSION AND REASON We now skip the meat in the Book II sandwich—​the moral psychology, to which we will return in c­ hapter 5—​to continue our examination of the account of the passions themselves in 2.3.3. Here Hume takes up the difficult question of the relation between the passions and reason. Nothing is more usual in philosophy, and even in common life, than to talk of the combat of passion and reason, to give the preference to reason, and to assert that men are only so far virtuous as they conform themselves to its dictates. Every rational creature, ’tis said, is oblig’d to regulate his actions by reason; and if any other motive or principle challenge the direction of his conduct, he ought to oppose it, ’till it be entirely subdu’d, or at least brought to a conformity with that superior principle. . . . In order to shew the fallacy of all this philosophy, I shall endeavor to prove first, that reason alone can never be a motive to any action of the will; and secondly, that it can never oppose passion in the direction of the will. [2.3.3.1, 413]

This strategy is absolutely fundamental to the project of the Treatise. Here Hume challenges a very standard (then and now) picture (in the European tradition) of human behavior as determined in part by the passions and in part by reason. On this picture, we are at our best, our most human, and our most morally responsible when we allow reason to guide our behavior and subdue the irrational, egoistic passions. We see this image of human behavior in Western philosophy shared by Plato and Aristotle, as well as by Locke, and enshrined in the most contemporary legal theory with its Kantian legacy. Hume will argue both that reason never is by itself a motive for action, and second, that it is our passions that not only determine our action, affect, and cognition, but that as human animals only our passions could ever do so. To understand human psychology, whether in the context of reasoning or of moral assessment, Hume will argue, is to understand the passions. Hume’s argument for the impotence of reason as a determinant of action is brief and straightforward: The understanding exerts itself after two different ways, as it judges from demonstration or probability; as it regards the abstract relations of our ideas, or those relations of objects, of which experience only gives us information. I believe it scarce will be asserted, that the first species of reasoning alone is ever the cause of any action. [2.3.3.2, 413]

When we reason, on this view, we determine what is true or false. We move from premises to conclusions; we judge the relations between concepts. All

The Pa s s i o n s a n d H u m a n  N at u r e 

( 81 )

of that gives us information. But in determining how to act, information alone is impotent. Any motive to act requires not just cognition but conation. We must want to do something, value something, desire something, fear something, care about someone, etc. Only when we join cognitive information with conation do we have a reason to act. And the motive is given by the desire, aversion, or affect; and these are not things that we do, or for which we have reasons. Instead, Hume argues, they are features of our personality that happen to us; in them we are passive; they are our passions. So, Hume concludes: We speak not strictly and philosophically when we talk of the combat of passion and of reason. Reason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them. [2.3.3.4, 415]

The task of reason is not to tell us what to want, or about what to care, but, once we have a motive, to assist us in achieving the desired end. Even the proof of a theorem requires us first to care about the answer. Only then does reason enter the theatre to assist us in getting from axioms to demonstration.7

3.  HOW THE PASSIONS OPERATE But what are the passions, and how do they act to determine our affective and cognitive lives? Hume’s spade will be turned a bit sooner than we might like, which is not surprising given that he is writing in the eighteenth century, prior to any serious empirical research on human affect. But he has some important things to say, much of which derives from his reading of Hutcheson and Shaftesbury. From Hutcheson he inherits the distinction we considered previously between the calm and the violent passions. Once again, as we saw, Hume draws this distinction differently from his predecessors:  the distinction between the calm and the violent passions in his hands is not a distinction grounded in their force or efficacy, but rather in their origins and onset; a matter of kind, not of degree or intensity. Calm passions are those that endure over time, and violent passions 7.  This view of the relation between reason and the passions is often regarded as original with Hume. But, as Laird (1932) observes, Hume follows Hutcheson and Mandeville very closely in this regard. Even the phrase “slave of the passions” was first used to describe reason by Mandeville in his essay Origin of Honor (1732). See Korsgaard (1986) for a critical discussion of Hume’s account of the limited motivational role of reason.

( 82 )  Book II

those aroused by particular stimuli that endure only for a short time. Hume emphasizes, contra Hutcheson, that often it is the calm passions that have the much greater effect on our motivational set and on action. Here custom is essential to the story. For custom, in this case in its guise as habit, is what accounts for the greater force of the calm passions. Since they induce standing dispositions, the regularity of their operation on our motive structure strengthens their influence. One’s admiration for honesty, for instance, is a calm passion; its steady operation in our lives makes us more honest and leads us to approve of the honesty of others. A violent passion, say anger at an instance of dishonesty may, on the other hand, lead us to act in a single case, but is a less important driver of our long-​term behavior. To be calm is therefore not to be weak; to be violent is not to be strong: There is not in philosophy a subject of more nice speculation than this of the different causes and effects of the calm and violent passions. ’Tis evident passions influence not the will in proportion to their violence, or the disorder they occasion in the temper; but on the contrary, that when a passion has once become a settled principle of action, and is the predominant inclination of the soul, it commonly produces no longer any sensible agitation. As repeated custom and its own force have made every thing yield to it, it directs the actions and conduct without that opposition and emotion, which so naturally attend every momentary gust of passion. We must, therefore, distinguish betwixt a calm and a weak passion; between a violent and a strong one. [2.3.4.1, 418–​19]

Hume emphasizes the importance of custom in this context and the importance of custom as an original, that is, an innate, principle in our psychology. Custom in this guise is the fundamental operating principle of cognition and motivation, and this emphasis on the role of habit in shaping not only our actions and personality, but also our cognitive lives, is one of many respects in which Hume anticipates modern behaviorism and its descendent connectionism. Hume points out that custom in this sense makes actions that are practiced—​again, whether actions of body or of mind—​both more likely to be performed in the future and more easily performed. The image—​ derived from Malebranche—​is of channels worn in the brain by the regular flow of the animal spirits that enable the sprits to flow more spontaneously in those directions in the future, as water flows through the canyons its flow creates.8 But custom has an analogous effect on the passions, with 8.  This metaphor is not so quaint as it may seem. Contemporary cognitive neuroscience uses the term canalization to refer to the establishment of efficient neural networks that subserve processes made automatic by frequent repetition.

The Pa s s i o n s a n d H u m a n  N at u r e 

( 83 )

repeated occurrence making future occurrence more likely. And here Hume is in agreement with Aristotle, as well as with Buddhist psychologists, who observed long before that affect is as subject to the effects of habit as is behavior: But nothing has a greater effect both to encrease and diminish our passions, to convert pleasure into pain, and pain into pleasure, than custom and repetition. Custom has two original effects on the mind, in bestowing a facility in the performance of any action or the conception of any object; and afterwards a tendency or inclination towards it; and from these we may account for all its other effects, however extraordinary. [2.3.5.1, 422]

An essential part of Hume’s program in the Treatise, and a strategy we will see adopted in virtually every analysis Hume offers, is the elevation of the imagination to pride of place as the faculty of mind responsible for cognition and experience.9 Hume assigns the imagination a central role in our experience of the external world, in our constitution of personal identity, and in our moral lives. (See c­ hapters 7–​10 for much more on this point.) But this recognition is limited: it is never clear what Hume takes the bounds of the imagination to be, or how precisely he takes it to operate; unlike Kant’s synthetic faculty, for instance, it remains a relatively flat domain of association that occasionally springs into action to create new ideas. For this reason, although the insight that imagination is so important in these aspects of our cognitive and affective lives is important, the psychology of its operation is unfortunately inadequate. Here Hume Cohon (2006, 2008) speculates on Hume’s use of the term will in contexts such as that just quoted, and suggests that he may understand the will as a passion, or as a sui generis impression. I think that this overreads Hume in this regard. Hume never develops a theory of the will, never treats it as a faculty, and never explains anything by it; indeed, his discussion of the fallacies involved in arguments for the so-​called freedom of the will suggest that he has no place for such a construct in his psychology. I take his use of this term to be casual and nonreferential. As Connolly (1987, 294–​95) notes, however, when Hume describes the will as “nothing but the internal impression we feel and are conscious of, when we knowingly give rise to any new motion of our body, or new perception of our mind” [2.3.1.2, 399], he does seem to take it to be an impression of reflection that could figure in the explanation of behavior. But immediately after that remark, Hume continues, “This impression ’tis impossible to define, and needless to describe any farther,” suggesting that he does not take this impression of agency to be all that important, a reading that we will see gains support from Hume’s discussion of freedom and agency. Connolly is thus correct to say (296–​97) that Hume does not take the will to be a cause of behavior, any more than it is a faculty. 9. Laird makes this point as well, and remarks on the degree to which Hume follows Malebranche in this regard (Laird 1932).

( 84 )  Book II

emphasizes the close relationship between the functions of the imagination and the passions. ’Tis remarkable, that the imagination and affections have a close union together, and that nothing, which affects the former, can be entirely indifferent to the latter. Wherever our ideas of good or evil acquire a new vivacity, the passions become more violent; and keep pace with the imagination in all its variations. [2.3.6.1, 424]

That is, Hume observes, the imagination is instrumental in exciting the passions, and the passions in turn call the imagination into operation. We are certainly aware of this tight union in our everyday affective lives. Our passions can be aroused by imagining a scene or an event, and this is the point of much of our moral and political rhetoric. Moreover, when our passions are aroused, we very often find ourselves imagining courses of action or events we would either like to see realized or that we fear. Hume’s location of the passions and the imagination at the center of our psychology reflects this close association between the two, and will be evident in his account of ethics, as we will see in ­chapter 11. Hume closes this discussion with observations regarding the original principles that govern the passions and their relation to the rest of our psychology, both the innate tendency to pursue the pleasant and avoid the unpleasant, and the tendency of passions to engage action: The impressions, which arise from good and evil most naturally, and with the least preparation are the direct passions of desire and aversion, grief and joy, hope and fear, along with volition. The mind by an original instinct tends to unite itself with the good, and to avoid the evil, tho’ they be conceiv’d merely in idea, and be consider’d as to exist in any future period of time. [2.3.9.2, 438] These indirect passions, being always agreeable or uneasy, give in their turn additional force to the direct passions, and encrease our desire and aversion to the object. Thus a suit of fine cloaths produces pleasure from their beauty; and this pleasure produces the direct passions, or the impressions of volition and desire. Again, when these cloaths are consider’d as belonging to ourself, the double relation conveys to us the sentiment of pride, which is an indirect passion; and the pleasure, which attends that passion, returns back to the direct affections, and gives new force to our desire or volition, joy or hope. [2.3.9.4, 439]

Once again, Hume emphasizes the social dimensions of our passions, and so their constitution by, and institution ofour sense of our role in our social networks, in a spiral of affect and self-​conception. While the foundations

The Pa s s i o n s a n d H u m a n  N at u r e 

( 85 )

of our passions are original in us, their articulation is, while also natural, ineliminably social. They are hence customary in both senses of that term.

4.  FREEDOM AND CAUSATION It is also important to note that Hume’s account of the influence of the passions on behavior is resolutely causal. Our passions cause us to act; they do not give us reasons that form the basis of deliberation leading to free action. Hume does not argue directly that we are not free—​he is not, that is, adopting a metaphysical position in a freedom-​determinism debate. Instead, consistent with the Pseudo-​idea Principle, he argues that the kind of freedom presupposed by friends of agent-​causation does not even make sense; we do not even have an idea of such freedom. While this insight connects to the impotence of reason as a determinant of action, it goes further. It nails down Hume’s naturalism both with respect to the explanation of human behavior and with respect to moral assessment. Hume begins his discussion by noting the universality of causal determination, a universality entailing that it applies to our own bodies: ’Tis universally acknowledg’d, that the operations of external bodies are necessary, and that in the communication of their motion, in their attraction, and mutual cohesion, there are not the least traces of indifference or liberty. Every object is determin’d by an absolute fate to a certain degree and direction of its motion, and can no more depart from that precise line, in which it moves, than it can convert itself into an angel, or spirit. . . . The actions, therefore, of matter are to be regarded as instances of necessary actions; and whatever is in this respect on the same footing with matter, must be acknowledg’d to be necessary. [2.3.1.3, 399–​400]

Hume then extends this principle to the mind, asking rhetorically: Are the changes of our body from infancy to old age more regular and certain than those of our mind and conduct? And wou’d a man be more ridiculous, who wou’d expect that an infant of four years old will raise a weight of three hundred pound, than one, who from a person of the same age, wou’d look for a philosophical reasoning, or a prudent and well-​concerted action? [2.3.1.7, 401]

Inasmuch as our conviction in the causal determinants of our mental acts, like those of our physical acts, is grounded in the regularities that govern them, the fact that we have as much conviction in psychological

( 86 )  Book II

regularity as we do in physical regularity forces us to adopt a causal attitude toward our own mental actions. Hume emphasizes this point when he notes that “in judging of the actions of men we must proceed upon the same maxims, as when we reason concerning external objects. When any phaenomena are constantly and invariably conjoin’d together, they acquire such a connexion in the imagination, that it passes from one to the other, without any doubt or hesitation” [2.3.1.12, 403]. Here Hume calls explicit attention to the account of causal reasoning he develops in 1.3.14, to which we return in c­ hapter 6. Hume concludes this discussion with a memorable, if macabre, confirmation of our certainty in the causal character of thought, and in the lawlike nature of the generalizations governing human behavior: A prisoner, who has neither money nor interest, discovers the impossibility of his escape, as well from the obstinacy of the gaoler, as from the walls and bars with which he is surrounded; and in all attempts for his freedom chuses rather to work upon the stone and iron of the one, than upon the inflexible nature of the other. The same prisoner, when conducted to the scaffold, foresees his death as certainly from the constancy and fidelity of his guards as from the operation of the ax or wheel. His mind runs along a certain train of ideas: The refusal of the soldiers to consent to his escape, the action of the executioner; the separation of the head and body; bleeding, convulsive motions, and death. Here is a connected chain of natural causes and voluntary actions; but the mind feels no difference betwixt them in passing from one link to another. [2.3.1.17, 406]

Nonetheless, Hume concedes, many, despite their reflective commitment to universal determinism both in the physical and in the mental domain, are convinced that human actions spring from a special freedom, or spontaneity, that characterizes human action in particular. We seem to be convinced that in acting we are free, and that in any action we could have chosen otherwise. He writes that “’tis difficult for us to perswade ourselves we were govern’d by necessity, and that ’twas utterly impossible for us to have acted otherwise; the idea of necessity seeming to imply something of force, and violence, and constraint, of which we are not sensible” [2.3.2.1, 407]. That is, although we recognize ourselves in the prisoner when we reflect from a third-​person point of view, it is hard to experience the same sense of constraint from the first-​person point of view. Hume offers a threefold diagnosis of the tendency to reject determinism in the first-​person case. First, we tend to confuse the absence of violence or constraint (what Hume calls “the liberty of spontaneity”) with the absence of any cause, or “the liberty of indifference” [2.3.2.1, 407]. In thinking of

The Pa s s i o n s a n d H u m a n  N at u r e 

( 87 )

ourselves as free, we are concerned that our own motives cause our actions, and not those of another, aspiring to freedom of the first kind—​the liberty of spontaneity—​but often take that to amount to acting causelessly. Second, Hume argues that we sometimes have a “false sensation or experience even of the liberty of indifference; which is regarded as an argument for its real existence” [2.3.2.2, 408]. That is, we mistake the experience of acting from a motive for an experience of causelessly initiating an action, a thought Hume probably derives from his reading of Locke, as Russell (1995) notes.10 Here we see Hume deploying the nominalist Pseudo-​idea Principle. We can never have an impression of indifferent freedom, since any action we regard as our own is an action we also take to be caused by our own motives and intentions. Since we can have no such impression, but only a “false impression” caused by the confusion of freedom with authorship, we can never even have such an idea. We can form the idea of freedom as spontaneity—​that is, as action caused by our own desires and motives as opposed to by external constraint, but that is not uncaused action, and we must always acknowledge that our motives and desire have causes, and so on ad infinitum. But the kind of freedom that would be inconsistent with causation, Hume argues, does not even make any sense. Penelhum (2000a) correctly connects this to Hume’s naturalism. He writes: This brings us to the bedrock of Hume’s understanding of what a science of human nature has to be like.  .  .  . [T]‌he common distinction between virtues and talents, which he finds a source of difficulty, exists because the popular ascription of virtue to someone involves ascribing some degree of what Hume calls liberty of indifference to that person. But Hume would respond that this entails the denial of the very predictability of human conduct that our ethical thinking requires, and is inconsistent with the scientific study of mankind. (2000a, 149)11

So, whence does our commitment to a freedom of indifference arise? Hume argues that religious doctrine motivates a strong thesis of indifferent

10. Russell (1995) also points out that Hume returns to this theme in the appendix, arguing that there is no special impression of necessity linking internal states that gives us greater insight into the causal link than do the connections between external events. 11. Nonetheless, Penelhum himself endorses the libertarian objection, arguing that Hume must be wrong on the grounds that we manifestly have free will. I  disagree; I think that Hume’s diagnosis shows why free will in the Augustinian sense is simply incoherent.

( 88 )  Book II

freedom. Hume is right about this. As I have argued elsewhere (Garfield 2014), the Western doctrine of the freedom of the will is an artifact of Augustine’s solution to the problem of theodicy arising from the Fall in the Garden of Eden. Since that time, many Christian theologians and Kantian ethicists have argued that freedom in this strong sense is necessary for moral responsibility, and even to make sense of the notion of virtue or vice, on pain of holding us responsible for actions or character traits over which we have no real control. Hume is at his best in refuting this idea. He points out that the doctrine of absolute freedom or agent causation that the Christian moralist desires would in fact undermine, rather than save, the only kind of freedom worth having—​that is, the freedom to act on our own motives. If our intentions, values, character, and motives did not cause our behavior, we would not be responsible for it; it would not be ours in any sense, and we would be unfree, in that we would have no control over our own bodies. Similarly, if our motives, intentions, and values were not in turn caused by other psychological states we regard as our own, we would be unfree in adopting them. Agency and responsibility are hence not inconsistent with causal determinism, but instead demand it.12 Hume’s naturalism hence leads him to an account of human action that makes sense of agency in the context of universal causation; at the same time, it reveals that the seductive misconception of agency as demanding transcendental freedom is in fact not even cogent. His account has ramifications for how we think about moral responsibility and moral evaluation as well. When we attribute responsibility, we are not, on this view, committed to the view that the person we deem responsible could have acted otherwise; only to the fact that that person is the author of her actions, in the sense that her actions are caused by intentions and motives with which she identifies or with which others identify her. This is hence a kind of hermeneutic view of responsibility (Garfield 2014). When we regard an action or a character trait as virtuous or vicious, that assessment—​as we will see in ­chapter 9—​is independent of any attribution of freedom from causal determination, but issues from our moral sentiments, as they are tempered by custom, leading to an assessment of 12. See Russell (1995) for a similar view of Hume’s position regarding the relationship between freedom and moral assessment and Garfield (2014) for an extended discussion of agency without freedom. One might suspect that Hume’s reliance on causal determinism in this context is inconsistent with his position on the idea of necessary connection. As we will see below, it is not. Hume’s naturalism demands a causal explanation of human behavior; his skepticism about causation is carefully framed to make sense of causal explanation.

The Pa s s i o n s a n d H u m a n  N at u r e 

( 89 )

the fitness of such actions and characters to conduce to a happy social life.13 That account, too, is thoroughly naturalistic in character.

5.  SUMMING UP Hume’s psychology, following Mandeville and Newton, is aimed in large part at locating human beings in the natural world as complex social animals, as opposed to embodied rational beings. Part and parcel of Hume’s naturalism—​and fundamental to his skepticism as well—​is a view of human psychology as more passive than active; driven by forces of which we are often unaware, or only dimly aware; and governed by psychological laws. Explaining the passions—​those forces that drive us in response to stimuli, and that are so constitutive in our representation of ourselves as subjects and as agents—​and the role of the imagination, a faculty intimately connected with those passions and again driven by associative laws rather than reason, is central to this project. Hume’s psychology is a precursor to a science of the nomic explanation of human behavior, not an account of rational agents deliberating regarding the true and the good. As we will see, it provides the foundation for a compelling vision of the structure of human being.

13. Bricke sums up the implications of Hume’s treatment of freedom for thinking about moral responsibility as follows: Hume is a naturalist about the mind. He purports to offer naturalistic characterizations and explanations of, among other things, the affective life of human beings, and in particular the so-​called indirect passions. . . . Hume’s concern, seen this way, is to identify the many sorts of considerations about agents and their circumstances that, as a matter of fact, do affect, in exculpatory ways, our affective responses to one another. (2011, 215) See also Connolly (1987, 287–​89) for a perceptive discussion of Hume on the importance of social explanations of behavior. Russell, on the other hand, takes Hume to task for reducing judgments of moral responsibility to sensations, claiming that on Hume’s account, if we did not experience these peculiar, “sui generis” moral sensations, “no one could be regarded as responsible” (1995, 14). As will become clear in ­chapter 11, I think that this argument confuses Hume’s project of providing a psychology of moral judgment with one of providing an a priori theory of moral responsibility. Hume explains our propensity to moral judgment and shows that our moral institutions are grounded in the customs they enable.

C H A P T E R  5

w

Hume’s Moral Psychology

T

he meat in the sandwich of Book II is Hume’s moral psychology (which, as we will see in c­ hapters 10 and 12, is further adumbrated in his account of morals proper in Book III—​but for now we focus on the account of the morally important passions in Book II). In this context, we see a deepening of Hume’s attention to the social dimension of human nature and to how custom—​understood here as social convention—​shapes our psychology. His concern in moral psychology, as in psychology more generally, is to understand the interplay between original or innate cognitive predispositions and those that emerge through natural human intercourse. This investigation will provide the foundation for his account of ethics. The focus shifts, however, from pride and humility to virtue and vice. The members of the former pair constitute, in Hume’s framework, the foundation of all of the positive and negative affective self-​regarding responses and constitute the basis of a classification of the passions with respect to their affective valence. Those of the latter form the foundation and constitute the principle of classification of the passions we allow to regulate our social and political life—​the moral passions.

1.  MORAL SENSIBILITY AND THE PASSIONS Hume opens his discussion of moral psychology by posing a question that was hotly contested among early modern ethicists and theologians (see Russell 2008, esp. chap.  4):  is our moral sensibility part of our natural endowment or does it arise from social convention? Hume, however, postpones this question to Book III, where he effectively deconstructs it in a classically Pyrrhonian fashion, by denying that the question even makes ( 90 )

H u m e ’ s M or a l P s yc h ol o g y 

( 91 )

sense. But for now, he leaves it open, turning the disjunction into a premise in a constructive syllogism: To begin with vice and virtue, which are the most obvious causes of these passions; ’twou’d be entirely foreign to my present purpose to enter upon the controversy, which of late years has so much excited the curiosity of the publick, whether these moral distinctions be founded on natural and original principles, or arise from interest and education. The examination of this I reserve for the following book; and in the mean time shall endeavor to show, that my system maintains its ground upon either of these hypotheses; which will be a strong proof of its solidity. [2.1.7.2, 295]

Hume then argues first that, if the foundation of morality is conventional, moral sensibility, however modulated by convention, must nonetheless have its origins in the passions: For granting that morality had no foundation in nature, it must still be allow’d, that vice and virtue, either from self-​interest or the prejudices of education, produce in us a real pain and pleasure. . . . Every passion, habit, or turn of character (say they) which has a tendency to our advantage or prejudice, gives a delight or uneasiness; and ’tis from thence that the approbation or disapprobation arises. [2.1.7.3, 295]

How is this argument meant to go? Hume takes it as a datum that virtue, whether contemplated in ourselves or in others, gives rise to approbation, and vice to disapproval.1 To approve of something is to see it as a source of pleasure; to disapprove is to find it painful. But pleasure and pain are not things that we do; they are things we experience. Hence they are the consequences of passions.2 So, even if our responses to virtue and vice are 1. The affinities of Hume’s views to those of Shaftesbury’s account of virtue and vice as kinds of second-​order passions and to Hutcheson’s view that we are endowed with an innate moral sense that tracks external properties are obvious here, but so is his divergence. He agrees that there may be an innate moral sense, but he does not take it to follow that there are independent phenomena that these passions track. Again, a comparison to the language faculty and the interplay in language acquisition between an original capacity and an environment in which it is realized is apposite. 2.  While Norton (1982) sees this as indicating a central role for reason in moral judgment, Kemp Smith (1941) actually has a more nuanced view, according to which reason may be necessary in determining the relations in which we stand to others, and even in determining the beneficial qualities of virtues like justice; but it is only the passions that can lead to specific moral judgments on particular occasions, and only the passions that can move us to act. He writes: Though our approbation of justice presupposes the discovery of its utility, we must not regard it as based on the discovery; if such were the case, the approval would

( 92 )  Book II

acquired, and not innate, a tendency for particular passions to be aroused in certain circumstances is acquired. It is acquired through a kind of classical conditioning, as opposed to a set of reflective judgments arrived at through deliberation on ethical matters or taught to us through moral education. Hume then turns to the second case, the possibility that the foundation of morality lies in human nature: The same unquestionable argument may be deriv’d from the opinion of those, who maintain that morality is something real, essential, and founded on 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, from a primary constitution of nature certain characters and passions, by 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. The pain and pleasure, therefore, being the primary causes of vice and virtue, must also be the causes of all their effects, and consequently of pride and humility. [2.1.7.5, 296]

The point is straightforward, and is the same in all relevant respects as that made in the first argument: whatever the basis is for the approbation or disapprobation of character that give rise to the moral sentiments, that approbation or disapprobation is the result not of reasoning, but of a passion, in this case, a passion that arises naturally. In either case then—​ whether the moral sense is conventional or natural, it is mediated by the passions. Ethical response, for Hume, is one more thing that happens to us, not something that we do.3 But here, as Brown (2011, 20) notes, Hume be due to reason, and that would be contrary to Hume’s fundamental thesis. To account for the approbation, therefore, we must raise the further problem: Why does utility please? Reason enables us to apprehend the utility, for the public good, of this and that action. But why should the individual approve the public good? . . . Hume’s answer . . . is given in terms of his doctrine of sympathy, which has there the same central position that belief occupies in his treatment of knowledge. The approval is owing to the particular fabric and constitution of our species, and above all to the operation of sympathy, whereby we enter into the sufferings of others as into sufferings of our own. (148) 3. And as Hume notes in the Enquiry Concerning the Principles of Morals, section I, The end of all moral speculation is to teach us our duty; and, by proper representations of the deformity of vice and the beauty of virtue, beget correspondent habits, and engage us to avoid the one, and embrace the other. But is this ever to be expected from inferences and conclusions of the understanding, which of themselves have no hold of the affections nor set in motion the active powers of men? . . . What is

H u m e ’ s M or a l P s yc h ol o g y 

( 93 )

diverges from Hutcheson in an important respect: Hutcheson takes us to be naturally disposed to approve each virtue and to disapprove each vice. Hume sees Hutcheson’s position as requiring a needless multiplication of instincts, preferring to see a general tendency to approve virtue and to disapprove vice modulated and tuned to each of the vices and virtues recognized in a culture by custom. This opens up the social dimension of moral sensibility and the need for moral sensibility to be cultivated. That cultivation is enabled by our susceptibility to pride and humility, as we respond to the approbation or condemnation of our peers and compare ourselves in imagination to those we admire or despise. And indeed, Hume argues that one of the principal functions of socialization is the cultivation and modulation of the moral sentiments:  “Tho’ this progress of the sentiments be natural, and even necessary, ’tis certain, that it is here forwarded by the artifice of politicians, who, in order to govern men more easily, and preserve peace in human society, have endeavour’d to produce an esteem for justice, and an abhorrence of injustice” [3.2.2.23, 500].4

2.  THE SOCIAL DIMENSION Hume opens the discussion of the social dimension of the human moral sentiments by noting that an account of the individual psychology of the moral sentiments would be hopelessly inadequate as a full account of human moral sensibility. We are, after all, social animals, tuned to resonate to each other (a metaphor also used by Shaftesbury [1964]). We therefore honourable, what is fair, what is becoming, what is noble, what is generous, takes possession of the heart, and animates us to embrace and maintain it. What is intelligible, what is evident, what is probable, what is true, procures only the cool assent of the understanding; and gratifying a speculative curiosity, puts an end to our researches. (EPM 1.7; SBN 172) Here Hume emphasizes—​in terms completely consistent with the Treatise account of morals, and connected to the discussion to which we will turn in c­ hapter 9 on the relation between reason and the passions—​that the goal of ethical instruction and of ethical practice and cultivation is the transformation of the passive part of our nature, of our reactions to others, and of the character of our moral perceptual responses, not the satisfaction of idle curiosity about the principles of morals. Again, the affinities to Śāntideva’s account in How to Lead an Awakened Life is striking. See Garfield (2015), for a discussion of Buddhist accounts of moral phenomenology and moral perception. 4. For a clear discussion of the respects in which Hume departs from Hutcheson in these matters, and of the influence of Butler on Hume in this regard, see Schneewind (1983). In this context, Schneewind also correctly emphasizes another respect in which Hume follows Hutcheson, viz., in his insistence that the point of morality is to make human life happier, and that the virtues are approved on the grounds that they conduce to happiness, the vices disapproved on the grounds that they do not.

( 94 )  Book II

develop as persons in response to social engagement, and as we mature, we in turn develop more sophisticated social responses: But besides the original causes of pride and humility, there is a secondary one in the opinions of others, which has an equal influence on the affections. Our reputation, our character, our name are considerations of vast weight and importance; and even the other causes of pride; virtue, beauty and riches; have little influence, when not seconded by the opinions and sentiments of others. In order to account for this phaenomenon ’twill be necessary to take some compass, and first explain the nature of sympathy. No quality of human nature is more remarkable, both in itself and in its consequences, than that propensity we have to sympathize with others, and to receive by communication their inclinations and sentiments, however different from, or even contrary to our own. [2.1.11.1–​2, 316]

We note several important points here. First, Hume draws the distinction between original, or innate, causes of pride and humility and secondary causes. The secondary causes are social, viz., the opinions of others. Note that once again, Hume does not distinguish here between natural and artificial causes, but rather draws a distinction within the natural between the original and the secondary. It is, on this account, perfectly natural for the opinions of others to influence our emotions and to give rise to moral sentiments. Social causes are causes—​things that happen to us, and that give rise to moral assessment—​ and it is natural for us to be responsive to them. Second, Hume draws attention to “our reputation, our character, our name” as “considerations of vast weight and importance.” Pride and humility, he has insisted earlier, are responses that always take the self as their object, whatever their causes. Here he emphasizes that those causes include the approbation and disapprobation of others. Our own view of ourselves is determined by the views of others, and the self we represent is an essentially social self.5 These two observations concern the social causes of our self-​regarding moral attitudes. The third locates in original human nature the cause of 5. This emphasis on the importance of pride and reputation and their relation to the construction of the self has a clear antecedent in Shaftesbury: For what Person is there who in his Imagination joins not something worthy and deserving with his true and native Self, as oft as he is refer’d to it, and made to consider, Who he is? Such is the natural Affection of all Mankind towards moral Beauty and Perfection, that they never fail in making this Presumption in behalf of themselves: “That by Nature they have something estimable and worthy in respect of others of their Kind; and that their genuine, true, and natural Self, is, as it ought

H u m e ’ s M or a l P s yc h ol o g y 

( 95 )

our responsiveness to others, that is, natural sympathy. This is the basis of the first two phenomena: we are sensitive to the opinions of others as the source of moral sentiments and of self-​assessment because of this inner propensity to care about what others think. But this very propensity is also the foundation of the states of character of which we approve in ourselves and others, and which form the foundation of social life and hence of morality. Natural sympathy is one of the most important concealed mechanisms of custom. It links the individual to the social in virtue of being a principle of individual psychology that connects us to our social milieu. Natural sympathy is hence the original disposition that makes both morality and the social life in the context of which we are constituted as persons possible. That natural sympathy is the biological and psychological foundation of the connections that enable us to respond to one another morally; these responses, in turn, form the basis for the complex conventions that extend our moral recognition and responsiveness to those more distant from us, enabling us to constitute an ethical community and standards of moral criticism.6 Hume devotes the next few pages of this section to an exploration of the mechanisms of this natural sympathy. First, he notes, when we sympathize with another—​when we come to share an emotion or an attitude—​we first observe the natural effects of that attitude or emotion in the other person, its “external signs in the countenance and conversation.” Impressions of

to be, of real value in Society, and justly honourable for the sake of its Merit, and good Qualitys.” (1964, 1:174) The impact of Hume’s discussion of the social dimensions of moral assessment, the interaction of social custom and natural moral sentiment, and the dependence of moral judgment on custom is evident in his friend Adam Smith’s Theory of the Moral Sentiments, published in 1759. Its origins are evident in the work of Mandeville: “The highest wish of the Ambitious Man is to have all the World, as to that particular, of his Opinion” (1989, 90). We can also note the similarity to the roles that hiri (shame) and devadhamkara (divine pride) play in the cultivation and reinforcement of ethical sensibility and conduct in the Buddhist moral tradition. 6. And herein lies another affinity between Humean thought and that of the Buddhist tradition, especially the Mahāyāna tradition, the emphasis on what Hume calls sympathy, and what the Buddhists call karuṇā, or care. Candrakīrti opens his Introduction to the Middle Way by characterizing karuṇā as essential to moral development at the beginning, middle, and end of practice (he refers to it as the seed of accomplishment, as the rain that nourishes the crop, and as the harvest). He emphasizes that our natural disposition to care for others motivates moral cultivation; moral development is sustained by the practice of care and by the care of others for us, and moral maturity is characterized by the achievement of a universal care. It is interesting to compare this to Hume’s suggestion that the moral point of view is the most general point of view from which we abstract from our own particular interests, taking those of others seriously as well.

( 96 )  Book II

these natural signs give rise not to an impression of that emotion, which, as an inner state, we cannot directly sense, but rather to an idea of that emotion: When any affection is infus’d by sympathy, it is at first known only by its effects, and by those external signs in the countenance and conversation, which convey an idea of it. This idea is presently converted into an impression, and acquires such a degree of force and vivacity, as to become the very passion itself, and produce an equal emotion, as any original affection. [2.1.11.3, 317]

The function of sympathy is to convert that idea into an impression, an impression which can “become the very passion itself, and produce an equal emotion, as any original affection.” This is important. Much is made of the so-​called copy principle in Book I, according to which all ideas derive from corresponding impressions. It is clear from this passage, and indeed from this entire discussion of sympathy, that this can also work in reverse in the case of socially mediated ideas and impressions, and the mechanism that makes that possible is sympathy. The “concealed influence of custom” in this case is the manufacture of an impression so powerful that it becomes a passion capable of driving our behavior. This occurs through the mediation of sympathy, an original principle of psychology that can supply the necessary force and vivacity lacking in the idea itself. Here is how that works: the idea that we form of the inner states of others delivered by “external signs in the countenance and conversation” is associated via sympathy with the impression of our own person—​ that socially constructed identity we explored in ­chapter 3—​to which we always have access.7 That association, facilitated by the sympathy that enables our recognition as beings of the same kind, calls up an impression of reflection, the impression of our own experience of inner states of that 7. Hume actually writes, “’Tis evident, that the idea, or rather impression of ourselves is always intimately present with us” [2.1.11.4,  317]. He elsewhere makes a similar claim: “’Tis evident, that as we are at all times intimately conscious of ourselves, our sentiments and passions, their ideas must strike upon us with greater vivacity than the ideas of the sentiments of any other person” [2.2.2.13, 339]. This might be taken to mean that we are always explicitly aware of ourselves, in a kind of perpetual narcissism. Not only is this implausible on its face, but it is inconsistent with Hume’s suggestion that when we take a moral point of view we adopt an impartial perspective, abstracting from our own perspective. The narcissistic reading of these passages is not correct. The “intimate” consciousness of which Hume speaks need not be thought of as occurrent awareness, but rather simply as immediate first-​person availability. We are, he emphasizes, always able to become conscious of who we are; we never misidentify ourselves as someone else, or our mental states as those of another. We can then understand the states of others on analogy with those we can always find in ourselves.

H u m e ’ s M or a l P s yc h ol o g y 

( 97 )

kind, and that impression of reflection—​a passion—​takes not only our self as its object, but also, again mediated by sympathy, our fellow, whose countenance and conversation are the cause of the passion. Now, sympathy, as a kind of natural force governing our affective life, is subject to natural laws; and, as we noted above, we must appeal to its rapid decay with social distance to explain its operation: Now ’tis obvious, that nature has preserv’d a great resemblance among all human creatures, and that we never remark any passion or principle in others, of which, in some degree or other, we may not find a parallel in ourselves. The case is the same with the fabric of the mind, as with that of the body. . . . Accordingly we find, that where, beside the general resemblance of our natures, there is any peculiar similarity in our manners, or character, or country, or language, it facilitates the sympathy. The stronger the relation is betwixt ourselves and any object, the more easily does the imagination make the transition, and convey to the related idea the vivacity of conception, with which we always form the idea of our own person. [2.1.11.5, 318]

This discussion is important in two respects. First, it represents a sketch of Hume’s theory of the mechanism of sympathy and its connection to the imagination. One of Hume’s distinctive contributions, as we will see in greater detail when we consider his skepticism in ­chapters 6–​8 as well as his account of moral development in c­ hapter 11, is the central role he assigns the imagination both in our psychology and in his epistemology. The tripartite account of mind as comprising the senses, imagination, and understanding was common currency in early modern European thought. But while Berkeley, for instance, prioritizes the senses as the origin of knowledge, and Descartes the understanding, for Hume the imagination takes center stage as the arena in which our innate psychological processes construct the world we inhabit. Much of the Treatise is devoted to showing that operations we might naively assign to the senses or to reason in fact belong to the imagination.8 For this reason, it may be worth pausing for a moment to reflect on just what the imagination is. The term is antique, and is deployed in diverse ways, none of which map neatly onto either the conceptual apparatus of contemporary cognitive science or onto the way the term imagination is

8.  See Hardin (2007) for a defense of this psychological reading of Hume. Hardin says that Hume “is, as he says forcefully and repeatedly, concerned to present a scientific theory of our moral beliefs—​not of the content of morality in principle but of our beliefs in fact and why we believe those things.” (2007, 200) I agree completely.

( 98 )  Book II

used in contemporary English discourse, either colloquial or technical. One way to see the difference between Descartes’s and Hume’s approach to the mind—​and, I think, a respect in which Hume makes a significant advance over Descartes—​is that while Descartes sees the mind in terms of interlocking faculties, each with its own domain of cognitive objects, Hume sees the mind as comprising a large set of law-​governed capacities. So, for Descartes, the imagination is the faculty that cognizes images. That is why it is incapable of knowing the piece of wax or of resolving the difference between a chiliagon and a 999-​sided polygon. The understanding cognizes abstract ideas, and the senses only sensations. As Garrett (1997) notes, Hume follows Locke and Berkeley in his thought about the imagination in two respects: first, he rejects the idea that beyond the imagination we must posit an intellect as a second representational faculty; second, he locates most cognitive functioning in the imagination, as an interplay of representational states. But there are differences between Hume’s account and that of his empiricist predecessors as well. For Hume, the imagination comprises the set of cognitive processes that operate on representations in blind, automatic ways. It is the domain of the construction of ideas from impressions; impressions from ideas; ideas from ideas; the domain where complex ideas are constructed from simple ideas; the domain in which sympathy operates. The senses comprise other cognitive capacities, those of transducing information from outside to produce cognitive objects on which the imagination can work. And the understanding comprises the faculties of deliberate, rational thought. Hume’s imagination is hence neither a recapitulation of Descartes’s imagination, nor a repetition of the Berkeley-​Locke account of the imagination, but a somewhat primitive model for Kant’s. On the other hand, Hume, unlike Kant, sees much of what passes for reasoning not as the result of deliberate and rational cogitation, or even active but unconscious synthesis, but rather as the result of blind, original cognitive processes governed by lawlike principles of association. In this case, sympathy is the cognitive process at issue, and the discovery of the laws governing it is a basic task of moral psychology. This leads us to the second point:  the structure of this discussion confirms that he is thinking of sympathy very much as a psychic force, akin to gravitation in the physical world. This idea is introduced here, and, as we shall see in c­ hapters  10 and 11, will play a crucial role in Hume’s account of justice and benevolence and his account of moral paideia. When Hume speaks of similarities and resemblances as increasing the strength of sympathy, facilitating the transfer of vivacity, he is constructing a multidimensional space of similarity, and concluding that the force of sympathy

H u m e ’ s M or a l P s yc h ol o g y 

( 99 )

is stronger the nearer the subject and object of that sympathy are to one another in that space. That is not a theory of reasoning, but a theory of forces, and it is a theory of forces that owes a clear debt to Newton. Here we see Hume limning the concealed influences of custom and, in particular, the complex interface between custom in the habitual senses and custom in the social sense.9 Hume reinforces this link in the following discussion: The lively idea of any object always approaches its impression; and ’tis certain we may feel sickness and pain from the mere force of imagination, and make a malady real by often thinking of it, . . . and ’tis there principally that a lively idea is converted into an impression. Our affections depend more upon ourselves, and the internal operations of the mind, than any of the impressions; for which reason they arise more naturally from the imagination, and from every lively idea we form of them. This is the nature and cause of sympathy; and ’tis after this manner we enter so deep into the opinions and affections of others, whenever we discover them. [2.1.11.7, 319]

First, Hume emphasizes that the imagination—​again, we must think of this as a collection of autonomic psychological processes—​is capable of the production of impressions from ideas, reinforcing the point that this traffic is not, as it is initially presented in Book I, one way. Hume then points out that our “affections,” or our emotions, are not in general directly elicited by

9. It is also plain that Hume thinks of sympathy in this broad sense simply as a force, and not as an experience that is, in itself, either pleasant or painful. In a letter to Adam Smith, he writes: I am told that you are preparing a new Edition, & propose to make some Additions & Alterations, in order to obviate Objections. . . . I wish you had more particularly and fully prov’d, that all kinds of Sympathy are necessarily Agreeable. This is the Hinge of your System, & yet you only mention the Matter cursorily in p. 20. Now it would appear that there is a disagreeable Sympathy, as well as an agreeable: And indeed, as the Sympathetic Passion is a reflex Image of the principal, it must partake of its Qualities, & be painful where that is so. Indeed, when we converse with a man with whom we can entirely sympathize, that is, where there is a warm & intimate Friendship, the cordial openness of such a Commerce overpowers the Pain of a disagreeable Sympathy, and renders the whole Movement agreeable. But in ordinary Cases, this cannot have place. An ill-​humord Fellow; a man tir’d & disgusted with every thing, always ènnuié; sickly, complaining, embarrass’d; such a one throws an evident Damp on Company, which I suppose wou’d be accounted for by Sympathy; and yet is disagreeable. It is always thought a difficult Problem to account for the Pleasure, receiv’d from the Tears & Grief & Sympathy of Tragedy; which would not be the Case, if all Sympathy was agreeable. An Hospital would be a more entertaining Place than a Ball. (Letter 169, to Adam Smith, 1759, in Hume 1932, 312–​13)

( 100 )  Book II

external phenomena, but rather from impressions of reflection produced in the mind. On this psychological theory of sympathy, the external contribution comprises perceptions of the behavioral and verbal manifestations of the emotional lives of others. We take those impressions and from them form the idea of the affect that must have given rise to them; we then automatically convert those ideas into impressions of reflection, adding the necessary vivacity through cognitive reflex. Today, we would talk about mirror neurons or cognitive simulation in this context. A bit later in this section Hume emphasizes this point: In sympathy there is an evident conversion of an idea into an impression. This conversion arises from the relation of objects to ourself. Ourself is always intimately present to us. Let us compare all these circumstances, and we shall find, that sympathy is exactly correspondent to the operations of our understanding; and even contains something more surprising and extraordinary. [2.1.11.8, 320]

What does Hume mean here by the fact that the “conversion arises from the relation of objects to ourself?” Well, there must be a source for the additional force and vivacity communicated to an idea to transform it into an impression. Energy in the mind, like energy in bodies, does not arise ex nihilo. The source of that psychic energy, as we saw previously, is the impression we have of our self, an impression readily available. The impressions that are most intimately related to those with which we identify are those arising from the behavior of others, with those closest to us in the social similarity space identified more closely than those more distant. The communication of psychic energy follows this proximity and allows a kind of expansion of our self to include others, and the attitudes we have toward ourselves to be extended to those others as objects. It is these causal principles that Hume argued in Book I  characterize the operations of the understanding that Hume finds in analogy in the imagination. When we reason, we use relations of contiguity and causation, among others, to draw conclusions. In reasoning, those relations are, according to Hume, both represented in and causally explanatory of our rational processes. In sympathy, we find the same thing. We represent relations of proximity and causation between ourselves and others, and those relations are causally relevant to the arousal of sympathy itself. What, then, is so surprising? Only this: that in the case of reasoning, we always move from impressions to ideas, and then reason about the relations between ideas; but in sympathy, we move the other way around. But that reverse movement, as we have also seen, is only apparent; the mediation of

H u m e ’ s M or a l P s yc h ol o g y 

( 101 )

the impression of ourself is what is supplying the power that generates the passion from the idea of the inner state of another. This is no small point, when human nature is the topic under discussion. Hume is, in the heart of his discussion of our moral psychology, pointing out just how deeply our social nature goes. Other persons are not simply objects among other objects:  they (and perhaps other animals by extension or analogy) are so intimately related to us that their very identities overlap ours. That is, unlike inanimate objects or their properties, the ideas of which can never generate impressions, the ideas of persons and their properties can go the other way: in virtue of their strong natural association with our impressions of ourselves, they can become impressions and can move us directly. This is the “something more surprising and extraordinary,” and it is what makes us human, what makes society and morality possible, and what shows just how much we are causally implicated in the social networks we inhabit. It is also “surprising and extraordinary” how much of who we are and what we do is to be explained through the passions automatically aroused by those with whom we associate. When we turn to Hume’s moral theory, we will find another important role for the imagination in relation to natural sympathy—​its role in extending our range of concern beyond our immediate circle to the others in our society and even beyond, as we construct the natural artifice of justice and related virtues.

PA R T  I I I

vwv Books I and III The Skeptical Framework Deployed

C H A P T E R  6

w

Epistemological Foundations

M

ost discussions of Hume’s Treatise begin at this point. I  have not done so, as I take the real foundations of the project of the Treatise to be the complex psychology and treatment of custom in both its individual and social dimensions, and against its legal backdrop. With that treatment in view, we can really appreciate the claims and arguments Hume offers when he advances his epistemology in Book I. In this chapter, we will examine some of the more explicitly epistemological foundations of Hume’s account of human nature. We begin with Hume’s own official methodological reflections in the introduction, where he makes explicit his commitments both to a naturalistic Pyrrhonism, and to an empirical approach to understanding knowl­ edge and morality. We will then turn to Hume’s account of ideas, and of meaning. In that discussion we will see how his nominalism often leads him to the conclusion that we can be mistaken in thinking that we have ideas that we take ourselves to have, and how his nominalism grounds a communitarian understanding of meaning and thought. We conclude with attention to Hume’s initial discussions of our knowledge of the external world and our use of causal reasoning. We will see in this context that the nexus of psychology and social science is the key to Hume’s naturalism, his skepticism, and his methodology in general. Let us begin with attention to the introduction of the Treatise.

( 105 )

( 106 )  Books I and III

1.  METHODOLOGICAL LESSONS FROM THE INTRODUCTION Hume presents the experimental naturalism he recommends in the philosophy of mind in the introduction. Many have noted that in the eighteenth century the word experiment could simply mean experience, as opposed to a controlled manipulation of variables, and Hume does sometimes use it that way. But not always. Here he directly advocates what we would recognize today as experimental philosophy or even experimental psychology as a way of bringing the study of the mind in line with that of the new experimental science of physics. Recall a passage we discussed in c­ hapter 1: For to me it seems evident, that the essence of the mind being equally unknown to us with that of external bodies, it must be equally impossible to form any notion of its powers and qualities otherwise than from careful and exact experiments, and the observation of those particular effects. [Introduction 8, xvii]

It is clear that when Hume talks about “the observation of those particular effects, which result from its different circumstances and situations,” he is suggesting that we come to know the “powers and qualities” of the mind through the experimental method. And he is sensible of both the limitations of this method and of its difference from what he regards as the pretensions of a more a priori philosophical approach. He rejects the philosophical program of his predecessors such as Descartes and even Berkeley, who are after the “ultimate original qualities of human nature” [introduction 8, xvii], and while striving for the theoretical virtue of simplicity, Hume takes that quest to be limited by what experience can deliver.1 With Book II in view, we can see clearly the sources of those limitations: human beings are determined in their operation not only by internal principles or mathematically characterizable relations to one another, like physical bodies, but also by social customs. These customs vary over time and circumstance, generating enough variance that experience will yield not a simple explanation of uniform behavior, but rather an account of the sources of variability. Nonetheless, they issue in behavior and cognition that 1.  We hear echoes of this insight of Hume’s in William James’s remark, “The attempt at introspective analysis  .  .  .  is in fact like seizing a spinning top to catch its motion, or trying to turn up the gas quickly enough to see the darkness” (James 1918, 244). Hume’s insistence on an experimental method (again, one we must concede that he does not always in fact employ) is a clear antecedent to James’s critique of introspectionism.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 107 )

is sufficiently regular that we can appeal to those customs in explanations. As a consequence, these customs warrant scientific hypotheses and inferences, just as social customs constitute warrant and rights in the legal sphere. This is why anyone who looks to the Treatise for the exceptionless uniformity of pure metaphysics is reading the wrong book. Hume amplifies this point in the next paragraph, a serious caution against an overly systematic reading of the Treatise, and one that emphasizes his methodology: I do not think a philosopher, who would apply himself so earnestly to the explaining the ultimate principles of the soul, would show himself a great master in that very science of human nature, which he pretends to explain, or very knowing in what is naturally satisfactory to the mind of man. For nothing is more certain, than that despair has almost the same effect upon us with enjoyment, and that we are no sooner acquainted with the impossibility of satisfying any desire, then the desire itself vanishes. When we see, that we have arrived at the utmost extent of human reason, we sit down contented; tho’ we be perfectly satisfied in the main of our ignorance, and perceive that we can give no reason for our most general and most refined principles, beside our experience of their reality; which is the reason of the mere vulgar, and what it required no study first to have discovered for the most particular and most extraordinary phaenomenon. And as this impossibility of making any farther progress is enough to satisfy a reader, so the writer may derive a more delicate satisfaction from the free confession of his ignorance, and from his prudence in avoiding that error, into which so many have fallen, of imposing their conjectures and hypotheses on the world for the most certain principles. [Introduction 9, xvii–​xviii]

Let us work through this discussion with some care. Hume begins by reiterating the imprudence of seeking a completely general account of the mind, suggesting that doing so demonstrates that one is no “master in that very science of human nature.” Why not? Well, first of all, there is the fact of variability: “Despair has almost the same effect upon us with enjoyment.” Different individuals respond to the same circumstances differently, or to very different circumstances in the same way. Moreover, there is a social dimension to behavior that induces more than mere individual variability, and that does more than obscure our access to the springs of our own nature: our original propensities manifest differently in different social contexts; so, when we study our own behav­ior and thought, we learn only about how our nature is manifest in the society in which we live, with no possibility of sorting out the effects of our original constitution from those of the environment in which we develop

( 108 )  Books I and III

our second nature. To generalize too hastily even from a careful survey of our own community would be like attempting to form a general theory of linguistics only by studying English: we would be doomed to failure, even were we to take into account the variations among English speakers. Hume follows this observation with a nice reflexive application:  the despair we might encounter in failing to formulate a complete, apodictic theory of human nature should lead us immediately to contentment: when we realize that there is no exceptionless generalization to be found, we can be satisfied with the fact that we can have only a general, fallible, empirical sense of the terrain, “the reason of the vulgar.” This fallibility is compounded by our own sense of the fallibility of our cognitive and introspective apparatus, and, as we shall see in a moment, by a final problem—​ that of ecological validity. The task of philosophy is thus, Hume argues, both an empirical survey of human nature and a limitation of the desire to transgress that survey; a prescription that puts one immediately in mind of those of Sextus Empiricus. Sextus, like Hume, thought that part of the point of philosophy was to bring us to the ataraxia that follows the recognition that the knowl­ edge we may seek is not possible to have. And like Hume, he argued that an understanding of the customs that guide our ordinary behavior is not only good enough, but all we could ever coherently desire. Hume concludes the introduction with this paragraph: But if this impossibility of explaining ultimate principles should be esteemed a defect in the science of man, I will venture to affirm, that ’tis a defect common to it with all the sciences, and all the arts, in which we can employ ourselves, whether they be such as are cultivated in the schools of the philosophers, or practiced in the shops of the meanest artizans. None of them can go beyond experience, or establish any principles which are not founded on that authority. [Introduction 10, xviii]

Again, a careful reading informed by the account in Book II shows just how important these methodological preliminaries are to the epistemological project of Book I, and how deep Hume’s Pyrrhonism goes. First, Hume emphasizes that the limitations of the science of human nature to which he adverts in the previous paragraphs are not restricted to the psychological domain; instead, because of the dependence of all knowledge on our psychological processes and limitations, they are common to all knowledge, whether academic or craft. All knowledge is beholden to empirical evidence, and so is hostage to its limitations. We will subsequently see this principle come home to roost in Hume’s discussion of skepticism

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 109 )

with regard to reason. The epistemic situation of the science of the mind, however, is special: Moral philosophy has, indeed, this peculiar disadvantage, which is not found in natural, that in collecting its experiments, it cannot make them purposely, with premeditation, and after such a manner as to satisfy itself concerning every particular difficulty which may arise. When I am at a loss to know the effects of one body upon another in any situation, I need only put them in that situation, and observe what results from it. But should I endeavor to clear up after the same manner any doubt in moral philosophy, by placing myself in the same case with that which I  consider, ’tis evident this reflection and premeditation would so disturb the operation of my natural principles, as must render it impossible to form any just conclusion from the phaenomenon. We must therefore glean up our experiments in this science from a cautious observation of human life, and take them as they appear in the common course of the world, by men’s behavior in company, in affairs, and in their pleasures. Where experiments of this kind are judiciously collected and compared, we may hope to establish on them a science, which will not be inferior in certainty, and will be much superior in utility to any other human comprehension. [Introduction 10, xix]

The problem Hume addresses here is not metaphysical, but methodological, and it is one familiar to any experimental psychologist: it is what J. J. Gibson (1979) called the problem of ecological validity. When we try to run an experiment to study how the mind works naturally, we “disturb the operation of [the] natural principles,” and so find out only how the mind operates in the experimental situation. And to think that we could remediate this by simply observing ourselves in ordinary life is foolish. I would only know how I appear to think when being observed. To find out by observation how the mind works when not being observed is simply impossible. So, all we can reasonably do is to engage in “a cautious observation of human life, and take them as they appear in the common course of the world, by men’s behavior in company, in affairs, and in their pleasures.” Now, psychology has come a long way since the eighteenth century, and we can now essay more than a mere observational study of the mind. Nonetheless, Hume’s caution is still important. First of all, it is a wise methodological admonition to those of Hume’s—​and our—​contemporaries who would study the mind introspectively or through the device of pure reason. Second, it reminds us that even the most exquisite understanding of how an individual human being behaves in a controlled situation can only be part of the story of human behavior. As social beings, and as beings whose

( 110 )  Books I and III

psychological states are responsive to norms, the science of persons is the science of “men’s behavior in company, in affairs.” Hume would call this the science of custom. We forget this at our peril.

2. IDEAS Hume’s distinction between impressions and ideas and his doctrine of the source of all ideas in impressions is well-​trodden ground, and I have nothing to add to that literature, save to point out that that very connection—​as well as our belief in it—​is a matter of custom, that is, custom in that sense of habit, or the way the organism customarily operates. For that reason, even the claim that all simple ideas originate in simple impressions also can be taken as nothing more than a general maxim, an empirical observation, but one subject to exception and approximation, as is shown, for example, in the case of the missing shade of blue, an exception that Hume acknowledges, and which he takes not to undermine the force of the gen­ eral principle.2 The fact that Hume says that the case of the missing shade of blue, while a counterexample, does not undermine his thesis should be taken neither as evidence of his disregard for counterexamples, nor as a suggestion that he has a way of accommodating the example up his sleeve to be discovered by his commentators. He says neither of these things. Instead, it should be taken as a positive example that confirms Hume’s metaphilosophical thesis enunciated in the introduction to the Treatise that all psychological generalizations are approximate and defeasible. This is no different from the way generalizations are treated in contemporary psychological research: a single counterexample to a statistically verified psychological generalization does not defeat it. We expect to confirm or to reject hypotheses by analysis of variance, not by discovering a complete absence of variance. And this is a point that applies to all science, not just to the science of human nature, as such diverse philosophers of science as Grünbaum,

2.  The problem of the missing shade of blue, in brief, is this:  imagine somebody who has never seen a particular shade of blue, say blue23. Now, suppose we arrange all shades of blue in order from blue1 to bluen, leaving blue23 out of the series. It seems that this person would be able to form the simple idea of blue23 from the series, even though she has never had the corresponding impression. So, Hume argues, there is a counterexample to the generalization that all simple ideas derive from corresponding simple impressions. On the other hand, Hume remarks, the instance is so particular and singular, that “’tis scarce worth our observing, and does not merit that for it alone we should alter our general maxim” [1.1.2.10, 6].

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 111 )

Quine, Feyerabend, van Frassen, and Cartwright, all in Hume’s lineage, have pointed out.3 One of the first uses to which Hume puts his account of impressions and ideas is in the investigation of the supposed ideas of substance and accident. This discussion is instructive because of the clarity with which it demonstrates the importance of the nominalist Pseudo-​idea Principle for interpreting the Treatise: we often believe that we have ideas that we in fact do not have. For instance, we think that we have an idea of a permanent self; we do not, only a set of conventions for referring to persons; we think that we have an idea of substance; we do not, only an idea of objects that may have properties. Hume’s diagnosis of this kind of illusion is that we confuse the fact that we use a term fluently with the putative fact that the term actually denotes something. It also illustrates the importance of the Cover Principle: when it appears that Hume is engaged in metaphysics, we should turn to the cover of the book and remind ourselves that he is actually studying human nature. Here is the relevant passage: I wou’d fain ask those philosophers, who found so much of their reasonings on the distinction of substance and accident, and imagine we have clear ideas of each, whether the idea of substance be deriv’d from the impressions of sensation or reflexion? If it be convey’d to us by our senses, I ask, which of them; and after

3.  There is an entire philosophical cottage industry devoted to the missing shade of blue. It would take us far afield to canvass all of the competing understandings of Hume’s quick dismissal of what looks like a counterexample to a central thesis of his psychology. Besides the explanation I offer here, the only other explanation that strikes me as plausible is that offered by Garrett (1997). He argues that Hume is untroubled by the example since the imagination should be capable of filling in the gap in an ordered sequence. It is worth considering that explanation for a moment, as seeing why, despite its plausibility, it doesn’t in the end work is instructive regarding Hume’s project more generally. There is a plausibility to the idea that Hume intends that the imagination can fill in the missing idea without the corresponding impression, for Hume assigns quite a lot of power to the imagination. On the other hand, we should note that Hume doesn’t say that, and gives no similar example of the imagination filling in gaps in sequences like this. It strikes me as more plausible that Hume simply sees these general maxims as susceptible of occasional exceptions to be explained as science progresses, a kind of recognition of the approximate and preliminary character of any principles of human behavior one might propose in the eighteenth century. The need to see what Garrett calls the “copy principle” as exceptionless derives, I suspect, from confusing the enterprise of psychology with that of metaphysics. This is not an a priori analysis of reason, but an empirical generalization. And we might note that even contemporary psychological laws often have a statistical character. Moreover, the missing shade of blue is an exception only to the generalization that every determinate idea comes from a corresponding determinate impression, not to the claim that all ideas derive from impressions in some way.

( 112 )  Books I and III what manner? If it be perceiv’d by the eyes, it must be a color; if by the ears, a sound; if by the palate, a taste; and so of the other senses. But I believe none will assert, that substance is either a colour, or a sound, or a taste. The idea of substance must therefore be deriv’d from an impression of reflexion, if it really exist. But the impressions of reflexion resolve themselves into our passions and emotions; none of which can possibly represent a substance. We have therefore no idea of substance, distinct from that of a collection of particular qualities, nor have we any other meaning when we either talk or reason concerning it. [1.1.6.1,  15–​16]

It might appear that Hume is asking whether things in the world are substances characterized by accidents, and answering in the negative. That would miss the point of this discussion pretty badly. Instead, Hume is showing us something not about basic metaphysics, but about ourselves and about philosophical practice: he is putting his basic psychology to work to show that often, when we think that we are engaging in profound philosophical reflection, we are in fact talking nonsense. Hume begins by asking whether those who speak of substance and accident really mean anything at all. If they do, their words must express an idea. If so, that idea must come from some impression.4 But the putative idea of substance is said to be the idea of something nonsensible, and so cannot come from any impression. Hence there can be no such idea. In a letter to Henry Home, Hume writes: As to the idea of substance, I must own, that as it has no access to the mind by any of our senses or feelings, it has always appeared to me to be nothing but an imaginary centre of union amongst the different and variable qualities that are to be found in every piece of matter. But I shall keep myself in suspense till I hear your opinion. (Letter 52, to Henry Home, May or June 1746, in Hume 1932, vol. 1, 94)

What are we meant to learn from this discussion? Not anything about the metaphysics of substance and attribute. First, we learn that we are capable of talking complete nonsense when we think we are making sense. But more important is the general moral Hume takes from this discussion, and 4. Note that the missing shade of blue problem does not undermine this principle as it is put to use here—​or for that matter, anywhere in the study—​as even the idea of the missing shade of blue originates in some impressions related as other determinates within the sensible determinable blue; it is not an example of an idea generated with no corresponding impressions, but an example of an idea generated in an unusual way from impressions.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 113 )

which gets put to work time and again in the Treatise: We often take ourselves to know our minds much better than we in fact do. We think we have ideas that we do not have. We then deploy those pseudo-​ideas not only in our understanding of the external world, but also in our understanding of ourselves, compounding nonsense with further nonsense. The most obvious first instance of this, Hume would argue, is the Cartesian idea that we are mental substances with mental attributes. This view might appear self-​evident. Even if it does not appear self-​evidently true, it might nonetheless appear self-​evidently possible; or even if it might not appear self-​evidently possible, it might appear self-​evidently to make sense. Hume denies even that. When we utter these words, they are not even false; they are simply nonsense, simply because the term substance conveys no idea, and fails to refer at all. We start with the illusion that we mean something, convince ourselves that what we mean is true, and end up with the illusion that we know our own natures. Much better, Hume urges, to give up the quest for natures, and to examine carefully, in terms derived from experience and therefore meaningful, our actual practice (as Sextus Empiricus advises in the Outlines of Pyrrhonism). And this short discussion is, he would aver, an example of doing just that: examining our verbal behavior to reveal its deceptiveness and nonsense. We will see this strategy deployed repeatedly in the Treatise. There is one more thing to say before we leave this paragraph. It is also a good initial example of Hume’s Pyrrhonian skeptical inversion. One can imagine two extreme positions regarding existence and substance: on the one hand, one might argue that we exist as minds—​or that my table exists as a material object—​because I  am, or it is. a substance with attributes; or, one might argue that since the metaphysics of substance and attribute makes no sense, there are no substances and no attributes, and so nothing really exists. The first position is the extreme of reification; the second the extreme of nihilism. Hume rejects both, finding a middle path neither through compromise, nor through indecision, but rather through epochē. That is, he rejects the very terms of the debate, the presupposition shared by both extreme positions, viz., that to exist is to be a substance with attributes. He rejects it as nonsense, and so sets aside the entire discussion. But he does not therefore conclude that neither he nor anyone else exists; the entire Treatise presupposes the existence of persons and their robust interaction with one another. Instead, as we will see in the context of the discussion of skepticism regarding the external world, Hume takes our ontology to be a matter of custom. What we take to be existent is what forces itself on us psychologically or what makes sense of our linguistic and scientific

( 114 )  Books I and III

conventions; that is, as a consequence of passions to which are susceptible. Custom, as Sextus noted in the context of the fourfold prescription, frames the positive side of the skeptic’s practice. Hume’s method is hence precisely that of a Pyrrhonian skeptic, and, just as we see it deployed here, we will see this strategy deployed throughout the Treatise. We now turn to Hume’s account of abstract ideas, an account developed in response to Berkeley’s failed attempt to characterize abstraction, and one in which, once again, custom plays a crucial role: Abstract ideas are therefore in themselves individual, however they may become general in their representation. The image in the mind is only that of a particular object, tho’ the application of it in our reasoning be the same, as if it were universal. [1.1.7.6, 20]

Thus far, Hume is simply following Berkeley, who argued that all ideas, being images, are particular, but that they can be used generally. As Sellars (1977) points out, Berkeley has no real official account of how that is possible, and rests his view on an uncritically inherited and incoherent Aristotelian account of abstraction. Moreover, any attempt to extend his view on its own terms leads to catastrophe, as Berkeley assimilates the generality of the content of an idea to the generality of the causal role of the idea. In other words, he assimilates a representation of a general concept to a general kind of representation. The fundamental problem is that Berkeley has no real semantic account of intentionality, and never links the intentionality of thought to that of language. Hume sees that problem, and the way out, and gives a central role to language, and hence to custom, in gen­ eral thought: This application of ideas beyond their nature proceeds from our collecting all their possible degrees of quantity and quality in such an imperfect manner as may serve the purposes of life, which is the second proposition I  propos’d to explain. When we have found a resemblance among several objects, that often occur to us, we apply the same name to all of them, whatever differences we may observe in the degrees of the quantity and quality, and whatever other differences may appear among them. After we have acquired a custom of this kind, the hearing of that name revives the idea of one of these objects, and makes the imagination conceive it with all its particular circumstances and proportions. [1.1.7.7, 20]

Here is Hume’s alternative explanation of general use of an idea, an explanation in which his naturalism and nominalism are very much in evidence.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 115 )

The important move is the association of a word with a set of objects. The association of a word with objects that resemble each other is nothing more than what later behaviorists would call “stimulus generalization,” and is, Hume sees, a simply causal notion. But when the association of the word with an equivalence class of stimuli enters the social domain, the use of the word to refer to that range of phenomena becomes rule-​governed and so customary in the social sense of custom. That custom in turn is internalized through training and habituation, and so in the habitual sense of custom, we come customarily to associate that word with things of that kind in accordance with the rule. As we have seen, Hume is committed to the view that it is custom that gives normative force to rules. This explains how we come to use the word and the associated idea in a rule-​governed way, and so intentionally.5 Now, this is not the sophisticated account of the relation between the intentionality of language and that of thought that we find in the later Wittgenstein or in Sellars, but it is a clear prototype, and a huge advance over Berkeley’s primitive notion.6 Hume summarizes the point as follows: But as the same word is suppos’d to have been frequently applied to other individuals, that are different in many respects from that idea, which is immediately present to the mind; the word not being able to revive the idea of all these individuals, only touches the soul, if I may be allow’d so to speak, and revives that custom, which we have acquire’d by surveying them. They are not really and in fact present to the mind, but only in power; nor do we draw them all out

5. We therefore begin by inculcating a habit (a custom in the individualistic sense); thereby come to participate in a social custom; and in virtue of participating in such a custom (again note the parallel to the legal sense of custom and its power to induce normative force that Hume has in the back of his mind) come to follow rules and to mean things by our words. One might think (particularly in the eighteenth century) that reference—​the association of an idea with a word—​is the only kind of meaning, or at least the foundation of meaning. But Hume is suggesting that rule-​following is the foundation, and that words are only even associated with referents—​including ideas, in virtue of being used in customary ways. Kripke (1982) correctly notes the parallel between Hume’s account of our idea of causation and Wittgenstein’s account of meaning as use. He is correct to remark on the apparent influence of Hume on Wittgenstein; but he misses the fact that the line of influence on Wittgenstein’s thinking about language may be even more direct, proceeding immediately from Hume’s own reflections on the centrality of custom to meaning. 6. See also Baxter (2008) for a clear account of the relationship between Hume’s own nominalist account of abstract ideas and those of his predecessors. Also see Ainslie (2005, 65–​66) who notes that Hume draws the parallel between the collective, rule-​ governed practice of language in constituting meaning and the role of collective rule-​ governed practices in instituting property rights [3.2.2.10, 492].

( 116 )  Books I and III distinctly in the imagination, but keep ourselves in a readiness to survey any of them, as we may be prompted by a present design or necessity. The word raises up an individual idea, along with a certain custom; and that custom produces any other individual one, for which we may have occasion. But as the production of all the ideas, to which the name may be apply’d, is in most cases impossible, we abridge that work by a more partial consideration, and find but few inconveniences to arise in our reasoning from that abridgment. [1.1.6.7, 20–​21]

There are a couple of nice points to note here. The first is that, again, unlike Berkeley, Hume does not see the general reference of general terms to be occurrent—​bringing all instances to mind at once–​but rather sees it as dispositional, offering a causal-​dispositional account of meaning. A word or idea has a general meaning when it can be used in a rule-​governed way in a range of circumstances.7 Second, Hume sees a role for the imagination in this process. Now, Hume does not theorize the productive imagination

7. Tanaka (2016) notes and solves an interesting problem one might raise for Hume’s account:  infralinguistic humans and animals do not use language, but nonetheless must recognize events or objects as instantiating types in order to engage in causal reasoning, or even to learn language, which itself requires recognizing both words and their referents as instances of types. So, if we can only recognize things as instantiating types if we already annex a general term, it would be impossible for infralinguals to engage in causal reasoning, and even impossible for language to ever be learned in the first place, given an insuperable bootstrapping problem. Moreover, as Tanaka points out, Hume himself sees it as one merit of his system that it provides a uniform account of the basic psychology of persons and animals, including the causal reasoning of animals (1.3.16). Tanaka, following the lead of Gamboa (2007), argues that Hume can distinguish between two levels of general thinking. Fully general thinking (what Tanaka calls “level-​II general thinking”) involves the recognition of generality as generality and requires language. We might say that it involves custom in the sense of social convention. Level-​I general thinking, however, simply reflects innate propensities to associate phenomena, or, as Tanaka puts it, “natural sensitivity to specific aspects of objects.” This set of innate sensitivities and dispositions to respond to sequences of stimuli of particular kinds, despite not being aware of them as of those kinds, is what enables basic causal reasoning and language acquisition. As Tanaka observes, this distinction is borne out by much recent work in cognitive science. Tanaka also directs our attention to EHU 9.2 (SBN 105), where Hume argues that “animals as well as men learn many things from experience, and infer, that the same events will always follow from similar causes,” using animal training as an example, but also pointing out that animals share our confidence in the reality of the external world (Tanaka 2016). It is evident that Hume takes such causal reasoning to involve custom in the sense of habit, but not in the sense of social convention. Hardin (2007) points out that Hume takes this concern for our moral relations to animals from Shaftesbury. This distinction is akin to that Sellars (1980) draws between what he calls “animal representation systems” and full-​fledged linguistic systems, and reflects the psychology of Tolman and his account of cognitive maps. See also Garfield (1988) for a discussion of the way that this distinction can solve bootstrapping problems in language acquisition.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 117 )

as Kant was to do, but Hume here and elsewhere in the Treatise, gives the imagination a central place in thought, and indeed grounds a lot of our most important cognitive powers in the imagination, an idea that Kant would take up in the Schematism in the first Critique.8 Unlike Descartes, who foregrounds reason, or Berkeley, who locates everything important in the senses, Hume sees the need for an active faculty that mediates between the two. This insight marks the discovery in the early modern period of the need for a theory of mental activity, even if it does not come with the necessary theory of that mental activity, and it is also the prescient linking of the theory of intentionality to an account of the publicity of language. Hume sums it all up as follows: A particular idea becomes general by being an annex’d to a general term; that is, to a term, which from the customary conjunction has a relation to many other particular ideas, and readily recalls them in the imagination. [1.1.7.10, 22]

8. But, as Garrett (1997) notes, Hume, despite his sense of the importance of psychological processes in the explanation of cognition and behavior, and despite his insistence on the relevance of content to the causal laws connecting perceptions one to another, remains fixated, like Locke and Berkeley before him, on the image as the model of a perception, as opposed to a judgment, and so impoverishes his psychology considerably. It was left to Kant to develop a theory of mental activity with room both for perceptual content and logical structure, and to identify the place of the imagination in thought with the requisite precision. In one sense, then, Kant’s richer and more specific sense of the imagination as the faculty of schematized synthesis is an improvement on Hume’s rather flat account of a domain of associative laws. The Kantian account reflects a sense of the mind as active in cognition and brings what we now call top-​down processing into the picture of cognition, allowing greater explanatory power. But in another sense, Hume has a more promising research program than does Kant. For it is one thing to posit an active synthetic activity, and another to explain how that activity is accomplished. Kant’s solution, of course, is to appeal to the “spontaneity” of thought. While that might do for transcendental philosophy, it is at least an obstacle to the naturalization of cognition and psychology in a law-​governed physical world. So, in the end, even if we take Kant’s point that a much richer account of cognitive processes is necessary than that Hume offers, we must also take Hume’s point that the account of those processes must, in the end, be causal. Wolff (1960) also argues that Hume “made the profound discovery that it was the activity of the mind, rather than the nature of its contents, which accounted for all the puzzling features of empirical knowledge” (99). Wolff credits Hume with launching Kant’s more extensive examination of the active role of the mind in mental representation. As Wilson (1997, 19) argues, however, Hume’s account remains associationist, and so the “activity” in question is not Kantian synthetic activity, but rather the law-​ governed interaction between mental events or phenomena. Wilson points out (10) that Hume would not have seen the first Critique as an advance on his own position, but rather as a retreat from a Newtonian model of explanation to the Aristotelian approach he rejected. Whether Hume would be correct in this regard is another matter.

( 118 )  Books I and III

The intimate connection between the role of the imagination, the role of habit, and the role of social convention substantially enriches early modern accounts of cognition and sets the stage for Kant’s synthetic theory of thought.

3.  THE EXTERNAL WORLD (FIRST PASS) AND CAUSATION (FIRST PASS) Hume is moving steadily to an account of the nature of our representation of the external world, and indeed to an account that is both skeptical and naturalistic. He will not develop that account in full until 1.4.1 and 1.4.2, in the two very complex sections on skepticism with regard to reason and with regard to the senses, which are closely linked to the profound discussion of necessary connection in 1.3.14. But the foundations are laid in part II and the early parts of part III in a discussion first of the idea of existence and then an initial discussion of the idea of a cause. It is to these discussions that we now turn. We will see that Hume regards each of these, like the putative idea of substance, as in the end, pseudo-​ideas, verbal habits with no real content. Once again, we must be careful in reading, remembering that Hume is discussing neither existence nor causation, but rather our ways of talking about them, focusing firmly on the implications of these façons de parler for human psychology. The real importance of these discussions will emerge when we discuss causality and skepticism with regard to the senses in subsequent chapters. Hume introduces the discussion of existence as follows: There is no impression nor idea of any kind, of which we have any consciousness or memory, that is not conceiv’d as existent; and tis evident, that from this consciousness the most perfect idea and assurance of being is deriv’d. From hence we may form a dilemma, the most clear and conclusive that can be imagin’d, viz. that since we never remember any idea or impression without attributing existence to it, the idea of existence must either be deriv’d from a distinct impression, conjoin’d with every perception or object of our thought, or must be the very same with the idea of the perception or object. [1.2.6.2, 66]

To sense something, or to think of it, Hume asserts, is to sense or to think of it as existent. This is an odd thought, as it seems that I can think of something, say the white whale, as non-​existent. Nonetheless, when we recall how Hume thinks of thought, we can make some reasonable sense of his point. A perception or an idea for Hume is always, if not straightforwardly

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 119 )

imagistic, at least quasi-​imagistic; and an image depicts its content. To depict content—​even the white whale—​is to show what it would look like were it real. There is no way to cancel reality in an image. That requires a more structured, logical model of thought, an innovation that would have to wait for Kant. Hume is wondering not whether we have a word for existence, but whether such a word has content—​whether it conveys an idea, a representation that depicts existence itself. He continues: The farthest we can go towards a conception of external objects, when suppos’d specifically different from our perceptions, is to form a relative idea of them, without pretending to comprehend the related objects. Generally speaking we do not suppose them specifically different; but only attribute to them different relations, connexions and durations. [1.2.6.9, 68]

Hume’s conclusion—​that we have no such distinct idea, since every idea represents its content as existent, and since every idea of existence is an idea of an existent—​is then not surprising, even if it is unsatisfactory. But in drawing this conclusion, Hume draws a distinction that will prove to be of the greatest importance when he gets to questions about identity over time, both in the case of external objects and in the case of personal identity—​that between numerical and specific identity. Hume claims that we cannot form an idea of external (actually existent) objects specifically (his italics) different from our ideas. Specific identity here is identity of kind, or of species. It contrasts with numerical identity, which is identity in the strict sense. Two identical twins are specifically identical to one another; each is numerically identical only to herself.9 In order to understand precisely what Hume means here, we must disambiguate the reference of the ambiguous anaphoric pronoun them in this passage. It is very commonly read as referring to external objects. But that

9. Baxter (2008, 2) correctly points out that Hume’s discussion of numerical identity presupposes his account of time. Baxter provides a very compelling account of Hume’s view of the idea of time and duration. I disagree with him, however, concerning the framing of those views:  he sees Hume as providing a metaphysical analysis of time per se; I see Hume as providing an account of our thought about time. But that disagreement is beside the present point. We agree that when Hume is worried about the nature of personal identity, he is worried about identity over time. See also Falkenstein (2006) for a careful discussion of the relationship between Hume’s account of space and time and the more general architectonic of the Treatise. In particular, Falkenstein shows that Hume’s discussion of the ideas of space and time are above all, discussions of our ideas, and so part of Hume’s psychology, and not, as Baxter suggests, an articulating of a metaphysical account of space and time themselves. (See esp. 61–​62.)

( 120 )  Books I and III

has to be wrong: If Hume were claiming that ideas and objects are even specifically identical to one another, that would be—​in virtue of saying that they are the same kind of thing—​to follow Berkeley into idealism. Not only is Hume no idealist, but, as I have argued, he is not even pursuing a metaphysical agenda here. So, let us consider an alternative reading. Note that Hume writes that he is talking about a “conception of external objects.” When he says in the next sentence that “we do not suppose them specifically different,” the antecedent of the pronoun them must be understood as those conceptions. The point Hume is after is therefore psychological, one about the relationship between our conceptions of external objects and our other perceptions: Hume is stating that our ideas of ideas and our ideas of the objects of those ideas are of the same kind. They are both representational mental states. So, for instance, if I consider my impression of my perception of my copy of the Treatise (an impression of reflection), and my impression of my copy of the Treatise, these two perceptions are specifically identical—​they are both impressions. The difference is in their content:  one takes a fleeting impression as its intentional object, and the other an enduring physical object. That difference, argues Hume, does not consist in annexing an idea of existence to one, but not the other. Instead, the difference in their content is a matter of the relations, connections, or durations we attribute to the things of which they are ideas. The relations I represent between my impression of reflection and other ideas and those I represent between my impression of the book and other ideas are very different. That is, we believe different things about the external things of which we have ideas than we do about the ideas of which we have ideas, even if they are not distinguished in appearance. We think, for instance, that the external objects are longer lasting than our mental states, more independent of us, involved in different causal relations, etc. This is the sense in which ideas of external objects are relative ideas. They are ideas of the relations into which something might stand, and not ideas that represent the objects themselves in some special way as existent. This contrasts, for instance, with Strawson’s (1989) reading of this characterization. He writes: We cannot “comprehend” external objects in any way on the terms of the theory of ideas; that is  .  .  .  we cannot in any way encompassingly take hold of their real nature in thought. We cannot form any positively contentful conception of their nature. . . . Because according to the theory of ideas (1) we can form a positively contentful conception of something only out of impressions-​and-​ideas material. . . . But (2) we have just supposed that external objects are entirely

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 121 )

different in kind (or species) from impressions and ideas. And Hume takes it that it follows immediately from (2) that (3) any impressions-​and-​ideas material we have is ex hypothesi useless. . . . One might as well try to get a congenitally blind person to form a positively contentful conception of particular colours by giving them shapes to feel. Nevertheless  .  .  .  we can still form a “relative” idea of such objects. It is a merely relative idea because we cannot in any way conceive of or descriptively represent the nature of an external object as it is in itself; we can only conceive it indirectly. We may for example conceive it as something that stands in a certain relation (the relation of cause) to our perceptions. . . . But a merely relative idea of . . . something X is not no idea of X at all . . . [I]‌t may and does refer to X, in this case. (1989, 50)

There is something right about this, and Strawson’s position in some respects is not so far from my own. Although Strawson reads Hume as a skeptical realist, and I do not, I agree with him that Hume denies that we have any idea whatsoever of the inner nature of external objects or of the inner nature of causation, and I agree with Strawson that Hume denies this for the same reasons in each case. I also agree with Strawson that in the end Hume takes the ideas that we think are of external objects to be ideas of things as standing in particular kinds of relations. But in other important respects, Strawson’s reading and mine are very different indeed. Unlike Strawson, I do not believe that Hume thinks that these ideas refer to external objects. That is precisely what Hume denies. Instead, he argues that these ideas lead us to believe in an external world, but not in virtue of having ideas of external objects as external. Instead, he argues, we have ideas of relations in which we stand to things in virtue of which we treat the objects of our ideas as external objects: relations such as independence, relative duration, etc. The distinction between believing in, or taking for granted, and having ideas of the external world is crucial for Hume in this discussion; he argues that we believe in an external world, not that we have ideas of external objects. That is why he is Pyrrhonian, not realistic, in his skepticism.10

10. One could, as Beebee (2006) notes, draw an analogy to Wright’s reading of Hume’s account of the causal relation. Wright (2000) takes Hume to argue that there are real necessary connections between events, although we can never have adequate ideas of them; I see Hume as arguing that we don’t have any idea of necessary connection, and so that we cannot even conceive what it would be for there to be such connections; instead we have another idea—​that of a determination of the mind in the case of constant conjunctions—​with which we confuse an idea that we don’t have, in virtue of a verbal convention that seems to express that other idea.

( 122 )  Books I and III

Hume is driven to the view that our idea of the external world is an idea of relations in which we stand to things instead of an idea or ideas of the things themselves in part because he has not yet shed the act-​object model of an idea, a model he shares with Berkeley. The idea, on this model, is the content of a mental act, not the act itself, and our beliefs or attitudes are in the first instance directed to the idea. As I note above, while the Kantian model of a mental act as a judgment was around the corner, Hume has not discovered it. In any case, Hume spells this out in much greater detail in his discussion of the importance of causation to drawing the distinction between the inner and the outer worlds, a discussion to which we now turn. Hume begins his discussion of causality by noting its importance in our experience of the world as external. We simply cannot begin to understand how we believe in the external world without understanding our instinct to interpret our experience as caused by objects outside of us. He argues (anticipating both Kant’s discussion in the second Analogy and Schopenhauer’s in The Fourfold Root of the Principle of Sufficient Reason) that our propensity to causal thinking lies both at the foundation of our confidence in the spatiality of the objects we encounter and in their continued existence. Each of these themes will be important in the context of the discussion of skepticism with regard to the external world, where they are treated in more detail. But they are introduced here to motivate the importance of causal thinking: We readily suppose an object may continue individually the same, tho’ several times absent from and present to the senses; and ascribe to it an identity, notwithstanding the interruption of the perception, whenever we conclude, that if

In each case there is a confusion of a psychological custom with an idea of something external; Wright, like Strawson, takes Hume to be committed instead to the reality of the putative external phenomenon, with only its inner nature lying beyond our ken. Beebee (2006) develops a similar critique. These readings patently violate the Cover Principle, and fail to appreciate the depth of Hume’s Pyrrhonism. It is worth noting that this reading may gain more aid and comfort from Hume’s treatment of causation and induction in the Enquiry Concerning the Human Understanding than from the Treatise. But I focus entirely on the Treatise here. See also Blackburn (2000), Millican (1998), and Winkler (2000) for a reading of the Treatise account in light of the Enquiry account; but neither attributes the kind of realism to Hume that Strawson and Wright do. Also see Allison (2008) for a critique of Strawson’s understanding of relative ideas. (And see Strawson 2002 for a defense of the claim that Hume is a realist about causation.) It is also useful to compare this sense of believing in—​or taking for granted, as Hume will put it in the context of his discussion of skepticism with regard to the senses—​ with Wittgenstein’s account of “hinge propositions” in On Certainty. In that case as well, we have an account not of believing that, but of believing in as prior to the consideration of the truth or falsity of declarative sentences.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 123 )

we had kept our eye or hand constantly upon it, it wou’d have convey’d an invariable and uninterrupted perception. But this conclusion beyond the impressions of our senses can be founded only on the connexion of cause and effect; nor can we otherwise have any security, that the object is not chang’d upon us, however much the new object may resemble that which was formerly present to the senses. Whenever we discover such a perfect resemblance, we consider, whether it be common in that species of objects; whether possibly or probably any cause cou’d operate in producing the change and resemblance; and according as we determine concerning these causes and effects, we form our judgment concerning the identity of the object. [1.3.2.2, 74]

Hume makes two important points in this paragraph. First, our judgment of continuity independent of perception is a causal judgment. The counterfactual supposition that we would have had an uninterrupted perception had we kept our eye or hand on the object11 suggests that we presuppose a causal link between the object and our senses, attributing spatiality to objects in virtue of positing them as effects of sensory experience as well as causal connections between the sequentially observed stages of the objects we take to exist. Second, our attribution of identity over time to objects of our experience is a form of causal reasoning—​in this case, a causal inference to the best explanation. These two respects in which the attribution of causality enters our experience are tightly intertwined, but distinct. It is one thing to reason about the causal relations between the objects we perceive and our perceptions of them; it is another to reason causally about the relations between successive stages of the objects themselves. In the first case, we are worried about whether we can infer from our experience to any conclusions about the external world. In the second case, we are asking about the relations that might confer identity—​and about the kind of identity they might confer—​ to external objects. While Hume does not distinguish these two cases carefully here, they will be treated distinctly a bit later, in the context of his treatment of skepticism with regard to the senses. There are two more methodological issues worth noting here. First, Hume is not writing about causality, but about our propensity to reason causally; second, he is carefully not ascribing to us an idea of a causal relation, but rather a propensity to think in a particular way; characterizing a cognitive process, not a content of cognition. Each of these will become important

11.  The pairing of visual and haptic perception is another anticipation of Schopenhauer.

( 124 )  Books I and III

in the next chapter when we consider the pseudo-​idea of necessary connection, as Hume himself foreshadows: Since it is not from knowledge or any scientific reasoning, that we derive the opinion of the necessity of a cause to every new production, that opinion must necessarily arise from observation and experience. The next question, then, shou’d naturally be, how experience gives rise to such a principle? But as I find it will be more convenient to sink this question in the following, Why we conclude, that such particular causes must necessarily have such particular effects, and why we form an inference from one to another? we shall make that the subject of our future inquiry. ’Twill, perhaps, be found in the end, that the same answer will serve for both questions. [1.3.3.9, 82]

Hume observes here that we have no theoretical arguments to the reality of causal necessity, although we presume this necessity in our ordinary life and in doing science. An account is therefore required of its origin, and that origin is not altogether obvious. A few pages later, Hume argues that our tendency to reason causally is grounded neither in the senses nor in memory, but in the imagination. As we have noted earlier, this location of essential cognitive functions in the imagination is one of Hume’s greatest innovations, and lays the foundation for Kant’s theory of mental activity in which the imagination functions so centrally (Beck 1978; Sellars 1978). Since it appears, that the transition from an impression present to the memory or senses to the idea of an object, which we call cause or effect, is founded on past experience, and on our remembrance of their constant conjunction, the next question is, Whether experience produces the idea by means of the understanding or of the imagination; whether we are determin’d by reason to make the transition, or by a certain association and relation of perceptions. If reason determin’d us, it would proceed upon that principle, that instances, of which we have had no experience, must resemble those, of which we have had experience, and that the course of nature continues always uniformly the same. . . . Our foregoing method of reasoning will easily convince us, that there can be no demonstrative arguments to prove, that those instances, of which we have had no experience, resemble those, of which we have had experience. We can at least conceive a change in the course of nature; which sufficiently proves, that such a change is not absolutely impossible. [1.3.6.4–​5, 88–​89]

It is plain that we have no sensory experience of a causal connection, but only of events we take to be causally connected. And Hume’s well-​known argument that there can be no demonstrative argument for the uniformity

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 125 )

of nature, and so no rational basis for induction, convinces him by elimination that the understanding cannot be the source of the idea.12 Once again, the faculty of the imagination will turn out to be essential to the account.13 Reason can never shew us the connexion of one object with another, tho’ aided by experience, and the observation of their constant conjunction in all past instances. When the mind, therefore, passes from the idea or impression of one object to the idea or belief of another, it is not determin’d by reason, but by certain principles, which associate together the ideas of these objects, and unite them in the imagination. Had ideas no more union in the fancy than objects seem to have to the understanding, we cou’d never draw any inference from causes to effects, nor repose belief in any matter of fact. The inference, therefore, depends solely on the union of ideas. [1.3.6.12, 92]

So, Hume concludes, the cognitive processes that lead us to unite objects into experienced unities over time, and that allow us to posit those objects as causes and not mere contents of our perceptual states, occur in the imagination.14 They are lawlike, governed by principles of association, and so, presumably, amenable to explanation by an articulated science of the mind, 12. It is interesting to compare this argument to Descartes’s piece-​of-​wax argument in the Meditations—​a very different argument by elimination to locate cognitive function in a particular faculty of the mind. 13. Contemporary readers are not the only ones who have read Hume as concerned with causation, per se, and as denying its reality. In response to the charge that Hume argues that there are no causes—​that this is a metaphysical argument against the reality of causes—​Hume writes in a letter to John Stewart: But allow me to tell you, that I  never asserted so absurd a Proposition as that any thing might arise without a Cause: I only maintain’d, that our Certainty of the Falsehood of that Proposition proceeded neither from Intuition nor Demonstration; but from another Source. That Caesar existed, that there is such an Island as Sicily; for these Propositions, I affirm, we have no demonstrative nor intuitive Proof. Wou’d you infer that I deny their Truth, or even their Certainty? There are many different kinds of Certainty; and some of them as satisfactory to the Mind, tho perhaps not so regular, as the demonstrative kind. (Letter 91, to John Stewart, February 1754, in Hume 1932, vol. 1, 187) It is plain here that Hume intends to be understood as discussing the nature and source of our conviction in the causal relation, and is not even interested here in whether there is indeed such a relation. We might also note both Hume’s emphasis on the social, not the individual, determinants of certainty, and the striking anticipation of Wittgenstein’s reflections on certainty in On Certainty. 14.  Laird (1932) concludes from Hume’s assignment of the role one might expect reason to take in causal explanation to the imagination to entail that Hume thinks that there can be no knowledge of causal relations, and that causal explanation is inadmissible in true science. But this is to confuse demonstrative certainty with knowledge. Hume is concerned not to deny the possibility of knowledge of causes, but to characterize that knowledge in a way that shows that it is never mathematical, but only

( 126 )  Books I and III

even if that science was not yet in hand. What is to be explained by such a science, though, are not explicit acts of inference, or perceptions of unities, but rather the customary—​in the sense now of habitual—​automatic operations of the imagination that cause us to experience a world of objects around us.15 Hume concludes: Thus tho’ causation be a philosophical relation, as implying contiguity, succession, and constant conjunction, yet ’tis only so far as it is a natural relation, and produces a union among our ideas, that we are able to reason upon it, or draw any inference from it. [1.3.6.16, 94]

Let us be clear about the distinction Hume is drawing here and its significance. Hume distinguishes two senses of the causal relation, or, more properly, two perspectives we can take on it. When we consider causality as a philosophical relation, we consider it as a cognitive reflex—​as the tendency of the mind to connect classes of events in virtue of the relations of contiguity, succession, and constant conjunction among token pairs of empirical in character. Laird gets closer to the mark—​although he fails to see how close he is to a better reading—​when he writes: Hume’s proper conclusion should therefore have been some doctrine of innate ideas, i.e. he should have held that men are equipped with an instinct of “reason” before experience, just as birds are equipped with a nest-​building instinct. And this, of course, was precisely what he denied. Yet what he said scarcely made sense. It was, in substance, that men did not “reason” about matters of fact, because, although they did “reason” about such matters, the “reason” they employed was an inexplicable natural endowment, just like an instinct. (1932, 144) Laird here misses Hume’s understanding of the role of the imagination in reasoning, his understanding of the centrality of custom, including our cognitive instincts, and the way that both of these doctrines interact with his Pyrrhonism. Hume does, after all, hold that we are equipped with cognitive instincts and that, like those of birds, they explain, but do not justify our cognitive achievements. All of this makes more sense than Laird can see. But that is because Laird does not read Book I in the context of Book II. Garrett (1997) gets this just right. And he notes correctly that it is for this reason that Hume can comfortably use inductive arguments throughout the Treatise and recommend an empirical science whose findings will be laws in which we have confidence in virtue of induction. His argument is meant to show not that we would be correct to reject induction, but that we would be incorrect to accept it on the grounds that it is rationally justified. Instead, our confidence rests entirely on custom, and needs no more secure foundation. See also Garrett’s excellent defense of a skeptical interpretation of Hume’s argument against the possibility of providing a rational foundation for induction. 15. As Wright (1983) points out, Hume’s conception of the relevant psychology was psychophysiological, drawing on hydraulic and geological metaphors for the action of the animal spirits on the brain, and the almost literal canalization of thought as the basis of the habits that undergird customary thought. As Wright also points out, Hume borrows this model directly from Malebranche.

E pi s t e m ol o g i c a l F o u n dat i o n s  

( 127 )

members of those classes. Philosophical relations, as Hume characterizes them [1.1.5.1–​2, 13–​14], are those that we use to connect various ideas to one another in the mind. Considered this way, causality is not a relation between events, but a propensity of the mind, mediated by processes in the imagination and by social practices of explanation. That relation, however, does not allow us to explain anything, or to develop scientific theories. Theories and explanations—​prominently including theories and explanations of human cognition and behavior—​ demand reference to real causal relations, not to tendencies of the mind. So we must acknowledge that when we attribute causal relations to classes of events, what we attribute—​for whatever reason, mediated by whatever processes—​is a natural relation, a relation capable of “producing a union among our ideas,” that is, having some ability to connect phenomena independent of our imaginative or rational activities. Solving this conundrum—​ coming to understand how we attribute a natural causal relation linking phenomena, or at least coming to understand better what we are doing when we engage in causal reasoning—​is the task of the complex analysis in 1.3.14, “Of the Idea of Necessary Connexion,” to which we now turn.16

16. Note that my view will be at odds with Strawson’s (1989) reading, according to which Hume never questions whether there is a real necessary connection between events in which the causal relation consists, but only doubts our ability to penetrate into its nature. Strawson argues for a sharp distinction between Hume’s metaphysics and his epistemology in this case, as well as in the context of Hume’s discussion of skepticism with regard to the senses. In each case, he argues that Hume is a metaphysical realist about the phenomena in question, but that he rejects our ability to know the relevant phenomena given our epistemological limitations. I will not discuss Strawson’s analysis in detail, as that would take me too far afield. But I can sum up my difference in approach in the following way: If one were to think that the subject matter of the Treatise is principally metaphysics and epistemology, one ought to agree with Strawson. Hume, in his own words takes for granted the reality of the external world, as well as the legitimacy of causal reasoning. And he denies that we know anything about either one. But to read the Treatise in this way, as I argued in the first chapter, is to miss the point of the text. The Treatise is neither about metaphysics nor about epistemology; it is about human nature. Hume is therefore silent about metaphysics, and even quiet about epistemology. He is writing instead about the psychology of our confidence in the reality of the external world and our causal reasoning. If one is convinced by my reading, Strawson’s exegesis is not refuted, so much as set aside.

C H A P T E R  7

w

Causality

1.  THE GENERAL ACCOUNT Hume’s account of necessary connection in 1.3.14 is certainly one of the most famous and most carefully discussed sections of the text. We will work through the argument in that section step by step, with a view to showing how it justifies each of the guiding principles that inform my reading of the treatise. That is, we will see that this section is not about causation or necessary connection, but rather about human nature—​in particular, how we confuse an idea we actually have with an idea of necessary connection that we could not possibly have.1 This discussion will also exhibit the Skeptical Inversion Principle. That is, Hume will argue that debates about causation are in fact best understood skeptically: those who assert that there is a real causal relation and those who deny that there is each presuppose that causality can be understood if and only if there is some special necessary connection; Hume will show that we could not even know what such a thing is. This conclusion, as Baier (1991) notes, this has far-​reaching consequences for Hume’s system, in virtue of his claim [1.3.14.32, 171] that “there is but one kind of necessity.” Necessity, for Hume, is the heart of normativity, and governs inference, explanation, and obligation in precisely the same way:  when we draw inferences, we are led to conclusions necessitated by premises; explanation shows us why an explanans necessitates an explanandum; and when we take ourselves to be morally obligated to perform a certain action, 1. Roth (2006) defends a similar view. ( 128 )

C au s a l i t y 

( 129 )

that action impresses itself on us as commendable, and its omission as reprehensible. In each of these cases, the activity in question is governed by norms, whether those norms are logical, epistemic, or ethical, and the norm-​governed character of these domains is grounded in the recognition of a kind of necessity. The investigation of necessary connexion in section 1.3.14, while it might occur in the context of a discussion of causation, is quite general in its import. Moreover, the argument will rely crucially on the role of custom in both senses:  first, custom understood as the ordinary operations of thought will provide a psychological explanation for our discourse about causality;2 second, custom understood as convention will be important in individuating events under descriptions suitable for causal generalization, as the kind of warrant that arises from long usage, as in the case of customary law (Baier 1991). Fourth, the discussion will be an instance of Hume’s general strategy of showing that we often believe that we have ideas that we do not, and replacing apparent pseudo-​ideas with actual ideas in order to solve philosophical problems. By showing that each of these four principles is operative in this argument, we have evidence for the fifth, namely the Principle of the Uniformity of Method.3 This discussion exhibits all of the features that we expect to find in a Humean argument in the Treatise. It can be read as a template for nearly every other major discussion in the text. Hume opens 1.3.14 by reminding us of the importance of causal reasoning in all of our cognitive activity. That is, he reminds us that even to go beyond our immediate experience and to posit the existence of objects in the external world, we must rely upon causality. Hume notes that in our ordinary causal discourse and thought we speak of the causal relation as one of necessary connection, and he undertakes in this section to examine that aspect of the apparent idea of causality. Having thus explain’d the manner, in which we reason beyond our immediate impressions, and conclude that such particular causes must have such particular effects; we must now return upon our footsteps to examine that question, which

2. See Bell (1993) for a similar observation. 3. I note that Beebee (2006) agrees that Hume’s approach to causality mirrors his approach to belief in the external world, although we disagree about what the structure of those arguments is. I agree with Kemp-​Smith (1941) that in each case Hume appeals to our natural propensities that have their roots in the customary operation of the imagination; I do not, however, agree with Beebee that this position commits one to skeptical realism of the kind attributed to Hume by such interpreters as Wright, Kail, and Strawson. See Dauer (2011) for a critique of this position.

( 130 )  Books I and III first occurr’d to us, and which we dropt in our way, viz. What is our idea of necessity, when we say that two objects are necessarily connected together. Upon this head I repeat what I have often had occasion to observe, that as we have no idea, that is not deriv’d from an impression, we must find some impression, that gives rise to this idea of necessity, if we assert we really have such an idea. [1.3.14.1, 155]

Note that Hume concludes this passage by warning us that we may not find that we have the idea whose origin we seek (and, as Dauer [2011] correctly notes, it must be our belief in necessity that mediates our causal inferences; the question is whether there is any idea that can serve as the content of that belief). That is, we may need to content ourselves not with finding the idea we believe that we have but instead with finding an actual idea—​an idea of some feature of our own psychology—​with which we confuse it.4 Hume begins the investigation by noting that when we observe regular successions of pairs of events we take to be causally related, all we observe is that succession and its regularity: In no one instance can I go any farther, nor is it possible for me to discover any third relation betwixt these objects. I therefore enlarge my view to comprehend

4. Fogelin (1985), following Kemp Smith, claims that “Hume never questions the assumption that some idea must be associated with the words ‘necessary connexion.’ He is not, that is, a conceptual skeptic concerning them. Thus any argument that makes our idea of a necessary connection either impossible or inexplicable runs counter to the main development in the text” (44). On one reading this is correct: Hume will explain what idea is in fact operative in our causal reasoning. But in another sense it is incorrect, for the idea that turns out to be actually operative in our causal reasoning is not the idea of a necessary connection between events that appears to be conveyed by those words. Hume is thus what Fogelin calls a “conceptual skeptic” about that idea, that is, he thinks that we have no such concept, and offers us a substitute. While, as both Kemp Smith and Fogelin correctly insist, Hume must distinguish mere regular sequence from causation, he does so in terms not of our awareness of an occult connection between events in causal sequences and its absence in noncausal sequences, but rather of the embedding of regularities within other regularities and the psychology and pragmatics of explanation, in short, in terms of custom. (But see Wilson 1997 for an alternative, more psychological, way of drawing the distinction between accidental and nomic regularities.) Hume’s argument for his account of the nature and origin of our idea of causality does not, as Fogelin insists, rely on his skepticism regarding induction. It is a straightforwardly naturalistic argument, and a causal account. Fogelin sides with Reid in arguing that Hume cannot distinguish regularity from causation. But the very fact that for Hume causal attributions are founded in inferential practices and not the other way around (“perhaps ’twill appear in the end, that the necessary connexion depends on the inference, instead of the inference depending on the necessary connexion” [1.3.6.3, 88])—​an expression that Fogelin also misreads as referring to induction—​allows Hume to distinguish causation from regularity on pragmatic grounds.

C au s a l i t y 

( 131 )

several instances; where I  find like objects always existing in like relations of contiguity and succession. At first sight this seems to serve but little to my purpose. (1.3.14.1, 155)

Repetition, he then points out, is just repetition. Seeing the same thing twice doesn’t give us a new idea of that thing but only produces two equivalent impressions that can only lead to the same idea. But, Hume notes, there is in fact a new impression that arises, albeit not an impression of the thing observed, but rather an impression of reflection directed upon the cognitive consequence of the repeated observation: The reflection on several instances only repeats the same objects; and therefore can never give rise to a new idea. But upon farther enquiry I find, that the repetition is not in every particular the same, but produces a new impression, and by that means the idea, which I at present examine. For after a frequent repetition, I find, that upon the appearance of one of the objects, the mind is determin’d by custom to consider its usual attendant, and to consider it in a stronger light upon account of its relation to the first object. ’Tis this impression, then, or determination, which affords me the idea of necessity. [1.3.14.1, 155–​56]

At this point, Hume has not offered much of an argument. He has only observed that custom—​in this case cognitive habit—​leads us to associate the idea of one of a pair of such events with the other. It is only the fact that Hume can imagine no other consequence of repetition that leads him to suspect that reflection on this cognitive habit is what leads us in turn to the idea that we call necessity. Note that this is all very preliminary in the section, and we should not attach much importance to the precise formulation. Hume himself notes that it forms a preamble to the detailed investigation and not a principal argument itself. He introduces the argument to follow in this passage: There is no question, which on account of its importance, as well as difficulty, has caus’d more disputes both among antient and modern philosophers, than this considering the efficacy of causes, or that quality which makes them be follow’d by their effects. But before they enter’d upon these disputes, methinks it wou’d not have been improper to have examin’d what idea we have of that efficacy, which is the subject of the controversy. This is what I find principally wanting in their reasonings, and what I shall here endeavor to supply. [1.3.14.3, 156]

Note the last remark. It is here that Hume reminds us of his strategy. Other philosophers believe that when they investigate necessary connections, they

( 132 )  Books I and III

know what they’re talking about; Hume suspects that they don’t. In particular, he suspects that we don’t even have an idea of necessary connection. If the reading I propose is correct, every member of the family of “skeptical realist positions,” such as the positions of Kail (2001), Strawson (1989), and Wright (1983), gets Hume wrong on causality. On those positions, as Beebee puts it correctly, “Hume thinks that causation consists in. . . ‘real powers.’ Real powers are thoroughly mind-​independent and are more than mere regular association. They are ‘secret,’ and Hume’s definitions are ‘imperfect,’ because we cannot grasp their nature, and we cannot grasp their nature because that nature is not revealed to us in sensory experience. But we believe in it and refer to it nonetheless” (2006, 173). Hume, however, argues that we don’t even have the requisite idea, although we do engage in explanatory practices that use words that appear to express that idea. Hume’s skepticism, therefore—​while every bit as constructive as that attributed to him by those who read him as a skeptical realist—​is constructive because it is Pyrrhonian, not because it is realistic.5 5. Loeb argues that Hume cannot be a skeptic about causation and induction partly on the grounds of Hume’s argumentative practices in support of his project. In the initial section of the work, Hume relies on inductive evidence to establish his principle that every simple idea exactly resembles a preceding impression [4–​5]. Just two sections following the main argument about induction, Hume appeals to “experience” and “experiments” to confirm—​indeed, “to prove”—​his associationist account of belief. (2006, 323) Loeb offers a number of other arguments as well for his conclusion that “it cannot be correct” that Hume is a skeptic, but those are answered in various ways elsewhere in this study, and in any case, this is the most persuasive of the lot. Nonetheless, we should not be persuaded. Indeed, the fact that Hume uses inductive arguments and causal reasoning is utterly irrelevant to the probity of a Pyrrhonian interpretation. If our customary epistemic practices include this kind of reasoning—​as they do, and as Hume in fact argues that they do—​there is no reason not to use them, and every reason to use them. Hume’s skepticism consists in his argument that there is no possibility of justifying these practices independent of those customs, that their normative force derives entirely from those customs, and that many of the discursive practices involved in these kinds of reasoning involve the use of terms to which no clear ideas correspond. None of this, for a consistent Pyrrhonian, argues against the use of these argumentative practices. Loeb’s argument is hence a straightforward non sequitur. Similarly, when Loeb argues that Hume is an epistemological externalist (which he surely is) as opposed to a skeptic (which he also surely is) on the grounds that “he labels inductive inference ‘just’ in the course of his main argument about induction and maintains his commitment to inductive inference” (333), he commits the same error of presupposing that a naturalistic, skeptical understanding of our justification of inductive reasoning must be the same as an antiskeptical demonstration that induction is independently justified (or “just”). Similar problems arise for Millican’s (1998) critique of Garrett (1997). Millican says that “on Garrett’s interpretation [according to which Hume is entitled to inductive

C au s a l i t y 

( 133 )

Hume therefore embarks upon a series of arguments designed to show us both that we have no such idea and to reconstruct the idea that we do have that guides our causal reasoning in the absence of any proper idea of causation itself. We will also see that this pattern of analysis and argument is followed in his discussions of skepticism with regard to reason and with regard to the external world; this identity of form of argument not only confirms the utility of the Principle of the Uniformity of Method in reading Hume, but also provides additional evidence for my interpretation of each of these arguments. We now turn to the first argument regarding causation.

2.  THE FIRST ARGUMENT The first argument is a bit of a warm-​up for the principal argument. It has a compact, clear structure that I represent in a series of numbered premises.6 Hume begins by reminding us that all ideas are derived from impressions and never originate in reason itself: (1) First, that reason alone can never give rise to any original idea. [1.3.14.5, 157]

From this it follows that reason cannot possibly be the source of the idea that there is a necessary connection between events: (2) Secondly, that reason, as distinguish’d from experience, can never make us conclude, that a cause for productive quality is absolutely requisite to every beginning of existence. [1.3.14.5, 157]

Hume then concludes that, since reason and experience are the only possible sources of ideas, any idea we have that is relevant to that of causation

inference and normative epistemic claims] I just do not see how Hume has any right to help himself to all of these ‘cannots,’ ‘can onlys,’ and ‘musts’ [in his discussion of causation and induction in the Enquiry Concerning the Human Understanding]” (152). As I have been arguing, however, there is no problem for Hume in this regard: he is arguing for a particular view of the status of inductive and normative discourse, not arguing for their elimination. 6. At several points in the representation of this argument and that of the principal argument, the numbering of the premises will not follow the order of their appearance in Hume’s text. I am numbering them in the order in which they appear logically in the arguments, in order to make their structures clearer. So, for instance, in this argument (6) will appear before (5) and (9) before (8), reflecting the fact that although (6) is stated first, it is inferred from (5); (9) depends on (8) although it precedes it in the exposition.

( 134 )  Books I and III

must come from our senses or from an impression of reflection directed upon an idea that originally derives from the senses: (3) I shall only infer from them, that since reason can never give rise to the idea of efficacy, that idea must be deriv’d from experience, and from some particular instances of its efficacy, which make their passage into the mind by the common channels of sensation and reflection. [1.3.14.6, 157]

It follows from this that if we have any idea of causal power or necessary connection, there must be some perceptible instance of that quality to which we can point as the origin of that idea: (4) If we pretend, therefore, to have any just idea of this efficacy, we must produce some instance, wherein the efficacy is plainly discoverable to the mind, and its operations obvious to our consciousness and sensation. [1.3.14.6, 157–​58]

So, since Hume has already refuted the doctrine of innate ideas, if there is no sensible property to which we can point as the origin of necessary connection, we must acknowledge that we have no such idea: (6) By the refusal of this, we acknowledge, that the idea is impossible and imaginary; since (5) the principle of innate ideas, which alone can save us from this dilemma, has already been refuted, or is now almost universally rejected in the learned world. [1.3.14.5, 158]

Now, Hume points out, there is a problem. Since the idea of necessary connection is an idea of a relation that is meant to connect instances of causes and effects that are regularly associated, this idea cannot come from any impression of any single pair of events; no single pair could allow us to discriminate between causation and mere coincidence: (7) Upon the whole, we may conclude, that ’tis impossible in any one instance to shew the principle, in which the force and agency of a cause is plac’d; and that the most refin’d and most vulgar understandings are equally at a loss in this particular. [1.3.14.7, 158–​59]

Moreover, since any such idea would be the idea of a relation that would make it impossible for the effect not to follow the cause, and since for any causally connected pair of events we can imagine, it is possible for the cause to occur without the effect, and since certainly no observation of any

C au s a l i t y 

( 135 )

individual pair of cause-​and-​effect events could give rise to an idea with this kind of necessity, Hume concludes that there is in fact no idea of necessary connection:7 (9) Now nothing is more evident, than that the human mind cannot form such an idea of two objects, as to conceive any connexion betwixt them, or comprehend distinctly that power or efficacy, by which they are united. (8)  Such a connexion wou’d amount to a demonstration, and wou’d imply the absolute impossibility for the one object not to follow, or to be conceiv’d not to follow upon the other: which kind of connexion has already been rejected in all cases. [1.3.14.12, 161–​62]

This argument takes us only halfway there, though. For while Hume takes himself to have shown that we can have no idea of necessary connection, he has not yet finished the job. He has not shown us what idea we do have that we take to be an idea of necessary connection; nor has he demonstrated its origin; nor has he shown how it is that we managed to confuse that idea with the idea of necessary connection. That is the task of the principal argument of section 14 to which we now turn.

3.  THE PRINCIPAL ARGUMENT This is Hume’s serious attempt to provide an exhaustive understanding of the idea that we do have: the idea that we confuse with an idea of necessary connection that we can’t possibly have; the distinctness of the idea we do have from that we don’t; and finally, its origin. In reading this argument with care, we see all of the characteristics of Hume’s philosophical method we have been discussing so far. He begins where the preliminary argument left off:  no observation of any pair of events supposed to be cause and

7. Henry Home in his Essays on the Principles of Morality and Natural Religion (Home 2000)  replies to Hume on this issue. He concedes that neither external perception nor reason can give rise to an impression of causality, but argues nonetheless that we do in fact have an impression of the connexion between causes and effects derived from introspection, viz., our experience of the connexion between our intentions and our actions mediated by the will. While this is a nice idea, it should be evident (and Home concedes this point) that it could never give rise to the necessity requisite to an idea of causality, as we would have to have an impression not only of willing, but of necessitating action. But there is a further problem with Home’s reply: it patently begs the question, for we would then require an impression of the connection between intending and willing, and between willing and acting, and so on ad infinitum. See also Dauer (2011).

( 136 )  Books I and III

effect can ever give us any perceptual access to any connection between them in virtue of which they are so related.8 (1) Suppose two objects to be presented to us, of which the one is the cause and the other the effect; ’tis plain, that from the simple consideration of one, or both these objects we never shall perceive the tie, by which they are united, or be able certainly to pronounce, that there is a connexion betwixt them. [1.3.14.14, 162]

That is, single instances of pairs of events can never suffice as the origin of the idea of causation. Hume draws this conclusion explicitly: (2) ’Tis not, therefore, from any one instance, that we arrive at the idea of cause and effect, of a necessary connexion of power, of force, of energy, and of efficacy. [1.3.14.14, 162]

It follows directly that if we did not both represent distinct pairs of such episodes as of the same kind, and see repeated instances of pairs we recognize of the same kind, there could be no source of this idea. Hume puts it this way: (3) Did we never see any but particular conjunctions of objects, entirely different from each other, we shou’d never be able to form any such ideas. [1.3.14.14, 162]

Now, Hume notes, our experience is not like that. We do not simply see single instances of pairs of events. We see multiple pairings of events of the type we denominate cause with those of the kind we denominate effect in each case where we attribute a connection to events of these types. It is when we observe this constant conjunction that we infer the occurrence of one from that of the other. Note that here the account is straightforwardly inductive, and is an account of a cognitive propensity or custom and nothing more. Nor is there the claim that the repetition provides evidence for a connection; only that it causes us to draw such inferences. (There is no begging of the question or 8. We should also note that Hume is a bit sloppy in his language here, often talking about objects when he intends events. I shall be a bit more regimented in my comments, using necessary connection as a relation between events, and talking about the causal powers of objects. We should also note that this argument involves a number of empirical premises of the kind that Hume thinks we should derive through experiment, but which he takes as obvious data of introspection, violating his own methodological canon, something he does frequently in the Treatise.

C au s a l i t y 

( 137 )

“Humean circle” here, either: in appealing to causation at this stage of the argument, the explanans is not the causal relation, but rather our propensity to speak of one): (4) But again; suppose we observed several instances, in which the same objects are always conjoin’d together, we immediately conceive a connexion betwixt them, and begin to draw an inference from one to another. [1.3.14.15, 163]

Hume infers from this that the multiplicity of resembling instances is the cause of the idea we in fact have: (5) This multiplicity of resembling instances, therefore, constitutes the very essence of power or connexion, and is the source, from which the idea of it arises. (1.3.14.15, 163)

Here we must tread with interpretative care. Hume is speaking loosely in this passage. When he says that “the multiplicity . . . constitutes the very essence of power and connexion,” he means not that power and connection are real phenomena that reduce to multiplicity; nor does he mean that we have an idea of power or connection and that that idea is the idea of multiplicity. He is entitled to neither of these claims, and he means to make neither. Hume’s point is made more clearly in the final clause, connected by an and that must be read as a viz. These multiple resembling instances are the source of whatever idea we have with which we confuse the idea of power or connection. We can confirm this reading by attention to the next sentence, which would be flatly inconsistent with a literal reading of the first half of the previous sentence. Hume points out here that that idea—​the one we actually have—​cannot be a new idea of something external, for all we get by repetition is repetition of the same thing. This claim, as Hume notes, follows from the principle, functioning as the sixth premise, that all ideas have their origins in impressions: (7) The repetition of perfectly similar instances can never alone give rise to an original idea, different from what is to be found in any particular instance, as has been observ’d, and as evidently follows from our fundamental principle, (6) that all ideas are copy’d from impressions. [1.3.14.15, 163]

Now, Hume argues, since whatever idea we have is a new one, not simply the idea of a pair of events, or even of the repetition of pairs of similar events in the same order, it cannot arise simply from the repetition, but

( 138 )  Books I and III

must be mediated by something produced by that repetition (and distinct from it). Here is that short subargument: Since therefore (8)  the idea of power is a new original idea, not to be found in any one instance, and which yet arises from the repetition of several instances, it follows, that (9) the repetition alone has not that effect, but (10) must either discover or produce something new, which is the source of that idea. [1.3.14.15, 163]

So, Hume concludes, if we want to understand the nature and origin of the idea we confuse with that of power or necessary connection, we must ascertain the effect of that repetition. Again, the reasoning here is straightforwardly causal, but not on that account question-​begging, given the presuppositions and purposes of this argument—​once again, an argument intended not to provide an analysis of causation, but an account of the source and content of the ideas we deploy in causal reasoning: (11) Every enlargement, therefore, (such as the idea of power or connexion) which arises from the multiplicity of similar instances, is copy’d from some effects of the multiplicity, and will be perfectly understood by understanding these effects. [1.3.14.15, 163]

First, Hume notes, the effect of the repetition is not observed externally. The events themselves, and their joint occurrence, are not affected by the repetition: (13) ’Tis certain that this repetition of similar objects in similar situations produces nothing new either in these objects, or in any external body. [1.3.14.15, 164]

This is because they are independent of one another. There may be a link between cause and effect, but there is no reason to believe that there is any such link between successive pairs of cause and effect of the same kind: For (12) ’twill readily be allow’d, that the several instances we have of the conjunction of resembling causes and effects are in themselves entirely independent, and that the communication of motion, which I see result at present from the shock of two billiard-​balls, is totally distinct from that which I saw result from such an impulse a twelve-​month ago. [1.3.14.15, 164]

Nonetheless, this repetition has some effect that leads us to the idea we are examining. Once again, Hume speaks loosely of the “ideas of necessity, of

C au s a l i t y 

( 139 )

power, and efficacy,” using these terms metonymically to denote whatever idea we do in fact have: (14) But ’tis from this resemblance, that the ideas of necessity, of power, and of efficacy, are deriv’d. [1.3.14.18, 164]

And we can confirm that metonymic reading once again by attention to the next step in the argument when Hume asserts that whatever idea we have is not an idea of anything connected to the objects themselves; it has to be on the side of the subject: (15) These ideas, therefore, represent not any thing, that does or can belong to the objects, which are constantly conjoin’d. [1.3.14.18, 164]

So, Hume concludes, from (15) and from (16), which immediately follows, the effect of the repetition of similar influences is not on the objects, but on the subject. We undergo a psychological change consequent upon observing the repetition of similar instances of pairs of events, and the source of our conviction that we have an idea of necessity lies in that psychological effect: (17) Tho’ the several resembling instances, which give rise to the idea of power, have no influence on each other, and can never produce any new quality in the object, which can be the model of that idea, yet the observation of this resemblance produces a new impression in the mind, which is its real model. [1.3.14.19, 164–​65]

The effect in question is a “determination of the mind” to expect an effect consequent upon the cause, to presume the occurrence of the cause when the effect has been observed. That is nothing but mere cognitive habit, or custom, and it is the sole effect of the observation of the series of resembling instances. (16) For after we have observ’d the resemblance in a sufficient number of instances, we immediately feel a determination of the mind to pass from one object to its usual attendant, and to conceive it in a stronger light upon account of that relation. [1.3.14.19, 165] (18) This determination is the only effect of the resemblance; and therefore must be the same with power or efficacy, whose idea is deriv’d from the resemblance. [1.3.14.19, 165]

( 140 )  Books I and III

It is important to note, as Baier (2008, 2) points out, that it is the feeling of the determination of the mind, not the determination itself, that gives rise to the idea we confuse with that of necessary connection. We do not, when we assert that one event causes another, assert that the relation between the two events mirrors the relation between two cognitive events of which we are antecedently aware; instead, we assert that we have the same feeling of expectation of one event when observing the other that we experience when we observe the relevant cognitive sequences. This distinction is important. Hume does not assert—​and in fact explicitly denies—​that we notice an introspectable occult connection between cognitive events that is invisible when we observe external events, and then extrapolate a causal connection we observe in our mind to the material world (as Beebee [2006], for instance, would have it). Instead, he asserts that we experience the same feeling when experiencing external sequences that we experience when we attend to our own cognitive states, and confuse that feeling with an idea of an occult connection, an idea we could never have. After all, we no more observe such a connection in our inner life than we do in the external world. The idea of reflection that emerges from the impression of reflection that in turn arises from noticing that determination is the idea we confuse with necessity. The several instances of resembling conjunctions lead us into the notion of power and necessity. . . . (19) Necessity, then, is the effect of this observation, and is nothing but an internal impression of the mind, or determination to carry our thoughts from one object to another. Without considering it in this view, we can never arrive at the most distant notion of it, or be able to attribute it either to external or internal objects, to spirit or body, to causes or effects. [1.3.14.19, 165] The necessary connexion betwixt causes and effects is the foundation of our inference from one to the other. The foundation of our inference is the transition arising from the accustom’d union. These are, therefore, the same. [1.3.14.20, 165]

Here is also another instance of Hume not being nearly as precise as he should be in his language. The last remark might lead one to think that he takes himself to have discovered an idea of necessary connection, and to have pronounced it to be a transition between ideas. But if we read this claim in the context of the preceding sentences we see that that cannot be what he intends. He means to say that the only foundation for causal inference is

C au s a l i t y 

( 141 )

the impression of customary union. Therefore, the idea that underlies our custom of causal inference—​the one we mistakenly believe to be an idea of necessary connection—​is in fact the idea of a customary transition from one idea to another. While we might have thought that we could explain causal inference only by finding an idea of causality that grounds it, we in fact find that our idea of causality is only that of customary causal inference. Here, for convenience, is a summary of this long argument: 1. No single instance of a pair of events can give us any impression of a causal connection between them (psychological observation). 2. So, we do not get the idea of necessary connection from any experience of a single pair of events (from point 1). 3. So, if we did not represent event pairs as specifically identical we would never have the idea of causation (from point 2). 4. But when we see many events pairs that are specifically identical we form the idea that the event types they comprise are causally related (psychological observation). 5. It is therefore the observation of the multiplicity that is the origin of the idea we seek (from points 2, 3, and 4). 6. All ideas are copied from impressions (established early in Book I). 7. Repetition cannot give rise to a new impression (from point 6). 8. The idea of (or that we confuse with) causal power is a new idea (from point 1). 9. Repetition alone cannot give any new ideas (from points 6 and 7). 10. Repetition must produce a new impression that is the source of the idea (from points 6 and 9). 11. The idea we seek is the effect of an impression caused by repetition (from points 6, 8, and 10) 12. Each of the multiple specifically identical instances of event pairs is independent of the others (premise). 13. So, the repetition doesn’t change the nature of any of the events or our ideas of them (from points 6, 7, and 13). 14. But the similarity of the events is causally relevant (from points 4, 7, and 9). 15. The idea we seek hence does not represent anything we observe in the objects (from points 6, 11, and 13). 16. The observation of the repetition causes us to feel a “determination of the mind” to pass from one event to the other (psychological observation). 17. We have an impression of reflection of this determination (psychological observation).

( 142 )  Books I and III

18. This is the only effect of that repetition (from points 13–​16). 19. Therefore, the idea we confuse with necessity is that derived from the impression of our determination to pass from one idea to the other. Let us reflect for a moment on what this argument is intended to show before moving to Hume’s restatement in summary form, a restatement that will confirm this reading. Hume has not analyzed causation in this context; he has neither argued that there is a causal connection nor that there is not. Instead, he has, following Pyrrhonian skeptical practice, suspended that debate; he has dissolved, rather than solved, the problem of understanding the idea of necessary connection. Moreover, he has not suspended it by shrugging his shoulders and declaring that each side is equally plausible; he has not done so by compromise—​by defending the reality of some weaker relation between events midway between causality and complete independence. Instead, he has suspended the debate by showing that it is predicated on a false thesis that both parties endorse:  that our causal discourse—​our customs regarding causal explanation, in the conventional sense of custom—​ is grounded in an idea of the necessary connection between events. The causal realist takes these customs to be justified because of our access to causal links that gives rise to these ideas; the causal antirealist argues that since we have no such access, our causal discourse is nonsense. Hume’s skeptical solution consists in rejecting that underlying biconditional assumption and reversing the order of explanation: our customs regarding causal reasoning are not grounded in the idea of necessary connection; instead, our conviction that we have such an idea arises from our customs (both in the conventional and the psychological sense of that term).9 Hume’s topic, therefore, in this discussion, is not causation, not necessary connection, but rather the psychology of our explanatory 9. Beebee writes, “As a matter of brute psychological fact, our minds are endowed with some sort of mechanism . . . by means of which one idea ‘attracts’ another, and this mechanism somehow or other tracks our causal beliefs” (2006, 16). This is close to correct: Hume is explaining our causal reasoning and our idea of causality by appeal to “brute psychological fact,” and he does understand that fact as a kind of associative mechanism. But it does not “somehow or other [track] our causal beliefs.” Instead, as we have seen, that mechanism gives rise to the idea we confuse with an idea of necessary connection among events, and so is the origin of our causal beliefs. So Beebee is correct when in the next paragraph she writes, “The mechanism simultaneously generates both the idea of (in fact, belief in) the effect and the judgment that the first event will cause the second” (17). But it is not only the judgment and the idea of the second event that are at issue, but the custom of taking the events themselves to be connected.

C au s a l i t y 

( 143 )

practices, and the account is thoroughly Pyrrhonian.10 This is the sense in which Hume, as Baier puts it, seems to “be spreading something human and animate onto the inanimate world, talking in what is strictly a metaphor, personifying natural regularities into habits.” (2008, 2). This is not an idealist claim that causal relations are all psychological; it is instead the claim that causal inference is grounded in custom.11

4.  THE SUMMARY ARGUMENT Hume immediately provides a concise summary of this complex argument: ( 1) The idea of necessity arises from some impression. (2) There is no impression convey’d by our senses, which can give rise to that idea. (3) It must, therefore, be deriv’d from some internal impression, or impression of reflection. (4) There is no internal impression, which has any relation to the present business, but that propensity, which custom produces, to pass from an object to the idea of its usual attendant. (5) This therefore is the essence of necessity [or, more precisely, of the idea we confuse with one of necessity]. [1.3.14.21, 165]12 Once again, we must be more careful in reading than Hume, alas, is in writing. The statement of this summary argument is sloppy. Hume begins

Note also that Hume cannot be asserting that we have an immediate impression of a causal connection between our ideas: that would be for his position to collapse into Locke’s, with the assumption that there is some kind of epistemic asymmetry between our inability to observe causal connections in the external world, and our immediate access to those connections in our mind, an approach Hume clearly rejects. 10. Coventry (2006) puts this point in a slightly different way, referring to Hume’s account of causation as quasi-​realist, following Blackburn’s usage of that term in ethical theory. That is, she argues, while Hume denies that necessary connection has any mind-​independent reality, and hence that we can talk about causality independent of human explanatory practices, we can talk sensibly about those practices, and so about when it is correct or incorrect to assert causal connections. Causal inference and causal explanation are hence preserved as real phenomena, but the causal relations to which they refer are regarded as mere fictions. 11.  The affinities of Hume’s account of causation to some Buddhist account are striking. For some discussion of the relationship between his account and that of Nāgārjuna, see Garfield (1994). 12.  Hume presents a similar summary of the Enquiry version of this argument in Enquiry Concerning the Human Understanding (EHU 7.28; SBN 75).

( 144 )  Books I and III

with a claim about the idea of necessity, and ends with a conclusion that looks like a statement about the essence of necessity, something of which he has just argued we have no idea at all!13 Let us clean things up a little bit. Hume begins by stating that whatever idea we confuse with necessity must come from some impression. In (2) and (3) he makes the point that the impression that is the source of this idea must be an impression of reflection, since there is no possible external object of which it can be an impression. Statement (4) is again sloppy. Hume means to say that the relevant impression is an impression of the propensity, produced by custom (in the psychological sense of that term) to associate the idea of an event with the idea of its usual concomitant event. The conclusion cannot be a statement about the essence of necessity then, but rather about the idea we take to be an idea of a necessary connection, which is the idea derived from this impression of reflection—​the impression of our psychological disposition that arose from the observation of those repetitions.14 Hume sums this up in very much these terms, returning to the clear assertion that we indeed have no idea of necessity, and putting things in clearly psychological terms: Upon the whole, necessity is something, that exists in the mind, not in objects; nor is it possible for us ever to form the most distant idea of it, consider’d as a quality in bodies. Either we have no idea of necessity, or necessity is nothing but that determination of the thought to pass from causes to effects and from effects to causes, according to their experienc’d union. [1.3.14.21, 165–​66]

The object of that idea, that is, is not something external, but something internal that we confuse with an external relation. The Humean method of demonstrating that we often take ourselves to have ideas we do not have, and take our verbal conventions to reflect very different ideas, is

13. Allison (2008), for instance, on these grounds, takes Hume to be analyzing causal power, rather than our idea thereof. 14. Pears (1990) misses this point when he argues that Hume takes us to have a perfectly good idea of necessity, one we use in causal inferences, grounded in an internal impression of necessity. Pears takes the impression of the determination of the mind to pass from one idea to another to be, according to Hume, an impression of necessity itself. This is of a piece with his understanding of Hume’s project as one of the investigation of the warrant of our beliefs, as opposed to their origin and dynamics. This issue arises again (82ff.) when Pears takes the question that Hume is trying to answer in his account of causal reasoning to be whether “audacious” inferences—​that is, inferences that take us beyond experience—​are warranted or not, as opposed to what causes us to engage in such inferential behavior.

C au s a l i t y 

( 145 )

hence amply in evidence. Hume now generalizes this point about necessity, emphasizing that it is not a thesis about the nature of causality itself, but rather about the structure of human psychology:15 Thus as the necessity, which makes two times two equal to four, or three angles of a triangle equal to two right ones, lies only in the act of the understanding, by which we consider and compare these ideas; in like manner the necessity or power, which unites causes and effects, lies in the determination of the mind to pass from the one to the other. The efficacy or energy of causes is neither plac’d in the causes themselves, nor in the deity, nor the concurrence of these two principles; but belongs entirely to the soul, which considers the union of two or more objects in all past instances. ’Tis here that the real power of causes is plac’d, along with their connexion and necessity. [1.3.14.19, 166]

As Beck (1978) and Allison (2008) each note, this position, locating necessity in the mind rather than in the world, lays the foundation for Kant’s location of causation as a pure concept of the understanding. It is also important to see that Hume has not drawn any metaphysical conclusion regarding causality or necessary connection. In particular, he has neither asserted nor denied that there are real causal connections in the world (and indeed, as we have seen, many commentators have ascribed each of these positions to him). Not only is that not his point; by his own lights he cannot do so, since he has argued that there is no idea that corresponds to cause or to necessary connection. Either of these dogmatic assertions would therefore be nonsense. Like a good skeptic he has suspended the debate by denying what both causal realists and causal nihilists accept, viz., that causal discourse can only be warranted by appeal to an independent causal relation to which we have epistemic access. Nor, however, does Hume argue that causal explanation and inference are either impossible or unwarranted (another position frequently imputed to him). Instead, he locates that warrant where he locates all warrant—​ in custom. It is our cognitive habits and our social practices that warrant causal attribution and causal inference, and the success and long usage of these customs, nothing more. So, the conclusion of his argument with respect to causal discourse is semantic, not metaphysical: the meaning of

15. Here Hume anticipates Schopenhauer’s unification of different kinds of necessity under a single principle of sufficient reason, a unity that on Schopenhauer’s view reflects not the nature of things as they are independent of the mind, but rather the structure of the human understanding and the role of causal reasoning in perception and cognition.

( 146 )  Books I and III

cause is given through its use in common and scientific practice, not by its denoting any necessary connection or occult power. Causal attributions, on Hume’s view, reflect rather than ground explanatory practices. Allison (2008), following Pears (1990), offers a very useful observation in this context, one that further demonstrates the importance of the Principle of Uniformity of Method. Hume’s commitment to the location of necessity in the mind, rather than in the objects or events to which we attribute necessary connection, is isomorphic to his assertion in the case of ethics that when we judge an act or a character trait as virtuous or vicious, the basis of that judgment is a moral sentiment in the mind (as we will see in ­chapter 9). In each case, it is appropriate usage to attribute the property in question—​causation in the present case, or virtue or vice in the moral case—​to the object we contemplate, but, in each case, the origin of the relevant idea is an impression of reflection. This is an important isomorphism. Nonetheless, as Allison argues, there a significant difference between the cases that must also be borne in mind: while the origin of the idea of virtue or vice is internal, the virtuous or vicious trait itself is actually in the object we contemplate. This is unlike the case of causation, in which what we notice is purely mental, although we ascribe it to the external objects. This difference, Allison points out, flows from another feature of Hume’s system to which Pears also draws our attention: there is a contextualism in Hume’s account of these impressions of reflection. The determination of the mind that gives rise to the impression that grounds the idea of causality is present only when we reason about regularities; similarly, the sentiments that give rise to moral judgments (as, again, we will see in ­chapter 9) are sentiments particular to the case of moral assessment, and are different, for instance, from the sentiments of approbation or disgust that might arise in aesthetic contemplation (Allison 2008).

5.  OBJECTIONS ANSWERED Hume concludes this section by facing the obvious objections to this position. It seems patently obvious to us that causation is a real relation between events, and that we have an idea adequate to it that derives from observation of causal connections. Hume’s contention, the first imaginary objector replies, flies in the face of this obvious fact. Hume replies not that there is no such relation—​as he might if the argument were intended to be one against the reality of causality in nature—​but rather that our idea of necessary connection does not derive from the observation of any external

C au s a l i t y 

( 147 )

connections, reflecting the fact that the real topic of this section is the origin of the idea of causation, not an inquiry into its object. Instead, he points out, the intuition that lies behind this objection is “easily accounted for” by the fact that the mind has a “propensity to spread itself in external objects” [1.3.14.24, 167].16 That is, we easily confuse our inner experience with the external world.17 The examples Hume chooses to make this point are classic secondary qualities—​sounds and smells—​which, he supposes, exist only in us as the consequences of interactions with external objects, and not in the objects, and one might think that the point of these examples is the fact that they are secondary. But this is not correct. Note that Hume does not include another classic example of a secondary quality—​color. The point of these examples is instead that we assign spatial locations to properties that we know are not in any sense spatial. (“I hear a sound over there”; “There is a bad smell in that room.”)18 His point is that we have a well-​documented tendency to assign external spatiality to phenomena we know on reflection

16.  Allison (2008) points out that Hume borrows this metaphor of the mind spreading itself out over reality from Malebranche, who argues that when we attribute secondary qualities to objects, the mind spreads out upon the objects that which is properly only in the mind itself. 17. For this reason we must reject Beebee’s reading of Hume on necessary connection, according to which the impression of the determination of the mind (to which she refers as the NC-​impression) detects a real causal process in the brain, on the basis of which we infer a like causal process in external events. She writes, “Those causal judgments that are based on what is immediately present to the senses are formed not by looking inward and examining our own reactions to the scenes before our eyes; rather, they are formed merely by attending to those scenes themselves. When we judge how things causally are—​whatever Hume thinks that amounts to—​our judgments are based on how they seem” (2006, 89). On Beebee’s view, we have an actual idea of necessary connection between external events, and we get that idea from an impression of reflection of an actual causal connection between events in our minds. The “spreading” of the mind over the world to which Hume refers, on this view, is merely argument from analogy. But if that were so, Hume would never regard his position as so incredible or astonishing. We argue from analogy all the time (as Hume himself notes in EHU 4.20 [SBN 36]: “All our reasonings concerning matter of fact are founded on a species of Analogy” [104]). Instead, Hume argues clearly that we have no idea of necessary connection between events, only of a determination of the mind to pass from one event to another, an idea of which we confuse with one of necessary connection. That is surprising, and is an instance of the mind projecting itself outward; this interpretation, following the principles of uniform interpretation and of the nominalist commitment to pseudo-​ideas, tracks Hume’s own words much more closely. 18. Compare Hume’s remark in the context of his critique of the idea that we must posit a soul as the locus of our ideas: “An object may exist, and yet be no where: and I assert, that this is not only possible, but that the greatest part of beings do and must exist after this manner” [1.4.5.10 235]. Hume grants that even though nonspatial existence is obviously possible, we have a tendency to attribute spatial location (or metaphorical spatial location in the case of the soul) to things we take to exist.

( 148 )  Books I and III

not to be external or spatially located.19 Necessary connections, he suggests, should be thought of in the same way: ideas in us that are caused by the observation of external objects, but which are not themselves observed in those objects:20 This contrary biass is easily accounted for. ’Tis common observation, that the mind has a great propensity to spread itself on external objects, and to conjoin with them any internal impressions, which they occasion, and which always make their appearance at the same time that these objects discover themselves to the senses. Thus as certain sounds and smells are always found to attend certain visible objects, we naturally imagine a conjunction, even in place, betwixt the objects and qualities, tho’ the qualities be of such a nature as to admit of no such conjunction, and really exist no where. But more of this fully hereafter. Mean while ’tis sufficient to observe, that the same propensity is the reason, why we suppose necessity and power to lie in the objects we consider, not in our mind, that considers them; notwithstanding it is not possible for us to form the most distant idea of that quality, when it is not taken for the determination of the mind, to pass from the idea of an object to that of its usual attendant. [1.3.14.24, 167]

Hume’s imaginary objector will not quit the ground. The second objection adverts to the manifest absurdity of the claim that physical causation depends upon the mind, rather than the mind’s idea of causation depending on its physical reality. Hume’s account, this objector claims, “reverses the order of nature.” But tho’ this be the only reasonable account we can give of necessity, the contrary notion is so riveted in the mind from the principles above-​mention’d, that I doubt not but my sentiments will be treated by many as extravagant and ridiculous. What! The efficacy of causes lie in the determination of the mind! As if causes did not operate entirely independent of the mind, and

19.  I  thank Kanazori Sawada for calling my attention to this fact, an insight he attributes to Prof. Akira Kume (personal communication). 20. For a slightly different account of this spreading, couched in the Freudian terms of projection, see Kail (2007). I  disagree with Kail on the analogy. Ideas that are projected originate with us and then are attributed to external phenomena. But in this case, the determination of the mind is caused by repeated exposure to pairs of external phenomena. We agree that an inner phenomenon is attributed to the external events; we disagree about the source of that inner phenomenon. But Kail is surely correct when he concludes that Hume takes the phenomenology of feeling that the cause necessitates the effect to derive entirely from the determination of the mind to pass from one to the other.

C au s a l i t y 

( 149 )

wou’d not continue their operation, even tho’ there was no mind existent to contemplate them, or reason concerning them. Thought may well depend on causes for its operation, but not causes on thought. This is to reverse the order of nature, and make that secondary, which is really primary. To every operation there is a power proportion’d; and this power must be plac’d on the body, that operates. If we remove the power from one cause, we must ascribe it to another: But to remove it from all causes, and bestow it on a being, that is no ways related to the cause or effect, but by perceiving them, is a gross absurdity, and contrary to the most certain principles of human reason. [1.3.14.25, 167–​68]

Once again, while Hume’s objector takes Hume to be drawing a metaphysical conclusion about causation, Hume’s response reveals that he is instead making a psychological claim about our causal reasoning, and about the absence of any idea of causation. As Hume points out in the rejoinder, if we have no idea of causation, there is no sense at all in our making claims about causal relations. We are simply talking nonsense when we do, using words to which no ideas correspond, although we are under the misapprehension that they do. Hume here may be the first Western philosopher to have discovered the opacity of mind—​the fact that we can be wrong not only about the truth of our thoughts, not only about their content, but about whether they have any content at all: I can only reply to all these arguments, that the case is here much the same, as if a blind man shou’d pretend to find a great many absurdities in the supposition, that the color of scarlet is not the same at the sound of a trumpet, nor light the same with solidity. If we have really no idea of a power or efficacy in any object, or of any real connexion betwixt causes and effects, ’twill be to little purpose to prove, that an efficacy is necessary in all operations. We do not understand our own meaning in talking so, but ignorantly confound ideas, which are entirely distinct from each other. I am, indeed, ready to allow, that there may be several qualities in both in material and immaterial objects, with which we are on utterly unacquainted; and if we please to call these power or efficacy, ’twill be of little consequence to the world. But when, instead of meaning these unknown qualities, we make the terms of power and efficacy signify something, of which we have a clear idea, and which is incompatible with these objects, to which we apply it, obscurity and error begin then to take place, and we are led astray by a false philosophy. This is the case, when we transfer the determination of the thought to external objects, and suppose any real intelligible connexion betwixt them; that being a quality, which can only belong to the mind that considers them. [1.3.14.26, 168]

( 150 )  Books I and III

The argument of section 14 is the heart of Book I of the Treatise. Here we see Hume’s strategy entire. Hume addresses what looks like a question in metaphysics—​the reality of the causal nexus—​but in fact is addressing a question about human nature, viz., “Do we have an idea of necessary connection, and if not, what idea do we have with which we confuse it, and whence does it come?” He answers that it is an idea of the associative power of the mind that derives from our awareness of our own psychological processes. This is further evidence that Hume adopts a Pyrrhonian strategy with reference to debates about the metaphysical question:  instead of taking sides or withholding judgment, he suspends the debate by demonstrating the falsity of its presupposition, viz., that we even know what we are talking about when we ask whether there is a causal nexus. The key to Hume’s account is his appeal to custom, including both the customary habit of mind that grounds our causal discourse, and the customs that govern that discourse, thus presupposing the results of Book II here in Book I.  Hume deploys the nominalist Pseudo-​idea Principle, showing that apparently cogent philosophical problems arise from our mistaken supposition that we have an idea when in fact we only have a custom for using a word. But why give this investigation of the idea of causation such a central place in the work? Let us return to the Cover Principle. Hume is writing a treatise of human nature, that is, a psychology. That psychology proposes to explain our minds on the analogy of external objects, and we explain the behavior of external objects by appeal to causal laws. Hume is hence appealing to causal laws in order to explain how our minds work, how we get to be the kinds of beings we are. But if he is to understand his own enterprise, Hume must understand what those explanations amount to.21 There is an interesting analogy here to Descartes’s method. Descartes argues that we can’t know anything about the external world until we understand how our own minds—​the instruments by means of which we know that external world—​work, and sets about first to develop that understanding, an understanding which he takes to be different in kind from that understanding it enables of the external world, in virtue of its immediacy and indubitability. Hume agrees with Descartes that a science of the external world requires an understanding of the mind, but argues that we know the mind in the same way that we know the external world. For that analogy to help, we need first to understand the nature of that knowledge of the external world,

21. Coventry (2006) gets this point just right.

C au s a l i t y 

( 151 )

and so the nature of the knowledge we seek. Without the understanding of our own explanatory practices, those practices do not ultimately explain anything. With this understanding in hand, we can proceed to understand all the rest. Finally, to understand our explanatory practices is to understand our customs for explaining phenomena, and to see that those customs are sufficient to explain the normative force we find in nomic claims. The sense of, and the warrant for, causal explanations derives, on this account, like the force of law, from customs. And those customs in turn depend on cognitive processes in which we are largely passive, dispositions of the mind that we can only understand as passions—​calm passions, perhaps, but passions all the same.

C H A P T E R  8

w

Skepticism with Regard to Reason

A

s I have argued, Hume’s method in the Treatise is that of Pyrrhonian skepticism. (Laird and Popkin agree.)1 He inherits this skepticism, as we have seen, from Pierre Bayle, and Hume has clearly read Sextus Empiricus with care, given the understanding he evinces of debates between the Academic and Pyrrhonian schools.2 Of course Hume also denies that he is in fact a skeptic, asserting that nobody ever endorsed that

1.  See Pringle-​Pattison (1890) for a perceptive analysis of the relation of Hume’s skepticism to his adoption of empiricist principles from Berkeley and Locke, and for a discussion of Reid’s critique of Hume’s skepticism. But he fails to see the Pyrrhonian character of Hume’s project. Popkin has done more than anyone to bring Hume’s Pyrrhonism into sharp focus. But Popkin (1980, 103)  takes Pyrrhonian skeptics to be that “fantastical sect,” although Hume’s own method is thoroughly Pyrrhonian. Indeed, as Popkin notes (1980, 103), Hume took the principal difference between himself and the Pyrrhonians to be his refusal to believe the conclusions of skeptical arguments; but that is disingenuous, as Hume himself was familiar with the fact that neither did the Pyrrhonians. Both saw belief as more a matter of instinct, or custom, than a consequence of reason, including their own reasonings. For similar reasons, I disagree with Garrett’s (2015) reading of Hume’s skepticism as non-​Pyrrhonian. Garrett argues that Hume’s skepticism is distinct from that of the Pyrrhonians because it is grounded in the observation of facts about how the mind works (242). I agree that it is; but so, I have argued, is the Pyrrhonian program. The modes of Anasademus, such as the mode from admixtures, or the mode from disagreement of the senses, often refer to the mechanics of the mind, and its infirmities. Hume’s skepticism is distinctive in respect of the detailed account he gives of the workings of the mind, and of the sources of its infirmities, but not in form. 2. James Balfour ([1768] 2000) charges Hume both with failing to see the distinction between Academic and Pyrrhonian skepticism and with falling into a Pyrrhonism that he cannot consistently maintain. Balfour defends Academic skepticism, as a kind of intellectual caution against what he sees as a nihilistic Pyrrhonism. I think that it is clear, however, that it is Balfour who misunderstands the resources of Pyrrhonism. ( 152 )

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 153 )

position:  “Whoever has taken the pains to refute the cavils of this total scepticism, has really disputed without an antagonist” [1.4.1.7, 183]. But that assertion must be taken in context. Hume here avers that he does not assent to the argument he has just offered (to which we will turn in a moment). But that refusal to assent is itself skeptical, and is entirely consistent with the Pyrrhonian attitude toward argument, an attitude Hume confirms when in the next paragraph he asserts that his “intention then in displaying so carefully the arguments of that fantastic sect, is only to make the reader sensible of the truth of my hypothesis, that all our reasonings concerning causes and effects are deriv’d from nothing but custom; and that belief is more properly an act of the sensitive, than of the cogitative part of our natures” [1.4.1.8, 183]. A more succinct statement of Sextus’s position in his exposition of the fourfold prescription for the skeptical life could not be found.3 Hume’s apparent disavowal of skepticism is hence, instead, an explicit embrace of the Pyrrhonian method. And in Hume’s system, that embrace occurs in the context of his naturalism, leading to a concern to trace the mechanisms of that sensitive part of our nature, and the respect in which the passions, not reason, guide our cognitive life.4 We now turn to the sections of Book I of the Treatise in which this program is developed. We

3.  Baier correctly rejects Fogelin’s (1985) assertion that Hume is only a “theoretical” and not a “prescriptive” skeptic (in Fogelin’s terms). Fogelin argues that Hume rejects the warrant of causal inferences on skeptical grounds, yet continues to make them. Baier (1991, 54–​69) replaces Fogelin’s distinction with that between a “true” and a “fantastic” skeptic, reflecting Hume’s terminology more closely. Baier argues that “true” skepticism is the reflexive metaskepticism that calls its own conclusions into question, and that “fantastic” skepticism comprises the “smiling” skepticism of the Pyrrhonists and the “despairing” skepticism of the Academics. This is not the place to rehearse all of the debates regarding the interpretation of ancient skepticism. Suffice it to say here that while I agree with Baier that Fogelin is wrong to say that Hume fails to adhere to his own prescriptions—​simply because Fogelin fails to understand the content of those prescriptions—​I disagree with Baier as well in this regard, both concerning her interpretation of ancient skepticism and concerning where she locates Hume on the skeptical spectrum. Baier, I think, fails to appreciate the degree to which Hume’s “true” skepticism aligns him with the Pyrrhonians. And indeed, I think that Baier comes around to this view a few pages later when she notes correctly that “Hume is giving us an account of how custom, and the ‘accustom’d unions’ it forms in our imaginations, can have ‘equal weight and authority” (EHU 5.2, SBN 41, emphasis Baier’s) (159). 4. Joining Hume’s Pyrrhonism with his naturalism allows us to forgo the complex classification of kinds of skeptical attitudes that Fogelin (1985) deploys, and shows why Fogelin is simply wrong when he writes, “I do not think that Hume is a conceptual skeptic concerning any of the concepts that arise in daily life. Hume will often argue that the plain man is deeply mistaken concerning the true nature of the ideas that correspond to the terms that he uses, but he does not argue that the plain man uses terms that lack a corresponding idea. To the extent that it exists, Hume’s conceptual skepticism seems to be limited to philosophical conceits” (7).

( 154 )  Books I and III

first consider his discussion of skepticism with regard to reason, and in the next chapter, will turn to the much more complex discussion of skepticism with regard to the senses.

1.  THE SUBVERSION OF REASON Hume begins his skeptical argument with an argument for the conclusion that reason itself has no independent warrant as an instrument for gaining knowledge. His argument is a recognizable descendent of Sextus’s discussion of the problem of the criterion in Outlines of Pyrrhonism, but is inflected by Hume’s understanding of the operations of probabilistic reasoning. In all demonstrative sciences the rules are certain and infallible; but when we apply them, our fallible and uncertain faculties are very apt to depart from them, and fall into error. We must, therefore, in every reasoning form a new judgment; as a check or controul on our first judgment or belief, and must enlarge our view to comprehend a kind of history of all the instances, wherein our understanding has deceiv’d us, compar’d with those, wherein its testimony was just and true. Our reason must be consider’d as a kind of cause, of which truth is the natural effect; but such-​a-​one as by the irruption of other causes, and by the inconstancy of our mental powers, may frequently be prevented. By this means all knowledge degenerates into probability; and this probability is greater or less, according to our experience of the veracity or deceitfulness of our understanding, and according to the simplicity or intricacy of the question. [1.4.1.1, 180]

It is easy to misunderstand what Hume is up to in this argument. His principal aim is not to show that we can never have confidence in our

We have already seen that Hume does argue that we often use terms to which no ideas correspond, although we take them to express ideas; we will also see that his skepticism is a uniform policy applied across his philosophical project; although it has different consequences for theoretical and practical life, those divergent consequences do not reflect divergent kinds of skepticism. For an excellent discussion of the way Hume’s naturalism informs, rather than undermines, an account of epistemic norms (while at the same time radically transforming how one might think of epistemic normativity itself) see Broughton (2003). We also see here that Loeb (2005) is wrong to see stability as the primary justificatory principle in the Treatise. Hume’s Pyrrhonian skepticism entails that it must be custom. While, as we noted above, custom can induce stability, it is custom itself, not the stability of that custom, that carries justificatory force.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 155 )

reasoning (this is the kind of skepticism he rejects, and which he correctly points out that nobody ever defends); rather it is to determine why we do have confidence in our reasoning, and to show that the grounds for our confidence are not themselves given by reason. He uses mathematics as his principal example, because it is the domain of mathematical, or logical, reasoning in which we might think we have the best case for reason itself as a justifier. It is also the case mobilized by those such as Descartes, who would argue that knowledge can be attained by a solitary thinker using his or her own cognitive powers, and perhaps most importantly because it is the domain in which certainty is most often claimed. Hume’s analysis will show that, even here, reason is insufficient to confer warrant on our judgments; solitary thought cannot lead to confidence; certainty can never be attained. Hume is concerned here to give his opponent all of the benefits of home court advantage the better to demonstrate the decisiveness of his victory.5 It is very important in this context (as Ainslie [2015] correctly emphasizes, but as Stove [1973] and Fogelin [1985] seem to miss) that when Hume uses the terms knowledge, belief, probability, and evidence, he uses them in a very specific sense, reflecting not only eighteenth-​century philosophical usage that differs from our own, but also Hume’s own idiosyncratic philosophical lexicon. Hume, following Locke (who in turn follows Descartes in this respect), reserves the term knowledge for that of which we can be absolutely certain—​relations of ideas—​such as that conviction which Hume’s targets alleged we get from mathematical proof or philosophical analysis. A belief, on the other hand, is a particularly lively idea, an idea that animates and gives rise to action, a conviction. We may believe something because we know it, but beliefs are caused in a variety of ways, and part of the project of the Treatise is to understand the mechanisms of belief formation. The most important of these terms to consider, though, is probability. It is tempting to think that probability in Hume’s thought is something like the object of the modern probability calculus (as Stove [1996], for instance, reads it). But this would be anachronistic and would indeed make the argument look pretty bad. There simply is no sound argument in the 5. Interested readers might compare Nāgārjuna’s argument for a similar conclusion in Vigrahavyāvartanī—​Reply to Objections. See Garfield (2015) and Westerhoff (2010). There Nāgārjuna argues that our epistemic instruments (pramāṇas) and the objects we come to know through deploying them (prameyas) warrant one another, and that there are no foundations for empirical knowledge, and no Archimedean fulcrum for epistemic justification; justification, he argues, is simply one more conventional epistemic practice.

( 156 )  Books I and III

probability calculus that takes us from premises regarding the probability of our being correct in our assessment of probabilities to the conclusion that it is rational to assign a probability of 0 to a statement that had a nonzero initial probability. And so, if one takes Hume to have a strict mathematical understanding of probability in mind, it is necessary either to rework his argument substantially (as does Bennett [2003]) or judge that Hume has simply offered a patently unsound argument (as does Stove). But we need not proceed this way. It is pretty clear that Hume could not have had in mind a sense of this term in which probability is quantifiable, and represented by parameters over which calculation is possible. For one thing, the very example that Hume chooses to place at the center of this discussion is that of mathematical knowledge. Were he to have had something like the probability calculus in mind, it would have been open to him to draw a nice distinction between our apparent knowledge of the probability, say, of rolling doubles in a game of backgammon and our lack of knowledge that we will or will not roll doubles on the next throw. He does not do so. The closest Hume comes to any mathematical understanding of probability is in his discussion of chance when he says, “Suppose a person to take a dye, form’d after such a manner as that four of its sides are mark’d with one figure, or one number of spots, and two with another. . . . ’Tis plain, he must conclude the one figure to be more probable than the other” [1.3.11.9,  127]. But all Hume can say here is that the die coming up with the figure inscribed on four sides is more probable than it coming up with the figure inscribed on two sides, not how much more probable, or what the probabilities are. There is no quantification of probability here; still less is there a calculus defined that would permit the calculation of probabilities. For another, (if only as a backgammon player) Hume as a natural philosopher, and as an observer of commerce would have had an interest in odds, either with respect to understanding variation from generalizations or in the calculation of interest and insurance. Given the range of issues he brings to his discussion of knowledge and the science of human nature, if he was aware of the probability calculus as a means of gaining knowledge of quantities, it is highly likely that he would have discussed it. He did not, and this suggests that the idea of the quantification of parameters of uncertainty never occurred to him. So, any reading of Hume’s account of the relation between knowledge and probability that ascribes to him our current understanding of probability is probably an anachronistic error. Instead, to say that something is probable in Hume’s sense is to say that although it is not known (in the sense just adumbrated), but to say that it

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 157 )

is something in which we have some degree of defeasible confidence.6 But the term denotes more than this in Hume’s lexicon, and with attention to the legal background that informs his thought about custom, we can add further nuance, nuance that will be important in interpreting these arguments. The word probable is cognate with probative, and connotes an argument that can be used to establish a point, even if nondemonstratively. So, in a court of law—​the context from which Hume’s understanding of custom derives—​adversaries might each advance probable arguments, arguments that aim at (nondemonstratively) proving the correctness of their respective cases. A judge must weigh probable arguments, consider their relative merits, and render a decision. But in doing so, she is not computing probabilities, or otherwise assigning parameters to the arguments or their conclusions over which computations are performed. And her decision does not amount to knowledge of the correct outcome of the case; it is a judgment of probability: in this sense, an assessment of which arguments are most probative, of which have the greatest probative force or weight. It is more likely that it is this process that Hume has in mind when he talks about probability and its relationship to knowledge.7 6. In the Enquiry Concerning the Human Understanding, Hume uses different terminology, distinguishing between reasoning regarding relations of ideas and reasoning regarding matters of fact. The former, he says, are “discoverable by mere operations of thought, without dependence on what is anywhere existent in the universe” (EHU 4.1; SBN 25). On the other hand, “Matters of fact . . . are not ascertained in the same manner; nor is our evidence of their truth, however great, of a like nature with the foregoing. The contrary of every matter of fact is still possible; because it can never imply a contradiction” (EHU 4.1; SBN 25). The relevant distinction is hence slightly different and is drawn on more explicitly epistemological, as opposed to psychological, grounds, reflecting some evolution in Hume’s thought about these matters. So, while the structure of the two discussions may be similar in some respects, they should not be assimilated to one another, and indeed the specific arguments are very different, with the Enquiry argument resting on the impossibility of justifying confidence in conclusions regarding causality, an issue not central to the Treatise argument. The Enquiry argument is of considerable interest, though, in that it connects a number of issues explicitly that are addressed more independently in the Treatise, including the connection between skepticism about causality and the problem of the origin and justification of inductive arguments, and between Hume’s positions regarding causality and induction and animal and infant cognition. And these arguments have a more normative epistemological tendency, despite their grounding in psychology, than those in the Treatise. It is therefore important to treat them independently in any study of Hume’s thought and not to read them back into the Treatise. In this study, I focus exclusively on the argument as presented in the Treatise. 7.  The literature on evidence, probability, and proof is vast, and more recent discussions in fact do turn to probability theory and develop Bayesian models of the weight of evidence at trial. But for accounts of the relation between probable reasoning and probative evidence as they would have been understood in English law at Hume’s time, see (Westerhoff 2010), Cohen (1977), and Wigmore (1983).

( 158 )  Books I and III

If we bear this in mind, we will see both that Hume’s arguments in this section are better than they are often taken to be and that they are homologous with the skeptical arguments he offers elsewhere in the book, adding additional probability to this interpretation.8 Moreover, as Garrett (2015) notes, Hume’s own account of probable reasoning is distinctive. Unlike his predecessors, including Locke, Berkeley, and Malebranche, Hume does not take probable reasoning to involve consciously calling to mind memories of past events or of propositions believed and using them in logical inference. On that more familiar model, the distinctively probable character of such reasoning rests in the uncertainty regarding the truth of these premises, and the looseness of evidential relations linking premises to conclusions. Instead, Hume takes past experience and the beliefs that it induces to function causally, not as premises for the deduction of, but rather as the causes of confidence in, our conclusions. That is, he takes reasoning about matters of fact to be a fundamentally causal process, and any confidence we have in such reasoning must devolve into confidence in the reliability of the causal process, not in the validity of arguments. The output of probable reasoning is evidence, that is, just the property of being evident to us, that is, beliefs of which we are subjectively certain. It is therefore important when reading Hume not to confuse our more contemporary sense of evidence as denoting premises in argument with his understanding, which more concerns the quality of the conclusion, nor to confuse this with Locke’s use of the term evidence, which is closer to our contemporary understanding of that term.9 Hume begins this paragraph by setting aside the question of whether mathematical or logical rules are infallible, and hence whether even perfect deductive mathematical reasoning can yield knowledge; the infallibility of the rules of reason would be insufficient for their generating knowledge, for we—​the users of those rules—​would need to know that we are infallible in employing them. Since we know by experience that our cognitive faculties are fallible, we can have no reason to believe that. Moreover, given our general fallibility, we can have only a probable confidence even in our grasp of the rules of logic and mathematics themselves, so that even if we grant their apodictic character, we cannot be completely confident that we grasp them correctly.

8. See Ainslie (2015) for a similar reading. 9. See also Garrett (1999) and Owen (1999) for careful treatment of Hume’s epistemological vocabulary. Owen, in particular, offers a very detailed account of Hume’s account of probable and demonstrative reasoning that strikes me as completely correct.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 159 )

Therefore, even if deduction carried out properly is guaranteed to deliver truth, it could never be a foundation for genuine knowledge.10 It follows that the most confident epistemic attitude it is rational to assume is a sense of the probability that we are right when we reason; that even our best reason only gives us good reason, or causes us, to be confident; it cannot independently warrant confidence. Hume then turns to a case to make the point: There is no Algebraist nor Mathematician so expert in his science, as to place entire confidence in any truth immediately upon his discovery of it, or regard it as any thing, but a mere probability. Every time he runs over his proofs, his confidence encreases; but still more by the approbation of his friends; and is rais’d to its utmost perfection by the universal assent and applauses of the learned world. Now ’tis evident, that this gradual encrease of assurance is nothing but the addition of new probabilities, and is deriv’d from the constant union of causes and effects, according to past experience and observation. [1.4.1.2, 180–​81]

Here Hume explains why reason itself cannot be the cause of our confidence in reason and also tells what the actual cause of that confidence is—​custom. First, he notes, no expert mathematician would ever trust his own reasoning the first time he proves an important result. Any mathematician of any experience knows that he might have made a mistake, and so he checks his proofs. The very fact that he does so and that his confidence increases each time he checks shows that at each stage his confidence is merely a matter of probability, not of certainty. And this means, as we just saw, that our confidence in the correctness of even demonstrative reasoning can only be an effect of our previous experience with such reasoning, just as legal arguments acquire weight from precedent, and just as custom acquires normative force from long usage. Therefore, the degree of evidence of the conclusion depends on prior cognitive causes and the customary associations they establish, not on valid argument itself. That is enough to show that reason itself is not sufficient for justification; for if

10. And this, as Allison (2008) suggests, is a reply to Malebranche, who, like Descartes, takes our knowledge of arithmetic truths to be guaranteed by simple direct perception. Hume can grant that we might detect truths in this way, but points out that knowledge requires that we can also endorse that perception, which requires confidence in our faculties and reliance on memory. Marušić (2016), however, provides good reason for thinking that Hume may have Locke in his sights. Locke is keen, she argues, to demonstrate the clear distinction between knowledge and probability; Hume is concerned to undermine precisely that distinction here.

( 160 )  Books I and III

reason were, no additional check would be necessary. But it also shows that reason is not the cause of confidence, which is the real issue here; instead, repeated checks are. That confidence is increased again when the rough draft is sent to colleagues who concur, and still more when the reviewers accept the piece for publication and it receives no refutations. Again, the cause in each case of the mathematician’s increased confidence is not the addition of more reasoning, but rather custom, in both senses:  the agreement of others constitutes a customary review of the proof, and the effect of that agreement on the mathematician is the increase of confidence, a matter of psychological custom. As a consequence, what causes belief in—​or, in Hume’s sense of the term—​the evidence of the conclusion is not knowledge, but the addition of more of what we trust in a merely probable sense. Hume then generalizes the case from the abstract realm of mathematics to the marketplace, indicating again that our confidence in reasoning is itself dependent upon custom in these senses: In accompts of any length or importance, Merchants seldom trust to the infallible certainty of numbers for their security; but by the artificial structure of the accompts, produce a probability beyond what is deriv’d from the skill and experience of the accomptant. For that is plainly of itself some degree of probability. . . . But knowledge and probability are of such contrary and disagreeing natures, that they cannot well run insensibly into each other, and that because they will not divide, but must be either entirely present, or entirely absent. Besides, if any single addition were certain, every one wou’d be so, and consequently the whole or total sum. . . . I had almost said, that this was certain; but I reflect, that it must reduce itself, as well as every other reasoning, and from knowledge degenerate into probability. [1.4.1.3, 181]

The first step here is the generalization: it is not only mathematicians whose certainty is both justified and caused by custom, but each of us in our daily lives. Second, Hume points out, this fact simply rules out any analysis of epistemic confidence as certainty grounded in rational justification. We must bear in mind that this is what Hume intends by “knowledge” in this case, and that this is also the sense in which the term is used by Bayle in his own discussion of skepticism, as well as by Locke in drawing the distinction between probabilistic belief and genuine knowledge. Finally, Hume lays his Pyrrhonian cards on the table, applying his own skeptical principle to his own argument in the reflexivity Sextus compares to the action of a purgative that must expel itself with the matter it is taken to purge, a reflexivity that distinguishes the Pyrrhonian from the more

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 161 )

conservative Academic skepticism.11 Hume is not even willing to be confident that his own lack of confidence in reason is justified. He now turns to the problem of coming to understand the way in which custom in fact leads to confidence. Here we will see the union of Hume’s naturalism and skepticism in his quest for an understanding of the mechanisms of cognition. Since therefore all knowledge resolves itself into probability, and becomes at last of the same nature with that evidence, which we employ in common life, we must now examine this latter species of reasoning, and see on what foundation it stands. [1.4.1.4, 181]

Note that Hume does not assert here that the consequence of the arguments he has offered is that we have no knowledge at all, but rather that “knowledge resolves itself into probability.” His skepticism with regard to reason is then not an attack on the possibility of knowledge in its ordinary sense, but a naturalistic demystification of knowledge and an exploration of what the knowledge “we employ in common life” really is. That exploration involves a search for the nature of epistemic warrant as we find it in our actual, customary practices. The first premise of the argument that leads to Hume’s view about this matter is that we are more confident of our views the more experience we have in a domain: ’Tis certain a man of solid sense and long experience ought to have, and usually has, a greater assurance in his opinions, than one that is foolish and ignorant, and that our sentiments have different degrees of authority, even with ourselves, in proportion to the degrees of our reason and experience. [1.4.1.5, 182]

Nonetheless, the more experience we have, the more errors we will have committed, and the more seriously we will take the possibility of error: In the man of the best sense and longest experience, this authority is never entire; since even such-​a-​one must be conscious of many errors in the past, and must still dread the like for the future. [1.4.1.5, 182] 11.  Candrakīrti uses the same metaphor for his own skeptical arguments in Prasannapadā (Lucid Words), echoing Nāgārjuna’s emphasis in Mūlamadhyamakakārikā (Fundamental Verses on the Middle Way) that the emptiness of emptiness is what allows us to recover conventional truth. Candrakīrti’s point is that the skeptical arguments do not establish an alternative view regarding the fundamental nature of reality to that against which they are mounted, but rather that they set aside the entire program of establishing a fundamental nature of reality, simply leaving us to rest content with our conventions for engaging with reality as we experience it. See Nāgārjuna (1995) and Garfield (2015).

( 162 )  Books I and III

Now things get complicated. This is the point where it is tempting to contemporary readers to take Hume to be presenting a fallacious argument within probability theory. But if we bear in mind the context of his language and the sense of the terms he is deploying, it is plain that he is not. He is instead calling our attention to the accumulation of causes of uncertainty in our estimation of the degree of confidence we ought to have in our own reasoning, just as a defense attorney would accumulate reasons for jurors to entertain uncertainty regarding the testimony of witnesses for the prosecution, as she builds a legal case. Our “man of solid sense and long experience,” aware of his own fallibility, decides to adjust his level of confidence in his own judgment by reducing it in proportion to his error rate on such matters. This need not involve a calculation; no numerical parameters are involved. After making these adjustments, he then realizes that to be sure that he has done so correctly, he would have to know his own error rate. He is fallible about that, too, and so he needs to adjust that confidence before he can adjust the first. Here then arises a new species of probability to correct and regulate the first, and fix its just standard and proportion.  .  .  . [H]‌aving adjusted these two together, we are oblig’d by our reason to add a new doubt deriv’d from the possibility of error in the estimation we make of the truth and fidelity of our faculties. [1.4.1.5–​6, 182]

So now we not only know that we are fallible, but we know that we don’t know how fallible we actually are, and each attempt to make the requisite adjustment in our estimation of our own fallibility presupposes knowledge of our error rate at that level that is equally inaccessible: But this decision, tho’ it shou’d be favourable to our preceding judgment, being founded only on probability, must weaken still further our first evidence, and must itself be weaken’d by a fourth doubt of the same kind, and so on in infinitum; till at last there remain nothing of the original probability, however great we may suppose it to have been, and however small the diminution by every new uncertainty. [1.4.1.6, 182]

Hume is not arguing—​as it might appear—​that we must say that the probability we assign to the original belief reduces to zero. That would be fallacious, and, as I noted above, it would also be an argument not available to Hume, as he does not even have the probability calculus at his disposal.12 12. Popkin (1980, 109–​10) and Allison (2008) attribute this fallacious argument to him and defend it. Stove (1966, 1973) and Fogelin (1985) also seem to miss the structure of this argument completely, but they charge Hume with a crude error in the use of the probability calculus.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 163 )

We must remember that Hume’s use of the term probability is not ours: by probability Hume means confidence, our willingness to assert, and probative force. Our merchant starts out confident in his books. He then realizes that his accountant may have made an error, and so decides to revise down his confidence by some amount; but he then realizes that he can’t be confident of that amount, and so on ad infinitum. The result is that he can’t even know how confident he can be of his accounts, let alone whether they are accurate. So at this point, if reasoning alone is what warrants belief, he simply has no warrant for any assertion at all, since warrant at least requires us to know how confident we are regarding that which we assert. And so, Hume concludes, if we only consider the objects of our reasoning, there can be no basis for belief at all. Reason itself subverts any reason for any assertion: I have less confidence in my opinions, than when I  only consider the objects concerning which I reason; and when I proceed still farther, to turn the scrutiny against every successive estimation I make of my faculties, all the rules of logic require a continual diminution, and at last a total extinction of belief and evidence. [1.4.1.6, 183]13

Note carefully the conclusion of this argument. Hume argues not that he does or should “have less confidence in [his] opinions” tout court. Instead he argues that if reason is the foundation for warrant, or is taken to be the cause of that confidence, then we could have no warrant, and nothing could account for that confidence. Nor is this an argument against the probative use of reason; it is an argument against the probative use of reason alone, and against the idea that the probative force of reason can be found within reason.

2.  THE SKEPTICAL REHABILITATION OF REASON One might think this is the end of the matter, but the Pyrrhonian turn is yet to come. Hume immediately takes the moral of this argument not to be that we are not warranted in belief, but that reason cannot be the warrant for our beliefs; not that a survey of this argument reveals that we in fact believe nothing, but that rational argument is not the principal cause of belief: Shou’d it here be ask’d me, whether I sincerely assent to this argument, which I  seem to take such pains to inculcate, and whether I  be really one of those

13. The echo of Bayle here is unmistakable. Bayle writes that Pyrrhonism leads to “the total extinction not only of faith, but of reason” (1965, 207).

( 164 )  Books I and III sceptics, who hold that all is uncertain, and that our judgment is not in any thing possest of any measures of truth and falshood; I  shou’d reply, that this question is entirely superfluous, and that neither I, nor any other person was ever sincerely and constantly of that opinion. Nature, by an absolute and uncontroulable necessity has determin’d us to judge as well as to breathe and feel; nor can we any more forbear viewing certain objects in a stronger and fuller light, upon account of their customary connexion with a present impression, than we can hinder ourselves from thinking as long as we are awake, or seeing the surrounding bodies, when we turn our eyes towards them in broad sunshine. Whoever has taken the pains to refute the cavils of this total scepticism, has really disputed without an antagonist, and endeavour’d by arguments to establish a faculty, which nature has antecedently implanted in the mind, and render’d unavoidable. [1.4.1.7, 183]

This is one of the most important passages for understanding Hume’s proj­ ect in Book I of the Treatise. Let us work through it with care. Hume first affirms that he does not deny, on the basis of this argument, that we are ever committed to our judgments. Nobody ever sincerely refuses to assent or to deny. The important question is, then, “Why are we committed to them?” Hume’s answer is that it is not because there are good reasons for believing, reasons that we somehow overlooked in the previous argument, but rather because nature causes us to do so. What nature? Our nature. Belief isn’t something at which we arrive for reasons, Hume argues, but something that is caused in us, albeit often by arguments, as, perhaps, his argument might cause us to believe that beliefs are caused, not justified (but not only by arguments—​a point Wittgenstein makes with particular force in his Humean analysis of knowledge in On Certainty). Once again, it is not only customs in the sense of social conventions such as checking each other’s work, or relying on testimony that causes beliefs, but also in the sense of a “customary connexion” between a typical cause of a belief and that belief. This is not a reason to suspend belief, but rather to acknowledge that our belief-​forming mechanisms are simply natural.14 Nor is it a retreat from the project of explaining and grounding 14. Baier puts the point this way: “[Hume defends] natural sentiment not as mere distraction, but as the replacer of reason. Reason must be worked through, taken to the end of its tether, before sentiment can take over the guiding role” (1991, 20). But this is not quite right, although it is close. It is more precise to say that Hume is arguing that sentiment (a shorthand for habit, instinct, custom) is and always has been the guide to belief formation, and to the extent that we reason, that is simply to engage in one more custom. A few lines later, Baier writes, “Here, at the transitional point between Book One’s reductio ad absurdum of Cartesian intellect and the rest of the Treatise’s development

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 165 )

norms of assertion in favor of “mere” description. For, as we have seen, on Hume’s understanding, custom is not a substitute for warrant, but a source of warrant. Hume sums this up in language that recalls Malebranche (La rechèrche de la verité), but is entirely Pyrrhonian in character: My intention then in displaying so carefully the arguments of that fantastic sect, is only to make the reader sensible of the truth of my hypothesis, that all our reasonings concerning causes and effects are deriv’d from nothing but custom; and that belief is more properly an act of the sensitive, than of the cogitative part of our natures. [1.4.1.8, 183]

Hume drives this point home forcefully with the following observation about the power of skeptical arguments themselves. If belief were a rational matter, rather than the effect of discursive and perceptual causes, then, given the manifest cogency of arguments such as the one just offered, we would believe nothing at all. The fact that we manifestly do have beliefs therefore shows that it is not:15 If belief . . . were a simple act of thought, without any peculiar manner of conception, or the addition of a force and vivacity, it must infallibly destroy itself,

of its more passionate and sociable successor, Hume gives us a short . . . preview of the dialectic of the passions which is to be more fully developed in Books Two and Three” (1991, 21). Again, this is close, but not precise enough. As I have been arguing, this discussion is not a preparation for, but rather presupposes, the discussion of the passions in Book II. It simply makes no sense on its own, or is at best a promissory note. Hume has chosen a path of exposition that begins in Book I and moves to Book II, but the conceptual structure of the Treatise is the reverse. Just as one might show someone through a house first on the main floor and only take them to the basement later, to inspect the foundations, although those foundations hold up the house, Hume is taking us through the epistemology that rests on the doctrine of the passions before taking us to the foundation that grounds them. 15.  This is probably where I  disagree most directly with Qu’s (2015b) account of Hume’s epistemology—​although I  agree with him in most matters. Qu argues that Hume’s positive epistemology is in a sense opposed to his skepticism and mitigates it. I see the positive epistemology as constituting a descriptive account of custom within the scope of his Pyrrhonian skepticism, as part of the “fourfold prescription.” So, I agree with Qu that Hume’s project is to ground epistemic norms on the basis of naturalistic descriptions of our epistemic practice (and I endorse his lucid exposition and defense of that Humean project). But I disagree that those norms are meant to have a status that transcends custom. In the end, their normativity derives from the power of custom to regulate our lives, a power that we have seen ascribed in the legal theory of Hume’s time. And this is true as much in the moral as in the epistemic domain, as we shall see. (About this we agree [Qu 2014].) For similar reasons I disagree with the critique of what Qu calls the “traditional interpretation” of Hume’s epistemology, which he attributes to Kemp Smith (1941). Loeb

( 166 )  Books I and III and in every case terminate in a total suspense of judgment. But as experience will sufficiently convince any one, . . . tho’ he can find no error in the foregoing arguments, yet he still continues to believe, and think, and reason as usual, he may safely conclude, that his reasoning and belief is some sensation or peculiar manner of conception, which ’tis impossible for mere ideas and reflections to destroy. [1.4.1.8, 184]

This passage is neither entirely precise nor entirely clear, but it is an important comment on the import of the previous arguments. The point is that belief cannot simply be a product of reason; if it were, we would believe nothing. And this leads to a final puzzle about the impotency of skeptical arguments. After all, Hume himself is offering arguments, and seems committed to the power of the rational argument to compel assent. His skeptical arguments in this section, like the arguments of Sextus Empiricus that inspire them, have as an apparent conclusion the fact that none of our beliefs are rationally warranted. Why, he asks, do we not draw just this conclusion, if not rationally, then at least as the effect of these arguments? But here, perhaps, it may be demanded, how it happens, even upon my hypothesis, that these arguments above-​explain’d produce not a total suspense of judgment, and after what manner the mind ever retains a degree of assurance in any subject? . . . ’Tis therefore demanded, how it happens, that even after all we retain a degree of belief, which is sufficient for our purpose, either in philosophy or common life. [1.4.1.9, 184–​85]

Hume’s answer does not betray his skepticism, nor does it abandon his naturalism: (2011b), another proponent of this “traditional interpretation,” argues that Hume—​far from undermining the rational credentials of induction (for instance)—​expresses “approval of inductive inference” and “persists in attributing causal inference to ‘reason,’ which he reconstructs as a component of the faculty of association carrying epistemic pride of place” (Kemp Smith 1941) (Loeb 2011b, 115). While the distance between my reading and Loeb’s is not great, I do think that it is important to see that Hume does not return rational credentials to inductive reasoning in approving it, but rather sees it as grounded in custom. The naturalism and skepticism are firmly in place, and trump any rationalist defense of any kind of reasoning. In this, Kemp Smith was correct. A bit later Loeb writes, “In attributing inductive inference to custom, Hume sees himself putting it on a firm epistemic footing. . . . [This brings to light] externalist strands in Hume’s thinking that begin to explain how he could assign inductive inference a positive epistemic status: because it results from custom” (2011b, 117). Depending on how one reads this, it is either absolutely correct or dead wrong. If Loeb means that custom gives rational warrant to inductive reasoning, that is incorrect; but if he means that the only warrant any practice can have is customary, then that is exactly right. But that is a skeptical conclusion, not an antiskeptical conclusion.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 167 )

Reason first appears in possession of the throne, prescribing laws, and imposing maxims, with an absolute sway and authority. Her enemy, therefore, is oblig’d to take shelter under her protection, and by making use of rational arguments to prove the fallaciousness and imbecility of reason, produces, in a manner, a patent under her hand and seal. This patent has at first an authority, proportion’d to the present and immediate authority of reason, from which it is deriv’d. But as it is suppos’d to be contradictory to reason, it gradually diminishes the force of that governing power, and its own at the same time; till at last they both vanish away into nothing. . . . The sceptical and dogmatic reasons are of the same kind, tho’ contrary in their operation and tendency; so that where the latter is strong, it has an enemy of equal force in the former to encounter; and as their forces were at first equal, they still continue so. . . . ’Tis happy, therefore, that nature breaks the force of all sceptical arguments in time, and keeps them from having any considerable influence on the understanding. Were we to trust entirely to their self-​destruction, that can never take place, ’till they have first subverted all conviction, and have totally destroy’d human reason. [1.4.1.12, 186–​87]

Hume here presents a psychological process as the explanation of the impotence of skeptical arguments even over those who endorse them. We begin, as we began this discussion, with full faith in reason as the origin and validator of our beliefs. Nonetheless, rational arguments themselves prove that reason is incapable of this task. The effect of those arguments is to weaken our faith in reason; but since it is our very faith in reason that leads us in the first place to endorse those arguments, that diminution in confidence in reason weakens the force of skeptical arguments themselves. There is thus an equilibrium of forces at play in the mind. Rational arguments cause us to disbelieve reason; the disbelief in the probative force of rational arguments causes us to withdraw assent from those arguments. There is probability in each side, leading to a kind of hung jury. Skeptical arguments may then be convincing, but they can never be powerful enough to overcome the other factors that fix belief, including prominently the customary force of argument as a cause of belief. Hume thus, like any good Pyrrhonian, turns his skeptical arguments on those arguments themselves, trusting them to act as a dialectical purgative.16 16. Fogelin (2009) asserts that in his section concerning skepticism with regard to reason, Hume shows no sign of recognizing the precarious character of his own position relative to his own skeptical argument. Hume seems to see himself as standing above the fray while the skeptic and the dogmatist engage in mortal combat that inevitably leads to their mutual destruction. It doesn't seem to cross his mind that he himself could be swept up in the combat with a similar outcome. (54)

( 168 )  Books I and III

Baier notes a nice parallel between this discussion and Hume’s discussion of the idea of causality. In each case, we begin with the natural assumption that the relevant connection (cause to effect in that case; premise to conclusion in this) is transcendental to and known by the mind; in each case, we discover that it instead lies within the mind and is projected onto the phenomena we take to be linked by it: The [epistemic] authority [of logical and mathematical arguments] is redescribed by Hume as a special case of the more comprehensive epistemic authority that he is suggesting that we should acknowledge. Even in our demonstrative thought moves, he claims, the necessity that we take to license and require the move to the conclusion is like causal necessity in belonging “entirely to the soul,” a projection onto our subject matter of “the determination of the mind” in inference. So all inference, demonstration as well as causal inference, traces a relation whose necessity is “spread” by the mind from itself onto its subject matter. (1994, 81)

This is further confirmation of the Principle of the Uniformity of Method, and of the remarkable uniformity in Hume’s use of custom as an analytic device to understand not only the operation of our passions, but also the epistemology and morality they ground. I take this uniformity to be an important argument in favor of this reading of Hume’s skepticism regarding reason. This latter point is emphasized by Broughton: Hume’s naturalism [is] at least partly constituted by his readiness to do in his philosophical inquiry what natural philosophers do in theirs and what we all do in everyday life: accept as authoritative various broad norms for the use of human faculties of cognition. He does not attempt to find some special vantage-​point from which to ground or validate the authority of these norms, and he does not raise skeptical worries about them, either.17 He simply takes them as his own, as anyone will who is discriminating in forming ideas and judgments. . . . But for Hume radical skepticism is all the same a conclusion we must reach: adhering to our cognitive norms requires us to regard our most general and basic beliefs about the world as utterly unjustified. (2003, 16)

As we have seen, this is a serious misreading of Hume’s attitude toward his own argument. Instead, Hume is explicitly concerned with the implications of his skepticism for the assessment of his own defense of skepticism. 17. Here Broughton seems to be using the term skepticism in its Cartesian sense, not the Pyrrhonian sense that informs the remainder of her discussion.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 169 )

Broughton sees clearly the seamless unity of Hume’s naturalism and skepticism. These positions are not in tension, and this is because the naturalism is the ground for the skepticism; moreover, since the skepticism Hume advances is Pyrrhonian in character, it also serves as support for Hume’s naturalism. One way to see this is that the positive side of Hume’s Pyrrhonism follows Sextus’s fourfold prescription, and so recommends adherence to the arts (including the arts of reason and persuasion), following our instincts (including our dialectical and logical instincts), our senses, and, most importantly for Hume, custom. And our reliance on custom recognizes that custom itself is ungrounded. This is why Broughton is correct to say that Hume argues that “adhering to our cognitive norms requires us to regard our most general and basic beliefs about the world as utterly unjustified.” It is the cognitive norms that are at fault here, not our beliefs. On Hume’s view, we are justified in our researches and in our actions because of custom, and for no other reason; nothing external to custom could be justificatory. But that is not to abandon the normativity of our epistemic and moral lives; it is to explain it, and to explain the source of its force in a way completely consistent with a naturalistic understanding of human nature.

3.  METHODOLOGICAL MORALS We see in this section confirmation and defense of several of the principles of interpretation I articulate at the outset of this study. First, there is the Cover Principle: Hume is not pursuing an epistemological agenda: he is neither defending the view that our beliefs are never justified nor providing an alternative to reason as an account of justification. Instead, his skeptical arguments are intended as an entrée into an account of human nature, in this case, the psychological principles governing the fixation of belief. He is concerned to dislodge one popular account of those principles as essentially rational, and to replace them with a purely causal account. We also see the skeptical inversion at work once we appreciate the truly Pyrrhonian structure of Hume’s account. We can imagine two dogmatic attitudes toward the relation between our conventions regarding the acceptance of reasons and arguments as justificatory: on the one hand, one might argue that these conventions are justified because reason can be shown to be a validator of our beliefs. This is the approach of Descartes. On the other hand, one might argue—​as Hume is often taken to argue, and as he is at pains to deny that he does—​that since reason has no such power, all of our conventions regarding belief fixation are unjustified and are to be rejected.

( 170 )  Books I and III

Each of these dogmatic extremes presupposes that our epistemic conventions require justification, and that they are justified if, and only if, good arguments can be provided for them. Hume, like Sextus, rejects both extremes not in favor of compromise, but in favor of an inversion achieved by rejecting that shared biconditional. That is, Hume argues that our epistemic customs require no justification at all: they are part of our nature. And it is our customs—​the brute facts about how we think and how we interact as members of epistemic communities—​that explain why arguments work in the first place, although that explanation reveals that they do not work in the way that the dogmatist might have thought that they must.18 And this in turn recalls the central role of custom in Hume’s enterprise. When Hume provides an account of human nature, and when he provides a skeptical analysis of a human practice—​in this case the practice of discursive justification—​he locates custom, comprising both the natural laws of human psychology and our common conventions, as the explanatory bedrock. Just as the Newtonian places laws of nature at the foundation of explanation, Hume places regularities of thought and behavior and social regularities at the bedrock. In doing so, he is consistent with the classical skeptical tradition, but advances that tradition in line with a modern conception of explanation and natural science and a social understanding of human nature. And, just as Hume took custom to ground the normative force of law—​not to undermine it—​he takes custom to ground the norms that govern epistemic practice—​not to undermine them. Finally, we can see in this discussion the fundamental role of the passions. Belief, Hume asserts, is sensitive, not cogitative. Both Malebranche and Hutcheson are in the background, but the architecture of the account is Hume’s own. In belief formation we are passive; and so it is the gentle passions modulated by our social relations that take us from the causes of belief to their effects. It is therefore the anatomy of the passions and of second nature developed with such care in Book II that undergirds the natural science of belief developed here in Book I. Many have seen Hume’s discussion of reason in this section and elsewhere in the Treatise as an attack on the authority of reason, per se. That would be an error. As Baier (1991, 14) reminds us, “Hume’s project all along

18.  As Baier notes, “We need each other’s help in judging the fidelity or ‘truth’ of representations, and we have that help. Persons among persons are the liveliest objects of our mental attention, in part because we depend in so many ways upon those persons” (1991, 47). And, as Baier also notes, this social dimension of epistemology is mirrored in Hume’s ethics.

S k ep t i c i s m wi t h Re g a r d t o  Re a s o n  

( 171 )

has not been so much to dethrone reason as to enlarge our conception of it, to make it social and passionate.” And as she remarks later in that study, “Reason in solitude becomes not just abstruse but monstrous. Reason’s appropriate nourishment is civilized debate within ‘the republic of letters,’ along with thoughtful response even outside that republic.  .  .  . Hume’s ‘deference’ to his readers is required of him by his own reconstruction of reason as social, concerned and responsive” (1991, 284). We will see this pattern continue as Hume now turns from reason to the senses. While reason governs the transition between beliefs, which we now come to understand as determined by causal processes, the senses fix belief in the first place. One might think naively of the senses as simply delivering the external world to consciousness, and then worry that the correct question to ask about them is, “How accurately do they do so?” Hume will once again utilize skeptical, naturalistic analyses in terms of custom and the passions that govern our thought to reject this approach and to reconstruct both the questions we must ask about our knowledge of the external world and the answers we can give to those questions. But there will be one important difference: whereas in the discussion of skepticism with regard to reason, the sense of “custom” at work has been social convention; in the discussion of skepticism with regard to the senses, Hume focuses on custom in its guise as habit. In each case, we encounter a naturalistic skepticism involving a central appeal to custom as an explanans. The only difference is in the sense of custom at issue: our commitment to reason, Hume has argued, is social; our commitment to the existence of the external world is instinctive.

C H A P T E R  9

w

Of Skepticism with Regard to the Senses

S

ection 1.4.2 is the most complex in the Treatise, and the argument can be hard to follow. We will begin with Hume’s framing of the problem—​his insistence that skepticism with regard to the senses does not arise from asking whether there is an external world, but rather from asking why we believe that there is one. It will be important to bear in mind Hume’s assertion that we take the existence of the external world for granted “in all our reasonings.” We will then follow Hume’s argument that the faculty responsible for this belief is neither the senses nor reason, but the imagination, and that it induces this belief through causing us to gain a kind of confidence in the “constancy” and “independ­ ence” of our perceptions. This conclusion, we will see, raises a deeper problem: the need to account for what appears to be an unreasonable confidence in the reality of the external world and our knowledge of it once we realize the slenderness of the thread on which our belief depends. It is at this point that the positive side of Hume’s Pyrrhonism comes to the fore, demonstrating the power of custom to explain this confidence and vindicating the naturalism that undergirds the explanation. The argument is complex. But if we keep our set of interpretative principles in the foreground, it will come into focus; in particular, we will see that the deployment of the naturalistic, psychologistic, skeptical method underwritten by a robust sense of the explanatory power of custom in both its collective and individual senses will reveal striking homologies between Hume’s account of our knowledge of the external world and his accounts of our causal reasoning and of our mathematical reasoning. This in turn will vindicate the Principle of the Uniformity of Method. Hume begins by ( 172 )

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 173 )

connecting this section to the foregoing and once again, very explicitly announcing a Pyrrhonian methodology:

1.  SETTING THE STAGE: THE EXTERNAL WORLD Thus the sceptic continues to reason and believe, even tho’ he asserts, that he cannot defend his reason by reason; and by the same rule he must assent to the principle concerning the existence of body, tho’ he cannot pretend by any arguments of philosophy to maintain its veracity. Nature has not left this to his choice, and has doubtless esteem’d it an affair of too great importance to be trusted to our uncertain reasonings and speculations. We may well ask, What causes induce us to believe in the existence of body? But ’tis vain to ask, Whether there be body or not? That is a point, which we must take for granted in all our reasonings. [1.4.2.1, 187]

One might think that a discussion of skepticism with regard to the senses would be framed with a question such as “Is there really an external world?” or “To what degree can we trust our senses?” That would be to mistake Hume’s Pyrrhonian skepticism for Descartes’s modern version.1 Hume begins by noting explicitly that just as the conclusion of the last section was not that the impotence of reason as a self-​justifier leads us not to trust reason, no demonstration that the senses cannot demonstrate to us the existence of the external world could ever lead us to believe that there is no such world. That, he points out, “we must take for granted in all our reasonings.”2 Moreover, the existence of the external world is no part of the subject matter of the Treatise, which is after all about human nature. So, Hume points out, the right question to ask is not whether there is an external world, but what could ever cause us to place such faith in its existence. 1. As Fogelin does, when he says that “Hume does not forward skeptical arguments against our perceptual faculties” (1985, 64). Fogelin clearly has Cartesian skepticism in mind here. We can grant that Hume doesn’t forward Cartesian arguments, but he does forward Pyrrhonian arguments. 2.  And indeed Hume does take it for granted at the outset of the argument, and returns at the end of the argument to explain why we must take it for granted. This idea that we must take the external world for granted, as I noted above, is an early antecedent of Wittgenstein’s idea of “hinge propositions” in On Certainty, propositions we do not know to be true, but which must be assumed, or taken for granted, in order for any inquiry or justification to be possible. There is, however, an important disanalogy between Hume’s assumption and Wittgenstein’s hinge:  Wittgenstein emphasizes that hinge propositions are never called into question in our epistemic life; Hume, in this argument, subjects this very assumption to searching critique before returning to it. While, like Wittgenstein, he eschews any transcendental epistemological justification of the assumption, he does provide a justificatory causal explanation of our confidence in it.

( 174 )  Books I and III

That is the question that occupies this section. It is recalling a point we made in ­chapter 6, in our initial discussion of Hume’s thinking about our belief in the external world, that the kind of believing at issue here is not so much a believing that as it is a believing in, a taking for granted of the external world.3 The reading I  advance will hence be at odds with those who defend a “skeptical realist” interpretation of the Treatise. Wright contends, for instance, that “a central aim of Hume’s philosophy of the understanding is to show that we retain a commerce with a world of independent objects through a species of natural judgments, which involve a systematic confusion of ideas” (1983, 5). I will argue that Hume is not even interested in the reality of external objects.4 I disagree with Strawson’s (1989) reading of Hume for the same reason: he, like Wright, reads Hume as a skeptical realist, because he notes that Hume takes for granted the reality of the external world and defends the probity of causal reasoning, while denying that we have adequate ideas of each. This reading would make sense if, and only if, one takes the Treatise to be a book about metaphysics and epistemology; that is, however, as I have been arguing, simply the wrong way to read the text. Hume notes that for us to take an object to be in the external world is to take it to have two characteristics: to exist continuously, even when not observed, and to be distinct from the mind that observes it. Hume therefore immediately resolves the fundamental question of what causes us to

3. See Strawson (1985) for a similar analysis. Wright’s view (1983, 39–​40) is in the same neighborhood, but somewhat different. Wright agrees with Strawson and with me that Hume takes for granted the existence of body at the outset, and that his primary goal is to determine the origins of this belief. Nonetheless, he also thinks that Hume seeks to justify this belief, where that amounts to showing that it is highly likely that the belief is true. There we diverge. Hume will argue both that we have no choice but to believe in the external world, and that custom grounds that belief. It is therefore justified in a customary sense, but not because we have independent reasons for believing it true; only because custom is as far as we can go in justifying such beliefs. Fogelin (1985) takes it that Hume claims that we have uniform, and uniformly false beliefs about our senses and the external world. This confuses believing that with believing in. My reconstruction of Hume’s argument will suggest a very different reading, closer to Strawson’s and to Baier’s. 4. Despite this difference in our readings, I agree with Wright regarding the consistency of Hume’s skepticism with a realism about the world, and regarding the crucial role of Hume’s naturalism and commitment to the role of the imagination in underpinning that skepticism. Our only disagreement—​and it is an important one—​concerns the aim of the Treatise. Wright takes it to be metaphysical and epistemological; I take it to be primarily psychological. While we both read Hume as a naturalist, I take that naturalism to be directed to Hume’s analysis of thought, while Wright takes it to be directed to Hume’s account of the nature of the external world.

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 175 )

believe in the external world into two subordinate questions: “What causes us to believe in the continued existence of objects of which we have only interrupted acquaintance?”; and “What causes us to believe in the distinctness of objects to which we have no access independent of our own perceptual and cognitive states?”: We ought to examine apart these two questions, which are commonly confounded together, viz., Why we attribute a continu’d existence to objects, even when they are not present to the senses; and why we suppose them to have an existence distinct from the mind and perception.  .  .  . These two questions concerning the continu’d and distinct existence of body are intimately connected together. For if the objects of our senses continue to exist, even when they are not perceiv’d, their existence is of course independent of and distinct from the perception; and vice versa  .  .  .  . But tho’ the decision of the one question decides the other; yet that we may the more easily discover the principles of human nature, from whence the decision arises, we shall carry along with us this distinction, and shall consider, whether it be the senses, reason, or the imagination, that produces the opinion of a continu’d or of a distinct existence. These are the only questions, that are intelligible on the present subject. For as to the notion of external existence, when taken for something specifically different from our perceptions, we have already shewn its absurdity. [1.4.2.2, 188]

In distinguishing these questions, Hume also notes that they are not independent of one another: continuity entails distinctness (and vice versa). Nonetheless, he suggests, there is a methodological purpose in pursuing them separately, for the causes of the relevant beliefs may not be identical. Hume frames the broader question in terms of which faculty is responsible for our belief in the external world. He will follow his distinctive strategy of focusing on the imagination, rather than on the senses or reason, as a source of this belief. Once again, he argues by a process of elimination, beginning with the senses, then eliminating reason, before turning to the imagination.5

5.  It is because of his commitment to the imagination, as opposed to the senses, as the source of our belief in the external world that Hume is not, as both Passmore (1952) and Fogelin (1985) suggest, a kind of phenomenalist. Our belief in the existence of body is not the belief in something ultimately composed of perceptions, or even as Fogelin suggests, “grounded in a system of relations among perceptions” (68). Instead, it is caused by the operation of the imagination, and is an instinctive, rather than a rational or even properly cognitive affair.

( 176 )  Books I and III

2.  THE SENSES AND THE EXTERNAL WORLD Hume begins the argument from elimination by considering the senses. He points out that since the idea of continued existence is the idea of continuing to exist when not sensed, the senses could never give us this idea; they would have to have an impression of something absent to them in order to do so, which is a contradiction in terms: To begin with the senses, ’tis evident these faculties are incapable of giving rise to the notion of the continu’d existence of their objects, after they no longer appear to the senses. For that is a contradiction in terms, and supposes that the senses continue to operate, even after they have ceas’d all manner of operation. [1.4.2.3, 188]

Hume then turns to the question of independent existence, a somewhat more difficult question. Hume argues that in order for the senses to be the source of the idea of the distinctness of their objects from the self, both the self and the object would have to be present to the senses as distinct entities: Now if the senses presented our impressions as external to, and independent of ourselves, both the objects and ourselves must be obvious to our senses, otherwise they cou’d not be compar’d by these faculties. The difficulty, then, is how far we are ourselves the objects of our senses. [1.4.2.5, 189]

Hume begins with a quick dismissal, grounded in the idea that the question of personal identity is a very hard one, too hard for the answer to be evident to the senses. If it is impossible for the senses to answer that question, it must be that they cannot take the self as an object and so cannot possibly determine whether the self is different from any other object. ’Tis certain there is no question in philosophy more abstruse than that concerning identity, and the nature of the uniting principle, which constitutes a person. . . . ’Tis absurd, therefore, to imagine the senses can ever distinguish betwixt ourselves and external objects. [1.4.2.6, 189–​90]

This is a manifestly terrible argument—​after all, something can be the object of the senses without their providing a clear idea of its identity—​and Hume himself immediately supplements it with a more careful one. Hume reposes the question, not as a question about whether it is possible for the senses to be the source of an idea of independent existence, but instead as a question of whether in fact we do construct that idea based on sensory information:

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 177 )

But not to lose time in examining, whether ’tis possible for our senses to deceive us, and represent our perceptions as distinct from ourselves, that is as external to and independent of us; let us consider whether they really do so, and whether this error proceeds from an immediate sensation, or from some other causes. [1.4.2.8, 190]

Hume begins by noting that at first glance it seems perfectly obvious that the senses do deliver the idea of external existence. After all, even our own body is perceived as an object in the external world, located in space, and other objects of the senses are experienced in space as related to our body, in determinate positions, and so outside of the mind: To begin with the question concerning external existence, it may perhaps be said . . . our own body evidently belongs to us; and as several impressions appear exterior to the body, we suppose them also exterior to ourselves. The paper . . . is beyond my hand. The table is beyond the paper. The walls of the chamber beyond the table. And in casting my eye towards the window, I perceive a great extent of fields and buildings beyond my chamber. From all this it may be infer’d, that no other faculty is requir’d, beside the senses, to convince us of the external existence of body. [1.4.2.9, 190–​91]

But Hume introduces this possibility only to refute it.6 He offers three distinct reasons to reject the idea that our senses actually convey the idea of external existence:  first, our senses deliver immediately only sense impressions; second, those sense impressions are clearly internal to the mind. Third, as Berkeley points out in the Dialogues, to conclude that external objects are the causes of impressions would require an inference, and not an obvious one. Hume puts the point this way: But to prevent this inference, we need only weigh the three following considerations. First, That, properly speaking, ’tis not our body we perceive,

6. Wright (1983) correctly notes that Hume follows Malebranche (and not Berkeley) in drawing a distinction between sense perceptions and external objects in this passage, and concludes from that similarity that he follows Malebranche in arguing directly for the reality of external objects. But this is not correct, for two reasons. First, as per the Cover Principle, and as I argued at the top of this section, Hume is not even asking a question about the existence of external objects; he is asking what causes us to believe in external objects. Second, Hume does not endorse the argument that our senses give us knowledge of the external world, but rejects it. So, even if he follows Malebranche in drawing a conceptual distinction between a sensation and an external object, there is no reason to think that he follows him in arguing that we can actually know that there are external objects.

( 178 )  Books I and III when we regard our limbs and members, but certain impressions, which enter by the senses; so that the ascribing a real and corporeal existence to these impressions, or to their objects, is an act of the mind as difficult to explain, as that which we examine at present. Secondly, Sounds, and tastes, and smells, tho’ commonly regarded by the mind as continu’d independent qualities, appear not to have any existence in extension, and consequently cannot appear to the senses as situated externally to the body . . . . Thirdly, Even our sight informs us not of distance or outness (so to speak) immediately and without a certain reasoning and experience. [1.4.2.9, 191]

Hume draws all of this together in the following summary of the argument: Thus to resume what I have said concerning the senses; they give us no notion of continu’d existence, because they cannot operate beyond the extent, in which they really operate. They as little produce the opinion of a distinct existence, because they can neither offer it to the mind as represented, nor as original. To offer it as represented, they must present both an object and an image. To make it appear as original, they must convey a falshood; And this falshood must lie in the relations and situation: In order to which they must be able to compare the object with ourselves; and even in that case they do not, nor is it possible that they shou’d, deceive us. We may, therefore, conclude with certainty, that the opinion of a continu’d and of a distinct existence never arises from the senses. [1.4.2.11, 191–​92]

The senses cannot be the source of the idea of external existence because they cannot produce an idea of continued existence in virtue of their intermittent contact with their objects. Following Berkeley once again, Hume points out that they cannot present ideas both of their own impressions and of the objects represented as distinct presentations to be compared with one another. This is simply because the only access the senses have to external objects could be through their own impressions; there is no independent direct channel to enable a comparison. Nor can they simply generate an original idea of externality, as that would require some impression simply of externality as such, of which there is none. The senses are hence ruled out as the source of this idea. Hume then turns to the faculty of reason.

3.  REASON AND THE EXTERNAL WORLD The argument against reason being the source of the idea of external existence is more compact. First, he notes, nobody ever really argues that there

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 179 )

is an external world, and lots of folks who wouldn’t even know what to do with an argument, or who are totally irrational, nonetheless believe in the external world. We may also observe in this instance of sounds and colours, that we can attribute a distinct continu’d existence to objects without ever consulting reason, or weighing our opinions by any philosophical principles. And indeed, whatever convincing arguments philosophers may fancy they can produce to establish the belief of objects independent of the mind, ’tis obvious these arguments are known but to very few, and that ’tis not by them, that children, peasants, and the greatest part of mankind are induc’d to attribute objects to some impressions, and to deny them to others. [1.4.2.14, 193]

Second, ordinary people actually confuse their perceptions—​which are interrupted and internal to the mind—​with external objects, thinking that they see external objects directly. Thus, they think that their very perceptions are external. This, Hume avers, is manifestly irrational, and so cannot be the work of the faculty of reason: Philosophy informs us, that every thing, which appears to the mind, is nothing but a perception, and is interrupted, and dependent on the mind; whereas the vulgar confound perceptions and objects, and attribute a distinct continu’d existence to the very things they feel or see. This sentiment, then, as it is entirely unreasonable, must proceed from some other faculty than the understanding. [1.4.2.14, 193]

So, Hume concludes by elimination that—​since neither the senses nor reason can be the source of our belief in the external world—​the only faculty left can be the imagination. Just as the imagination, we saw, was the faculty responsible for the genesis of our idea of the self, and of all of the essential passions connected to the self, it will turn out to be the seat of the belief in the world around us: So that upon the whole our reason neither does, nor is it possible it ever shou’d, upon any supposition, give us an assurance of the continu’d and distinct existence of body. That opinion must be entirely owing to the imagination: which must now be the subject of our enquiry. [1.4.2.14, 193]

4.  THE IMAGINATION AND THE EXTERNAL WORLD All of this has been by way of brush clearing prior to the real work of demonstrating how we come to believe in an external world. If we follow

( 180 )  Books I and III

this complex chain of reasoning, we will see all of the strategies that underpin Hume’s project in play. Hume begins by surveying the properties common to those things we regard as external. First, he notes, they seem to stay pretty much the same despite interrupted attention: After a little examination, we shall find, that all those objects, to which we attribute a continu’d existence, have a peculiar constancy, which distinguishes them from the impressions, whose existence depends upon our perception. These mountains, and houses, and trees, which lie at present under my eye, have always appear’d to me in the same order; and when I lose sight of them by shutting my eyes or turning my head, I soon after find them return upon me without the least alteration. [1.4.2.18, 194]

But this constancy is not complete. Things do change over time. They move; they age; they transform. Last night’s milk becomes this morning’s yogurt. Yesterday’s chrysalis becomes today’s butterfly. Despite this change, Hume points out, those things we regard as external change in an orderly fashion. There is, as he puts it, a coherence to their successive states, unlike the states of internal phenomena, such as sensations, or dreams: This constancy, however, is not so perfect as not to admit of very considerable exceptions. Bodies often change their position and qualities, and after a little absence or interruption may become hardly knowable. But here ’tis observable, that even in these changes they preserve a coherence, and have a regular dependence on each other; which is the foundation of a kind of reasoning from causation, and produces the opinion of their continu’d existence. [1.4.2.19, 195]

This, then, is the initial hypothesis regarding the cause of our belief in the external world:  some of our impressions—​those deriving from the senses—​are more constant and coherent than others, or at least constant and coherent in a different manner from others.7 As Hume notes in the next passage, our passions and other obviously internal states also exhibit some constancy and coherence: our inner lives are not characterized by random dramatic transformations from state to state. Nonetheless, Hume argues, inference to the best explanation yields a different account of the reason for the constancy of our sense perceptions from that it delivers for the coherence 7. Wright (1983) notes correctly in this context that Hume does not follow Berkeley or Malebranche in asserting that the involuntariness of perception is among the causes of our taking them to be caused by external objects, on the grounds that our passions are similarly involuntarily, and we do not therefore take them to arise immediately from external causes.

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 181 )

of our sense impressions. The external world is the best possible explanation for the degree and manner of the coherence of our sense impressions: Having found that the opinion of the continu’d existence of body depends on the coherence and constancy of certain impressions, I now proceed to examine after what manner these qualities give rise to so extraordinary an opinion. To begin with the coherence; we may observe, that tho’ those internal impressions, which we regard as fleeting and perishing, have also a certain coherence or regularity in their appearances, yet ’tis of somewhat a different nature, from that which we discover in bodies. Our passions are found by experience to have a mutual connexion with and dependance on each other; but on no occasion is it necessary to suppose, that they have existed and operated, when they were not perceiv’d, in order to preserve the same dependance and connexion, of which we have had experience. The case is not the same with external objects. Those require a continu’d existence, or otherwise lose, in a great measure, the regularity of their operation. [1.4.2.20, 195–​96]

Hume confirms this hypothesis in a long rumination in which he discusses his awareness of his study, his house, a porter who brings him a letter, the letter that tells him of the happenings in a far-​off land, the view from his window, the gradually changing fire in the grate, and so on, emphasizing that these impressions are not disconnected, but rather form a systematic whole, and that the best explanation for their regularity and systematicity is that they are caused by a reasonably stable world governed by regular principles [1.4.2.20, 196–​97]. This both delivers the conclusion that the imagination is the source of our belief in the external world and an account of the mechanism by means of which it does so. He concludes: There is scarce a moment of my life, wherein there is not a similar instance presented to me, and I have not occasion to suppose the continu’d existence of objects, in order to connect their past and present appearances, and give them such a union with each other, as I have found by experience to be suitable to their particular natures and circumstances. Here then I am naturally led to regard the world, as something real and durable, and as preserving its existence, even when it is no longer present to my perception. [1.4.2.20, 197]

5.  WHY BELIEVE? We might think that this is the end of the matter. But there is a rub. The account here, which seems so natural, is in terms of inference to the best

( 182 )  Books I and III

explanation. So now, if we are concerned to know whether this mechanism by means of which we are caused to believe in the external world also provides a justification for that belief, we must ask, is this inference any good? Hume worries that it is not. There are two issues intertwined here that Hume does not separate clearly, but each will be of importance in what follows, and so it is good to disentangle them here. One concerns the negative side of the skeptical argument, and the other the positive side. On the negative side, we need to worry about justification. Does the inference to the best explanation justify our belief in the external world? If not—​and Hume will argue that it does not—​we discover that just as we had no rational justification for reliance on reason, we will find that we have no rational justification of our belief in the external world. That belief and the practices it supports—​a belief and practices that we must remember Hume takes for granted in this entire discussion—​will have to be explained in another way. On the positive side, we need a causal explanation for that belief. Constancy and coherence might be expected to do that work, even if they cannot do the justificatory work undermined by the skeptical argument. Here Hume argues that although they can never do the justificatory work, they may suffice to do the causal work. So, as in the case of skepticism with regard to reason, we will end up with an account of a belief that cannot be justified, but which can be explained, and an inversion of explanatory order in which the custom that one might have thought would be rationally explained, will in fact itself explain the inferences we actually draw. That is, in lieu of a justification of custom, which might seem to be demanded, we will get a justification by custom, which will turn out to be all the justification we could ever in fact demand. In this regard, this discussion is of a piece with those we have already encountered in the context of our discussion of causation and of skepticism with regard to reason.8 But, as Hume notes here, the argument in this case will be considerably more complex, and custom will have a slightly different role to play in explaining our confidence in the existence of the external world, in virtue of the fact that the customs in question are highly constructive. As we will see, custom

8. In this context, it is worth remarking another instance of uniformity of method: In the case of the so-​called “naturalistic fallacy” Hume is credited with discovering, many take the moral of the argument to be that one can never derive normative conclusions from descriptive premises. But that is not quite right. Hume argues in that context that we cannot derive normative conclusions from premises that do not already have normative content. But statements about established custom do have normative content, and that normativity, according to Hume, arises precisely from custom. So, once again, we encounter not justification of custom, but rather justification by custom.

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 183 )

enters the picture in slightly different ways in this argument, and the skeptical argument takes some surprising turns. But tho’ this conclusion from the coherence of appearances may seem to be of the same nature with our reasonings concerning causes and effects; as being deriv’d from custom, and regulated by past experience; we shall find upon examination, that they are at the bottom considerably different from each other, and that this inference arises from the understanding, and from custom in an indirect and oblique manner. For ’twill readily be allow’d, that since nothing is ever really present to the mind, besides its own perceptions, ’tis not only impossible, that any habit shou’d ever be acquir’d otherwise than by the regular succession of these perceptions, but also that any habit shou’d ever exceed that degree of regularity. Any degree, therefore, of regularity in our perceptions, can never be a foundation for us to infer a greater degree of regularity in some objects, which are not perceiv’d. . . . But ’tis evident, that whenever we infer the continu’d existence of the objects of sense from their coherence, and the frequency of their union, ’tis in order to bestow on the objects a greater regularity than what is observ’d in our mere perceptions. We remark a connexion betwixt two kinds of objects in their past appearance to the senses, but are not able to observe this connexion to be perfectly constant, since the turning about of our head, or the shutting of our eyes is able to break it. [1.4.2.21, 197–​98]

Note that while Hume begins by talking about our “reasoning concerning causes and effects,” he moves quickly to talk about the acquisition of habits, before returning to reasoning. The issues are intertwined, but are best kept apart analytically. One is relevant to the skeptical problem, the other to the skeptical solution to that problem. The first thing to note then is that the major premise of this new argument is that no argument from constancy and coherence can justify the ascription of greater constancy and coherence to the explanandum than we find in the explanans; and, were this an argument for the existence of the external world, it would do just that. Hume concludes: But as all reasoning concerning matters of fact arises only from custom, and custom can only be the effect of repeated perceptions, the extending of custom and reasoning beyond the perceptions can never be the direct and natural effect of the constant repetition and connexion, but must arise from the co-​operation of some other principles. [1.4.2.21, 198]

He now returns to the causal side of the story, pointing out once again that coherence will not suffice to explain our confidence in the existence of the

( 184 )  Books I and III

external world; constancy will be necessary as well, but it will be necessary in a causal story about the origin of our psychological compulsion to take the world for granted, not a justificatory story of the rational legitimacy of that belief. But whatever force we may ascribe to this principle, I am afraid ’tis too weak to support alone so vast an edifice, as is that of the continu’d existence of all external bodies; and that we must join the constancy of their appearance to the coherence, in order to give a satisfactory account of that opinion. [1.4.2.23, 198–​99]

Hume begins with a sketch of the explanation, but this is only a sketch that sets up more problems than it solves. Nonetheless, it does represent the outline of the causal explanation of our propensity to believe in the external world: When we have become accustom’d to observe a constancy in certain impressions, and have found, that the perception of the sun or ocean, for instance, returns upon us after an absence or annihilation with like parts and a like order, as at its first appearance, we are not apt to regard these interrupted perceptions as different, (which they really are) but on the contrary consider them as individually the same, upon account of their resemblance. But as this interruption of their existence is contrary to their perfect identity, and makes us regard the first impression as annihilated, and the second as newly created, we find ourselves somewhat at a loss, and are involv’d in a kind of contradiction. In order to free ourselves from this difficulty, we disguise, as much as possible, the interruption, or rather remove it entirely, by supposing that these interrupted perceptions are connected by real existence, of which we are insensible. This supposition, or idea of continu’d existence, acquires a force and vivacity from the memory of these broken impressions, and from that propensity, which they give us, to suppose them the same; and according to the precedent reasoning, the very essence of belief consists in the force and vivacity of the conception. [1.4.2.24,199]

Let us note the principal features of this system, as Hume next takes care to explain it in great detail, and it is easy to lose sight of this forest for the trees. First, Hume is mobilizing his distinction between two senses of identity—​a distinction to which he returns frequently in the Treatise—​ strict numerical identity and specific identity, or identity of kind, that is, resemblance. The chief executive of the United States of America is numerically identical to the commander in chief of the armed forces of the United States, but one of a set of identical twins is only specifically identical to the

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 185 )

other. Just as in the case of personal identity, Hume suggests that we often confuse these two senses of identity, casually attributing numerical identity to two events or stages of a thing that are only specifically identical to one another. So, Hume notes in the passage just quoted, we notice that our perceptions of external objects are specifically identical to one another, but treat them as though they were numerically identical.9 In order to make sense of this transition, we posit external objects that remain numerically identical over time. We really believe in this because of the “force and vivacity,” that is, the causal potency, of the memories of specifically identical perceptions that combine to produce a very vivid idea of a single numerically identical phenomenon. Whether or not this is good psychology, it is psychology, not epistemology. We are firmly on the causal-​explanatory side of the skeptical program, not on the justificatory or critical side. But Hume recognizes that it is not enough to assert this explanation; the explanation itself must be justified, even if that explanation is not meant to constitute a justification of our belief in the reality of the external world: In order to justify this system, there are four things requisite. First, To explain the principium individuationis, or principle of identity. Secondly, Give a reason, why the resemblance of our broken and interrupted perceptions induces us to attribute an identity to them. Thirdly, Account for that propensity, which this illusion gives, to unite these broken appearances by a continu’d existence. Fourthly and lastly, Explain that force and vivacity of conception, which arises from the propensity. [1.4.2.25, 199–​200]

That is, we need to understand the kind of identity we attribute to external objects, explain how we come to attribute this kind of identity to them, then explain why that kind of identity leads to positing continued existence, and finally, explain why we actually believe in all of this to the degree that we do, which is a very considerable degree. Hume is setting out to explain one of the most profound aspects of human nature—​our taking for granted the existence of the world around us despite having no evidence whatsoever that it exists. 9.  Pears (1990) misses this point when he complains that “Hume’s intolerant downgrading of the identities of most material objects is obviously questionable and it has been criticized by many commentators. How can what we count as identity possibly fail to be identity?” (125). Hume’s point is that we count things as identical when they are merely specifically identical, but often conflate numerical and specific identity and slip into counting things as numerically identical when they are merely specifically identical.

( 186 )  Books I and III

In this investigation, he is effectively taking up Berkeley’s challenge and answering him by using the classical skeptical tradition on which Berkeley depends to pose the challenge. (He is also directly attacking Locke’s own account of our knowledge of the external world.) Kant was later to take a transcendental route to proving the empirical existence of objects in space; Hume instead takes a naturalistic route to explaining our conviction in the existence of objects in space, despite the fact that their spatiotemporal existence cannot be proven, once again joining his skepticism and naturalism through appeal to custom, this time in its guise as psychological instinct. After Hume notes that multiplicity can never give rise to an idea of identity [1.4.2.26–​29, 200–​220], he introduces the idea of time as something that can mediate the relation between multiplicity and identity: Here then is an idea, which is a medium betwixt unity and number; or more properly speaking, is either of them, according to the view, in which we take it: And this idea we call that of identity. We cannot, in any propriety of speech, say, that an object is the same with itself, unless we mean, that the object existent at one time is the same with itself existent at another. By this means we make a difference, betwixt the idea meant by the word, object, and that meant by itself, without going the length of number, and at the same time without restraining ourselves to a strict and absolute unity. Thus the principle of individuation is nothing but the invariableness and uninterruptedness of any object, thro’ a suppos’d variation of time, by which the mind can trace it in the different periods of its existence, without any break of the view, and without being oblig’d to form the idea of multiplicity or number. [1.4.2.29–​30, 201]

This completes the first task, filling in an idea that was sketched previously. The kind of identity we ascribe to objects when we say that they persist through time—​that they are constant, in the language of this section—​is that they remain uninterrupted through change, that there is a continuity uniting their distinct stages. This is not numerical identity, but a kind of specific identity; on the other hand, it is not the specific identity between two discontinuous objects, like a pair of identical twins, but the specific identity between the stages of a continuous causal sequence.10 Hume now

10. Fogelin (1985) argues that Hume’s introduction of identity as a new idea is illicit, given that the only ideas in the neighborhood here are unity and number, and that Hume is not entitled by his own lights to introduce a new idea without a corresponding impression. But Hume is careful here:  we start with similarity and move to specific identity, which is a kind of similarity relation, and then succeed in confusing specific

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 187 )

turns to the second question, that of why the observation of this kind of specific identity leads us to treat successive stages of objects as numerically identical despite their difference: I now proceed to explain the second part of my system, and shew why the constancy of our perceptions makes us ascribe to them a perfect numerical identity, tho’ there be very long intervals betwixt their appearance, and they have only one of the essential qualities of identity, viz., invariableness. That I may avoid all ambiguity and confusion on this head, I shall observe, that I here account for the opinions and belief of the vulgar with regard to the existence of body; and therefore must entirely conform myself to their manner of thinking and of expressing themselves. [1.4.2.31, 201–​2]

Hume emphasizes that by conforming himself to the “opinions and beliefs of the vulgar” he is explaining, not justifying, our ordinary practice, and that he is not even considering our theoretical practice. The question is not about how we deduce that later stages of objects are numerically identical to earlier stages; that would be impossible. Instead, the question concerns how we are caused to treat later stages as numerically identical to earlier stages although reflection would tell us that they are not. We are explaining a custom, not justifying a judgment. Hume begins by asking whether we

identity with the numerical identity that each thing has with itself. This is hence not the creation of a new idea out of unity and number, but the realization that we are already in possession of an idea that is intermediate between these two. And here I  disagree with Baxter. He argues that Hume’s account requires that we have an inconsistent idea of an object when we view it as identical over time, and so that, properly speaking, there is no “medium betwixt unity and number” to constitute identity. Rather, he suggests, “The fiction of an object being represented as being both enduring and steadfast requires switching back and forth between viewing the object as many and viewing it as one. Since there cannot be an idea literally of a medium between singleness and plurality, this mongrel, view-​dependent idea serves well enough. The switching of views masks the inconsistency” (2008, 63). Baxter is driven to this account because of the emphasis he places on the role of what he calls “steadfast objects” in Hume’s understanding of our idea of the external world. Steadfast objects, on Baxter’s view, endure through change and themselves are not successions of temporal parts. To be sure, if one thought—​as Baxter does—​that Hume is attempting to account for such an idea, there would be a manifest inconsistency to resolve. On that view, Hume would be trying to show why we have an idea of the numerical identity of objects over time, and if one were to try to account for such an idea, Baxter’s would be the best approach. But on my view, there is no inconsistency to mask. There is an idea “medium betwixt unity and number,” viz., that of specific identity. And that strikes me as a good independent argument for this reading. (I note in passing that Baxter’s earlier reading [2006], emphasizing the Pyrrhonism of Hume’s account, is much closer to my own.)

( 188 )  Books I and III

have any easy examples of a similar pattern of thought, and answers in the affirmative: Now what other objects, beside identical ones, are capable of placing the mind in the same disposition, when it considers them, and of causing the same uninterrupted passage of the imagination from one idea to another? This question is of the last importance. For if we can find any such objects, we may certainly conclude, from the foregoing principle, that they are very naturally confounded with identical ones, and are taken for them in most of our reasonings. [1.4.2.34, 203]

The relevant example is a succession of closely related experiences of objects. Hume claims that when we entertain very closely connected perceptions of a single object, as when we fix our gaze on a thing over time, the successive perceptions are so similar to one another that our cognitive state does not change perceptibly. That feeling that we remain in the same cognitive state leads us to ascribe an identity to the objects of the successive states. I immediately reply, that a succession of related objects places the mind in this disposition, and is consider’d with the same smooth and uninterrupted progress of the imagination, as attends the view of the same invariable object. The very nature and essence of relation is to connect our ideas with each other, and upon the appearance of one, to facilitate the transition to its correlative. The passage betwixt related ideas is, therefore, so smooth and easy, that it produces little alteration on the mind, and seems like the continuation of the same action; and as the continuation of the same action is an effect of the continu’d view of the same object, and ’tis for this reason we attribute sameness to every succession of related objects. The thought slides along the succession with equal facility, as if it consider’d only one object; and therefore confounds the succession with the identity. [1.4.2.34, 204]

Hume immediately calls our attention back to the structure of the argument, which is not at all easy to keep in mind! (He is now in the third part of the four-​part explanation of how it is that the constancy and coherence of perceptual impressions cause us to believe in the existence of the external world—​the explanation of how we manage to move from an awareness of discontinuous perceptions to a belief in continuous objects.) We are not, he reminds us, asking whether the external world exists, or even whether we believe that it does, but rather what the causes are of our believing in it. Note as well the remarkable parallel between this account and Hume’s discussion of the origins of the idea of necessary connection. In each case, the mind “spreads itself over the world,” and we confuse a characteristic

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 189 )

of our cognitive processes—​here the “smooth and easy” passage between ideas, in the case of causality the “determination to carry our thought from one object to the other”—​with something we perceive in the external world. In the case of causality, that was a real causal nexus; in the present case, it is continued existence. There is, however, an important disanalogy to bear in mind:  while in the case of necessary connection, the idea we turn out to have is one of a determination of the mind, Hume does not argue that the missing idea of continued existence is in fact the idea of the easy transition. The easy transition plays a different explanatory role in this context than that played by the determination of the mind in the context of the discussion of causality, this time explaining how we confuse one sense of identity with another. Nonetheless, the overarching strategy is the same. Hume now stirs in another ingredient to his characterization of the minds of the vulgar (where, he points out, the vulgar comprise philosophers as well, when they are not occupied with explicit philosophical reflection). That is, when we perceive, we tend not to distinguish our perceptions from the objects we take ourselves to perceive. When I gaze at the peony in the vase on my desk as I type these words, I take myself to be seeing the flower directly; I take the object of the perception of the flower to endure, even though I  know my perceptions to be fleeting. I  therefore take whatever I see (which, again, I know to be only a perception) to exist when it is unperceived and I take external objects (which, again, I know not to be the immediate objects of perception) to be immediately present to my mind.11 We may begin with observing, that the difficulty in the present case is not concerning the matter of fact, or whether the mind forms such a conclusion concerning the continu’d existence of its perceptions, but only concerning the manner in which the conclusion is form’d, and principles from which it is deriv’d. ’Tis certain, that almost all mankind and even philosophers themselves, for the greatest part of their lives, take their perceptions to be their only objects, and suppose, that the very being, which is intimately present to the mind, is

11.  Wright (1983) observes that the discussion of “smooth and easy” transitions probably owes a great deal to Malebranche’s psychophysiology, with its accounts of channels in the brain created by the regular flow of animal spirits. We might also note the resemblance between this account of unreflective perceptual experience and that we find in the Yogācāra Buddhist tradition, particularly in texts such as the Saṃdhinirmocana-​sūtra (The Discourse Untangling the Thought) and Vasubandhu’s Trisvabhāvanirdeśa (Treatise on the Three Natures). In each of those contexts we find assertions that whereas we are immediately aware only of states of our own consciousness, we unreflectively regard those states as external objects. See Garfield (2015).

( 190 )  Books I and III the real body or material existence. ’Tis also certain, that this very perception or object is suppos’d to have a continued uninterrupted being, and neither to be annihilated by our absence, nor to be brought into existence by our presence. When we are absent from it, we say it still exists, but that we do not feel, we do not see it. When we are present, we say we feel, or see it. Here then may arise two questions; First, How we can satisfy ourselves in supposing a perception to be absent from the mind without being annihilated. Secondly, After what manner we conceive an object to become present to the mind, without some new creation of a perception or image; and what we mean by this seeing, and feeling, and perceiving. [1.4.2.38, 206–​7]

Hume asks how we come to these obviously false views. But, we might also ask, why does Hume ask these questions at this point in the inquiry? The answer is that he is focusing ever more narrowly on the point in the psychological process where an awareness of our own inner states causes a belief in an external world. We are aware, in the case of some successions of perceptions, that those perceptions are very similar to one another (that they are specifically identical) and that our cognitive states directed upon these impressions are also specifically identical to one another; somehow this causes us to believe that the objects we perceive endure when we are not perceiving them. This in turn requires that we somehow make a transition from believing in our perceptions to believing in objects; but the fact that we don’t distinguish perceptions from objects means that we need instead to understand how we make a transition from taking ourselves to perceive something fleeting to taking ourselves to have a fleeting perception of something permanent. As to the first question;  .  .  .  Now as every perception is distinguishable from another, and may be consider’d as separately existent; it evidently follows, that there is no absurdity in separating any particular perception from the mind; that is, in breaking off all its relations, with that connected mass of perceptions, which constitute a thinking being. [1.4.2.39, 207]

Hume’s answer to the first question—​that regarding how we can imagine a perception to be outside of a mind and yet existent—​is a straightforward application of his principle that if two objects can be imagined to exist independently then they are metaphysically independent (what Garrett has called the “separability principle”). While this may be questionable (and, as we have seen, it may be the most important principle that Hume defends to which he is not, on his own terms, entitled), it is an operative principle in

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 191 )

the Treatise. Since each perception is logically independent of every other, we can imagine any perception detached from the bundle that constitutes our mind and existing independently. One might react to this with a bit of shock: independence or no independence, one might think, perceptions can only be the perceptions of minds, and it is simply incoherent to imagine a perception perceived by no mind at all. We should not be so hasty in ascribing such an error to Hume. Note that this exercise in imagination trades directly on the confusion Hume ascribes to us between perceptions and objects: while it might seem odd to think of a perception existing outside of a mind, it is not so odd to think of an object existing outside of a mind. To the extent that we confuse our perceptions with objects, we can easily simply ascribe independence to whatever it is we see, and allow the confusion to take care of the rest. And this confusion is also at work in Hume’s answer to the second question—​ that concerning what we mean when we talk about seeing objects, as opposed to perceptions: The same reasoning affords us an answer to the second question. If the name of perception renders not this separation from a mind absurd and contradictory, the name of object, standing for the very same thing, can never render their conjunction impossible. External objects are seen, and felt, and become present to the mind; that is, they acquire such a relation to a connected heap of perceptions, as to influence them very considerably in augmenting their number by present reflexions and passions, and in storing the memory with ideas. The same continu’d and uninterrupted Being may, therefore, be sometimes present to the mind, and sometimes absent from it, without any real or essential change in the Being itself. An interrupted appearance to the senses implies not necessarily an interruption in the existence. The supposition of the continu’d existence of sensible objects or perceptions involves no contradiction. [1.4.2.40, 207–​8]

Here, the confusion operates in the opposite direction. In the first case, we allowed our instinctive view that objects can exist independently of the mind to lead us to believe that perceptions could do so as well. In this case, our sense that perceptions are directly perceived causes us, via the same confusion, to believe that external objects are directly perceived. This is the third part of the explanation of how we come to believe in the external world:  we confuse our perceptions with the objects they present to us. Our belief in the former leads to a belief in the latter. But now comes the hard part. We need to explain how this confusion leads to a persistent and

( 192 )  Books I and III

powerful belief in the external world, and not, say, to a mere conjecture or a fleeting suspicion: But as we here not only feign but believe this continu’d existence, the question is, from whence arises such a belief; and this question leads us to the fourth member of the system. It has been prov’d already, that belief in general consists in nothing, but the vivacity of an idea; and that an idea may acquire this vivacity by its relation to some present impression. Impressions are naturally the most vivid perceptions of the mind; and this quality is in part convey’d by the relation to every connected idea. The relation causes a smooth passage from the impression to the idea, and even give the propensity to that passage. The mind falls so easily from the one perception to the other, that it scarce perceives the change, but retains in the second a considerable share of the vivacity of the first. . . . . . . Our memory presents us with a vast number of instances of perceptions perfectly resembling each other, that return at different distances of time, and after considerable interruptions. This resemblance gives us a propension to consider these interrupted perceptions as the same; and also a propension to connect them by a continu’d existence, in order to justify this identity, and avoid the contradiction, in which the interrupted appearance of these perceptions seems necessarily to involve us. Here then we have a propensity to feign the continu’d existence of all sensible objects; and as this propensity arises from some lively impressions of the memory, it bestows a vivacity on that fiction; or in other words, makes us believe the continu’d existence of body. If sometimes we ascribe a continu’d existence to objects, which are perfectly new to us, and of whose constancy and coherence we have no experience, ’tis because the manner, in which they present themselves to our senses, resembles that of constant and coherent objects; and this resemblance is a source of reasoning and analogy, and leads us to attribute the same qualities to the similar objects. [1.4.2.41–​42, 208–​9]

The explanation here is simple. Our confidence in the reality of the external world is achieved because the disposition to take for granted the continued and independent existence of objects, effected by the confusion diagnosed in the third part of the explanation, acquires sufficient force and vivacity to rise to the level of belief. It acquires this force simply from the massive number of times we experience similar perceptions and take them to be identical. As Malebranche (1879) would put the point, the relevant channels in our brains are deepened and the spirits flow more readily through them (see 1:150–​51). To put the point in the terms of the more recent behaviorism of Clark Hull (1943), the more the mind takes two similar perceptions to be identical, the greater the habit strength that response attains. When it

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 193 )

becomes sufficiently habitual to attribute that identity, we do so as a cognitive reflex whenever we encounter sufficiently similar perceptions in sufficiently coherent frameworks. That attribution of identity in turn entails the confusion of perception and object, and we attribute that identity to objects, hence believing in the external world. So much for an explanation of the tendency of the vulgar. But this gives rise to a new tension. We have a skeptical argument that we are never justified in believing that there is an external world, and a naturalistic explanation of our custom of believing in an external world.12 In one sense, given Hume’s Pyrrhonian method, this is no big deal. Custom, not argument, is what justifies our taking the external world for granted. In this sense, Hume’s treatment of skepticism with regard to the senses need be no different from his treatment of skepticism with regard to reason. But in another sense, a big problem remains, and solving this problem constitutes the second half of the project of this section. For we have not really accounted for the origin of the idea of externality, of objects. We have accounted for the propensity to believe in the external world, if that is merely, as Hume puts it, to take it for granted, and to treat things as constant, coherent, and independent of our minds. That will do for the vulgar. But as philosophers, we have been talking of objects, of a real external world—​even if it is one of whose existence we cannot be certain—​ with which our ideas are confused. What are we even talking about when we do this? If Hume’s analysis is correct, it would appear not only that the problem of how we come to believe in an external world is insoluble, but also that, without an idea of externality, we cannot even formulate it, and that in even asking this question we are talking nonsense. This is a far deeper skeptical abyss than we encountered in the discussion of skepticism with regard to reason. As we examine Hume’s treatment of the philosophical system, we are therefore brought face to face with another of our fundamental interpretative principles: Hume often intends to show us that we don’t have the ideas we think we do—​that often we do not even know what we are

12. Green (1968) argues that at this point Hume simply begs the question. He writes, “With ‘the vulgar’ that which is ‘immediately perceived’ is the real thing, just because it is not the mere feeling which with Hume it is. But under pretence of provisionally adopting the vulgar view, he entitles himself to treat the mere feeling, because according to him it is that which is immediately perceived, as if it were the permanent identical thing, which according to the vulgar is what is immediately perceived” (257). If Hume’s task were to provide an analysis of an idea of external existence, Green might well be right. But in fact Hume is instead arguing that we have no such idea. On this reading, the charge of petitio principii leveled by Green is unjustified.

( 194 )  Books I and III

talking about. And in this case, that conclusion—​that we do not even have an idea of externality, and so cannot even cogently ask the questions we find ourselves trying to answer in this section—​is particularly disturbing.13

6.  TAKING THIS SERIOUSLY It is at this point that Hume contrasts the philosophical and the vulgar systems. The vulgar we have just seen. The philosophical system is an attempt to make sense of the claim that the external world exists and that we know it through our senses. Hume had Locke in mind here.14 The philosophical system is of interest not because it might be correct, nor even because it demands refutation, but rather simply because it is apparently cogent. Whether it is correct or not, if we can articulate the philosophical system, then we can make sense of the very idea of an external world; and if that is the case, there is a genuine idea of externality that demands explanation. If, on the other hand, we cannot articulate it, we may find that even asking what causes us to believe in the external world may be an exercise in nonsense. Moreover, the cogency of the philosophical system is apparently a precondition of our very articulation and critique of the system of the vulgar. 13. Wright, in commenting on this discussion, states that “Hume draws the conclusion that we are aware only of mind-​dependent images. But he also goes on to draw the conclusion that these images are ‘nothing but . . . representations of other existences, which remain uniform and independent’ [EHU 12.8; SBN 118]” (1983, 51). He argues on this basis that Hume is defending an indirect realist theory of perception. But Wright misses both the point of the remark he quotes and the structure of the argument in which it figures. The claim that our perceptions are representations of external objects is not the claim that there are genuine external objects of which our perceptions are copies or effects—​this is not even the topic of Hume’s discussion; rather it is the claim that we take our perceptions to be representations of external objects. It is our so taking them that is the phenomenon to be explained, not any actual representational relation between them and any external phenomena, and it is explained without any reference to any actual causal or representational relations to any objects claimed by Hume to exist externally to the mind (although again, they are always taken to so exist). 14.  We might note another nice parallel between this argument and another argument in the Treatise also directed against Locke, Hume’s attack on the social contract as a foundation for social obligation. Hume argues in that context that Locke presupposes the very phenomenon he is trying to explain: if there were not antecedent obligation to keep our promises and to observe contracts, no social contract could be binding in the first place. In this context, Hume is considering Locke’s defense of our knowledge of the external world as grounded in our sensory experience. The task is to explain the origin of the idea of external objects. But in order even to infer them as the causes of our experiences, we would already have to have the idea in place. The question, on Hume’s analysis, is begged in exactly the same way.

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 195 )

For if the philosophical system is not at least cogent, we cannot even characterize the confusions in the vulgar system. And Hume argues, it is not cogent: it suffers from all of the defects of the vulgar system and then some. That is, with the vulgar, it assumes a double existence, but as philosophical, it recognizes the incoherence of that position, and, in reflection, must also recognize the incoherence of the very terms in which the problem and the putative solution is posed, undermining its own claim to greater sophistication than the vulgar. Hume’s strategy in this portion of the argument is first to show how deep the philosophical hole is into which we fall when we try to make sense of our idea of the external world when we reflect on it, and only then to dig his way out of it. But the way out, as we shall see, does not involve relinquishing the Pyrrhonian strategy that gets us into this predicament, but rather following that strategy to its conclusion. By doing so, Hume completes the project. Having first shown that it is the imagination that is the source of our idea of the external world, then having shown that the causes of our belief in the external world cannot themselves be justificatory, and—​as we are about to see—​that we can provide a coherent justification neither of the belief in the external world nor of our doubts regarding it, Hume argues that our only entitlement to any talk or thought about the external world is custom. The naturalistic causal story of the origins of conviction in the reality of the external world hence remains in place, as well as the skeptical demonstration that nothing cognitive could ever justify that conviction; these two contrary pressures are reconciled through the Humean strategy of appeal to custom. We now turn to the first part—​the descent into skeptical despair. Hume first argues [1.4.2.48–​49, 212–​13] that the philosophical system has no “primary recommendation” to the imagination, that is, that there is no reason to believe that we should have an innate idea of externality, and that the reasoning involved in the philosophical system cannot possibly have the power to create an idea of externality in the imagination. He concludes: As to the second part of the proposition, that the philosophical system acquires all its influence on the imagination from the vulgar one; we may observe, that this is a natural and unavoidable consequence of the foregoing conclusion, that it has no primary recommendation to reason or the imagination. For as the philosophical system is found by experience to take hold of many minds, and in particular of all those, who reflect ever so little on this subject, it must derive all its authority from the vulgar system; since it has no original authority of its own. [1.4.2.49, 213]

( 196 )  Books I and III

That is, even though philosophers talk about the external world, there is nothing but the confusions of the vulgar system behind the words they use. We know that the senses can’t inform us that there is an external world; there is no innate idea of externality in the imagination; and there are no good arguments for externality; moreover, if there were, they would be too abstract to account for a near universal belief in the external world. And the vulgar system, as we have seen, is the result of confusions that we cannot possibly endorse as giving us a distinct idea of an external object. So, the question remains:  when we even talk about confusing our perceptions with external objects, or confusing the properties of constancy and coherence with that of externality, what do we mean by the very terms external object and externality? If we cannot answer these questions, not only is the philosophical system incoherent, but so is Hume’s own diagnosis of its failings. For that diagnosis seems to rely on the very ideas whose impossibility vitiates the philosophical system. Nonetheless, says Hume, philosophers blithely and uncritically continue to talk as though this makes sense: The imagination naturally runs on this train of thinking. Our perceptions are our only objects: Resembling perceptions are the same, however broken or uninterrupted in their appearance: This appearing interruption is contrary to the identity:  The interruption consequently extends not beyond the appearance, and the perception or object really continues to exist, even when absent from us: Our sensible perceptions have, therefore, a continu’d and uninterrupted existence. But as a little reflection destroys this conclusion, that our perceptions have a continu’d existence, by shewing that they have a dependent one, ’twou’d naturally be expected, that we must altogether reject the opinion, that there is such a thing in nature as a continu’d existence, which is preserv’d even when it no longer appears to the senses. The case, however, is otherwise. Philosophers are so far from rejecting the opinion of a continu’d existence upon rejecting that of the independence and continuance of our sensible perceptions, that tho’ all sects agree in the latter sentiment, the former, which is, in a manner, its necessary consequence, has been peculiar to a few extravagant sceptics; who after all maintain’d that opinion in words only, and were never able to bring themselves sincerely to believe it. [1.4.2.50, 213–​14]

The “extravagant sceptics” are not the Pyrrhonians like Hume, but those who would simply deny outright that there is an external world—​those who would not “take the existence of body for granted in all of their reasonings.” So, Hume concludes at this point, despite compelling reasons to abandon

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 197 )

the philosophical system, philosophers remain committed to it, with the exception of those few who claim—​albeit insincerely—​that they reject the existence of the external world. Nonetheless, despite this widespread support, Hume concludes, the philosophical system is in fact entirely incoherent. It requires us to maintain that our perceptions are both continuous and uninterrupted and at the same time to maintain that they are discontinuous and interrupted; the only way to paper over this manifest contradiction is to introduce the distinction between mere perceptions and external objects, a distinction that presupposes that we have an idea of externality in the first place, an idea we have shown to be impossible. And in order to remedy this inconsistency, the proponent of this system is forced to beg the question, presupposing the very idea of externality whose origin the system is meant to explain: This philosophical system, therefore, is the monstrous offspring of two principles, which are contrary to each other, which are both at once embrac’d by the mind, and which are unable mutually to destroy each other. The imagination tells us, that our resembling perceptions have a continued and uninterrupted existence, and are not annihilated by their absence. Reflection tells us, that even our resembling perceptions are interrupted in their existence, and different from each other. The contradiction betwixt these opinions we elude by a new fiction, which is conformable to the hypotheses both of reflection and fancy, by ascribing these contrary qualities to different existences; the interruption to perceptions, and the continuance to objects. [1.4.2.52, 215]

But this conclusion, Hume realizes, leaves the original conundrum in place. Nature forces us to take for granted the existence of the external world; skeptical arguments tell us that there is a great difference between what we in fact know and the external world we take ourselves to know; but philosophical reflection tells us that we can’t even have an idea of the external world in the first place. These philosophical arguments therefore show not only that is there nothing to take for granted, but also that the skeptical arguments that demonstrate this make no sense at all on their own terms, as they purport to distinguish the external world from the perceptions with which we confuse it. If we follow Hume’s argument, we have effectively reasoned ourselves into a position in which we cannot even express our initial confusion, let  alone a solution to it. It is hard to imagine a philosophical abyss more difficult to escape. Before we can move toward that escape, Hume demonstrates that the abyss is even deeper than one might have suspected:

( 198 )  Books I and III Nature is obstinate, and will not quit the field, however strongly attack’d by reason; and at the same time reason is so clear in the point, that there is no possibility of disguising her. Not being able to reconcile these two enemies, we endeavor to set ourselves at ease as much as possible, by successively granting to each whatever it demands, and by feigning a double existence, where each may find something, that has all the conditions it desires. Were we fully convinc’d, that our resembling perceptions are continu’d, and identical, and independent, we shou’d never run into this opinion of a double existence; since we shou’d find satisfaction in our first supposition, and wou’d not look beyond. Again, were we fully convinc’d, that our perceptions are dependent, and interrupted, and different, we shou’d be as little inclin’d to embrace the opinion of a double existence; since in that case we shou’d clearly perceive the error of our first supposition of the continued existence, and wou’d never regard it any farther. ’Tis therefore from the intermediate situation of the mind, that this opinion arises, and from such an adherence to these two contrary principles, as makes us seek some pretext to justify our receiving both; which happily at last is found in the system of a double existence. [1.4.2.52, 215–​16]

So, only the “intermediate situation of the mind,” the situation of total confusion in which we simultaneously embrace contradictory views of the nature of our own ideas, could lead us to the doctrine of double existence, or the doctrine that we can distinguish internality from externality. Hume frankly concedes that these arguments not only undermine the naive view that we know the external world directly as well as the humanly impossible view that there is no external world; they even undermine the skeptical position they entail, and, if we take them seriously, they also undermine the naturalistic account of the confusion that leads us to talk of an external world. Nothing is left. This is why this discussion is so unsettling and vertiginous. It not only partakes of the purgative aspects of the section on skepticism with regard to reason, but also seems to eliminate even the possibility of the purge. Hume sets out the dismal consequences: I begun this subject with premising, that we ought to have an implicit faith in our senses, and that this wou’d be the conclusion, I should draw from the whole of my reasoning. But to be ingenuous, I feel myself at present of a quite contrary sentiment, and am more inclin’d to repose no faith at all in my senses, or rather imagination, than to place in it such an implicit confidence. I cannot conceive how such trivial qualities of the fancy, conducted by such false suppositions, can ever lead to any solid and rational system. They are the coherence and constancy of our perceptions, which produce the opinion of their continued existence; tho’ these qualities of perceptions have no perceivable connection with such an

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 199 )

existence. The constancy of our perceptions has the most considerable effect, and it is attended with the greatest difficulties. [1.4.2.56, 217]

Here Hume draws the pessimistic conclusion. This argument betrays its own premises, refuting not only any entitlement to confidence in the reality of the world, but also the foundation on which that challenge is based. On the other hand, even though this predicament seems to arise from the Pyrrhonian position, for a truly relentless Pyrrhonian like Hume, things are not so bad after all; for Pyrrhonian skepticism, as we have seen, has a positive side as well. We can always step up a level and consider the argument metaphilosophically, and this is precisely what Hume does. The idea of the external world cannot stand philosophical scrutiny. Fortunately, however, we do not need to scrutinize it. Nature has taken care of that problem, creating in us a disposition to act both as though there is an external world, and to accept the supposition that this makes sense in the first place.15 If Hume is right, even these very words don’t have a determinate sense. But this is one more thing that the Humean approach to philosophy tells us. Very often—​not only in the case of externality, but also in the case of necessary connection—​we use words, but mean nothing by them. We act as though we have ideas that we could never have. In doing so, we participate in useful verbal customs, customs for using words that, despite the lack of a noncustomary semantic foundation for those words, allows us to communicate with one another. The customs themselves warrant the use of the words. We may be confused about what we are doing, taking ourselves to express ideas we do not have, but fortunately, that confusion does very little harm. This skeptical doubt, both with respect to reason and the senses, is a malady, which can never be radically cur’d, but must return upon us every moment, however we may chace it away, and sometimes may seem entirely free from it. ’Tis impossible upon any system to defend either our understanding or senses; and we but expose them farther when we endeavor to justify them in that manner. As the skeptical doubt arises naturally from a profound and intense reflection on those subjects, it always encreases, the farther we carry our reflections, whether in opposition or conformity to it. Carelessness and in-​attention alone can afford us any remedy. For this reason I rely entirely upon them; and take it for granted, whatever may be the reader’s opinion at this present moment, that

15. See Allison (2008) for a similar reading.

( 200 )  Books I and III an hour hence he will be persuaded that there is both an external and internal world. [1.4.2.57, 218]16

And this shows the way to a higher-​level skeptical resolution of the present conundrum. We begin with an opposition between two positions: according to one, we are entitled to talk about the external world because we have an idea of such a world, whose origin is explained either by the vulgar or the philosophical system; according to the other, since neither of these systems can provide such an idea, we are not entitled to talk about external objects. It appears that the parties to the debate agree about nothing. But they in fact agree about all that is important, the claim that we are entitled to talk about the external world if, and only if, we have an idea of it. Hume adopts the Pyrrhonian strategy of suspending the debate by rejecting the biconditional presupposition shared by the parties, and by arguing that custom, and custom alone, entitles us to talk about our world, and that this is all the justification we require. But, as Hume himself recognizes, it is hard to be comfortable with this solution. Custom seems a thin reed on which to hang so much of our confidence in our grasp of reality; hence the need for a bit of “carelessness and in-​attention” if we are to get by. Malherbe (2010) sums this up nicely, drawing together Hume’s skepticism and naturalism as they are in evidence here: Hume’s positive doctrine of common belief in the existence of objects and of the external world might itself appear fictitious; but it depends on the science of human nature to prove that it is experimentally false and to correct it. His naturalistic doctrine of philosophical belief in the double existence of perceptions and ideas might appear unacceptable to philosophers:  so much the worse for philosophers! It is obvious that his skeptical critique is formidable: it is the best of philosophy! His whole analysis appears quite uncommon by this strange mixture of naturalism and skepticism and, accordingly, uneasy for our common way of thinking. (138)

16. And Hume thinks that this Pyrrhonian response to skeptical arguments is even shared by his religious opponents. Consider this remark in a letter to Gilbert Elliot: But to all Appearance the Sentiment of Stockholm, Geneva, Rome antient & modern, Athens, & Memphis, have the same Characters. And no thinking man can implicitly assent to any of them; but from the general Principle, that as the Truth in these Subjects is beyond human Capacity, & that as for one’s own Ease he must adopt some Tenets, there is more Satisfaction & Convenience in holding to the Catechism we have been first taught. Now this I have nothing to say against. I woud only observe, that such a Conduct is founded on the most universal & determin’d Scepticism, join’d to a little Indolence. For more Curiosity & Research gives a direct opposite Turn from the same Principles. (Letter 71, February 18, 1751, Hume 1932, vol. 1, 151–​52)

Of S k ep t i c i s m wi t h Re g a r d t o t he Se n s e s  

( 201 )

Malherbe is correct in taking this potentially toxic cocktail of extreme Pyrrhonian skepticism and relentless naturalism to prove medicinal when drunk to the dregs. It then allows the metaphilosophical assent that allows us to make sense of our actual practices and the source of their warrant, and explains just how the existence of the external world can be “taken for granted in all our reasonings.” This analysis therefore demonstrates the depth of Hume’s commitment to the Pyrrhonian method.17 Taken together, the treatments of skepticism with regard to reason and skepticism with regard to the senses do a great deal to entrench the role of the imagination as a determinant of our cognitive and behavioral lives, and of the role of custom in our reasoning.18 We also see Hume exposing the emptiness of much philosophical thought that seeks independent foundations for those customs, and we see the relentless consistency of method that characterizes his analysis:  naturalism, Pyrrhonism, deconstruction of philosophical ideas, and the replacement of epistemological and metaphysical questions with psychological questions.19 Let us remember: in none of this analysis does Hume ask whether there is an external world, as he reminds us at the outset. This is an examination of the reasons for our confidence in the reality of the external world. The

17. In another context, Hume emphasizes this very reflexivity, and the way that skeptical arguments both reinforce and undermine themselves: “The following of general rules is a very unphilosophical species of probability; and yet ’tis only by following them that we can correct this, and all other unphilosophical probabilities” [1.3.13.12, 150]. This reflexive application of arguments, and the use of our dialectical customs to guide our everyday life even when those customs deliver the result that they themselves are not to be trusted, is a hallmark of Hume’s thoroughgoing Pyrrhonism—​encompassing both its negative and positive phases—​and is in evidence throughout the Treatise, another instance of the uniformity of Hume’s method. (I thank Kazunori Sawada for drawing the importance of this remark to my attention.) 18. Baxter (2006) draws far too pessimistic a conclusion at this point. He argues that Hume concludes at this point that he cannot assent to anything or endorse any view. But Hume does assent to the existence of the external world; he just doesn’t do so on rational grounds. Baxter distinguishes in this context between “passive” and “active” assent, arguing that while Hume takes himself unable to assent actively to the existence of the external world, but takes himself to do so passively. This reading does have the virtue of aligning Hume closely to Pyrrhonism and the notion of katalepsis as a response to appearances. Whatever the virtues of this reading are, it underestimates Hume’s position here: Hume is trying to explain not only why we passively assent to the reality of the external world, but also why we actively do so, for it is in fact obvious that we actively accept its reality (or at least that most of us do!). 19. We might also note that we see here the antecedents of Quine’s argument for a naturalized epistemology. Quine inherits from Hume a deep skepticism of transcendental epistemological justification. He, like Hume, sees justification as one more socially constituted epistemic practice—​one more custom—​that must itself be explained scientifically, again, using science that respects those very customs.

( 202 )  Books I and III

answer is that philosophy and reason have nothing to do with that confidence, nor does any kind of empirical evidence delivered by the senses; it is rather a matter of instinct. That may be disturbing, but it should not be surprising. Given the impossibility of either reasoning our way or learning our way into realism, why shouldn’t evolution have wired it into us? If it had not, we would never have survived to pose these questions.20

Quine’s insight that we never get to stand outside of our epistemic practices in order to assess them, that there is no Archimedean fulcrum from which justification derives its authority, has its origins in Hume’s insight that we never stand outside of custom in order to justify it; instead, custom itself is often justificatory, and any justifications we can provide owe their normative force to the customs they reflect. This insight in turn, as I  argued in the opening chapter, is guided by Hume’s understanding of the normative force of custom in law. So, both Quinean philosophy of science and Wittgensteinian epistemology owe a great deal to Hume’s approach to naturalizing normativity through a science of human nature, which in turn has its origins firmly in Anglo-​Saxon legal theory. 20. One might think that this very reflection would constitute an argument for the external world, thus undermining Hume’s position that there can’t be any. But it is not. It is an explanation of why, despite having no good reason to believe in an external world, we do. If there is none, we believe it because we are wired to, for whatever reason; if there is, the reason we believe in it is still simply that we are wired to do so, even though, if that is the case, we then have an explanation for the wiring.

C H A P T E R  1 0

w

Personal Identity and Philosophical Method

W

e close our consideration of Book I with an examination of Hume’s account of personal identity and the self. We will once again see our set of interpretative principles confirmed. The structure of Hume’s account and argument is completely consistent with those regarding causation, reason, and the external world. He is concerned primarily not with questions of metaphysics, but with the psychology of our self-​understanding. His account is resolutely Pyrrhonian, involving a classical skeptical inversion, and he argues that in fact we fail to have an idea we take ourselves to have—​the idea of a self. Our thinking about ourselves is explained with reference to the operation of custom. I  begin by showing how Hume’s discussion of the self and personal identity reflects these methodological commitments and vindicates this interpretative strategy, and then turn to his explicit reflections on philosophical methodology.

1.  PERSONAL IDENTITY Hume opens his discussion of the self by announcing quite explicitly that he will demonstrate not that there is or is not a self, but that we have no idea of the self in the first place, and that our use of the term is referentially empty. His strategy, as is well known, and consistent with the strategy we

( 203 )

( 204 )  Books I and III

have seen him employ before, is to argue that there is no impression from which such an idea could arise: There are some philosophers, who imagine we are every moment intimately conscious of what we call our self; that we feel its existence and its continuance in existence; and are certain, beyond the evidence of a demonstration, both of its perfect identity and simplicity. . . . Unluckily all these positive assertions are contrary to that very experience, which is pleaded for them, nor have we any idea of self, after the manner it is here explain’d. For from what impression cou’d this idea be deriv’d? This question ’tis impossible to answer without a manifest contradiction and absurdity; and yet ’tis a question, which must necessarily be answer’d, if we wou’d have the idea of self pass for clear and intelligible. [1.4.6.1–​2, 251]

In Madhyamakāvatāra (Introduction to the Middle Way), Candrakīrti deploys an argument for the conclusion that the self we instinctively posit (despite its nonexistence) cannot be identified with anything given in introspection, since introspection only delivers momentary experiences and reflection only psychophysical processes that, while often taken to belong to a self, have no independent subjective substrate that remains when they are subtracted.1 Hume, in a strikingly similar argument, points out that the impression of the self cannot be one of the many fleeting impressions we experience, whether perceptual or affective, for two reasons. First, the self is understood to be persistent through all of our constantly changing experience. Second, the self is supposed to be not on the objective, but on the subjective, side of things—​it is meant to be something distinct from our experiences of which it is the subject; that is, the self, if there is one, is the witness, agent, and enjoyer, not any of the things experienced, done, or appreciated: It must be some one impression, that gives rise to every real idea. But self or person is not any one impression, but that to which our several impressions and ideas are suppos’d to have a reference. If any impression gives rise to the idea of self, that impression must continue invariably the same, thro’ the whole course of our lives; since self is suppos’d to exist after that manner. But there is no impression constant and invariable. Pain and pleasure, grief and joy, passions and sensations succeed each other, and never all exist at the same time. It cannot, therefore, be from any of these impressions, or from any other, that the idea of self is deriv’d; and consequently there is no such idea. [1.4.6.2, 251–​52] 1. See Huntington and Wangchen (1995) for the text and Garfield (2015, chap. 4) for a more detailed account of the argument.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 205 )

In an oft-​quoted passage, Hume drives this argument home. It is not even clear that positing a permanent self, whether substantial or insubstantial, would explain anything. We would still need to fill two gaps. First, we would have to explain why our various experiences stand in need of such a mysterious substance in which to inhere—​why they simply can’t exist in the way that other events do. Second, even if we thought they needed a substantial support, we would have to explain exactly what the relation is between the self and these perceptions in virtue of which it explains anything, and in virtue of which we can know it and identify with it. The route is far from obvious in each case. Hume then points out that when we actually speak of self-​knowledge, and when we consider the real objects of introspective awareness, we find not a purely subjective substance, but rather the very impressions it is meant to ground. To the degree that we know ourselves, or are acquainted with ourselves, it is not a pure subject or a substance that is the object of our knowledge, but our experiences. So, to the extent that there are any impressions that are the source of a conception of who we are, they are our multitude of successive impressions: But farther, what must become of all our particular perceptions upon this hypothesis? All these are different, and distinguishable, and separable from each other, and may be separately consider’d, and may exist separately, and have no need of anything to support their existence. After what manner, therefore, do they belong to self; and how are they connected with it? For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without perception, and never can observe anything but the perception. When my perceptions are remov’d for any time, as by sound sleep; so long am I  insensible of myself, and may truly be said not to exist. And were all my perceptions remov’d by death, and cou’d I neither think, nor feel, nor see, nor love, nor hate after the dissolution of my body, I shou’d be entirely annihilated, nor do I conceive what is further requisite to make me a perfect non-​entity. If anyone upon serious and unprejudic’d reflexion, thinks he has a different notion of himself, I must confess I can reason no longer with him. All I can allow him is, that he may be in the right as well as I, and that we are essentially different in this particular. He may, perhaps, perceive something simple and continu’d, which he calls himself; tho’ I am certain there is no such principle in me. [1.4.6.3, 252]

So, Hume concludes, there is no impression that could ever give rise to an idea of the self (as opposed to the idea of a person), and so no idea of the

( 206 )  Books I and III

self. Hume then draws what might appear to be an uncharacteristic metaphysical conclusion regarding personhood: But setting aside some metaphysicians of this kind, I may venture to affirm of the rest of mankind, that they are nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement. Our eyes cannot turn in their sockets without varying our perceptions. Our thought is still more variable than our sight; and all our other senses and faculties contribute to this change; nor is there any single power of the soul, which remains unalterably the same, perhaps for one moment. [1.4.6.5, 252–​53]

It might appear that Hume has here gone beyond his examination of human nature to make a metaphysical claim:  that there is no persistent self, or even any persistent element in the self. But there are two reasons to resist this reading. The first is pretty straightforward: to make a claim about the nature of the person just is to pursue the task of spelling out human nature. So, this is not a general excursus into metaphysics, but a continuation of the project of mapping what it is to be a human being. But we can go further: this mapping is not an exercise in metaphysics so much as a diagnosis of what we actually mean when we use the first-​person pronoun or other terms referential to persons. After all, we do have the custom of individuating ourselves and each other, of distinguishing among us, and of discussing our respective doings and experiences. If the terms we use to refer to persons do not refer to selves, we are owed an account of their use. Hume proffers one here: they refer not to substantial selves standing as autonomous subjects of our experiences and agents of our psychological actions, but to bundles of causally connected psychophysical events, to the conventionally constituted persons we encountered in Book II and which we discussed in ­chapter 4, not to metaphysically autonomous selves.2 This discussion is hence not so much metaphysical, as it is semantic—​an explanation of what our language about persons means.3 Hume next turns to the psychological question, the question about human nature: given that we have no idea of a self, but only a practice of using names and pronouns, what is it that gives rise to our tendency to ascribe unity over time to persons? This, again, is a question not about the

2. See McIntyre (1993). 3. Compare Nagasena’s response to the king in the Milindapañha (Questions of King Milinda), when the king asks him who he is. He replies, “I am called ‘Nagasena,’ but that is just a name that has no real referent.”

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 207 )

nature of identity, but about the psychological processes that underlie our conventions: What then gives us so great a propension to ascribe an identity to the successive perceptions, and to suppose ourselves possest of an invariable and uninterrupted existence thro’ the whole course of our lives? In order to answer this question, we must distinguish betwixt personal identity, as it regards our thought or imagination, and as it regards our passions or the concern we take in ourselves. The first is our present subject; and to explain it perfectly we must take the matter pretty deep, and account for that identity, which we attribute to plants and animals; there being a great analogy betwixt it, and the identity of a self or person. [1.4.6.5, 253]

We have already seen Hume’s discussion of personal identity “as it regards our passions or the concern we take in ourselves,” in ­chapter 3. There we saw that an ascription of identity is important in order to provide an object for the passions, and that the passions themselves are operative in constituting this idea through the operation of the imagination. In this context, Hume is after our attribution of identity in a cognitive, and not an affective, context. So we must explain three things. First, just as in the case of belief in the external world, in which we must explain why we in fact believe that there is an external world, and do not just form a general idea of a world outside of ourselves, here we must explain why we believe that we have a self, as opposed to just forming some kind of general idea of one. Second, we must explain why we are willing to assert that a bundle of perceptions whose unity is determined affectively also has a practical and empirical unity. Finally, we must also explain why we never take ourselves to be indulging in a merely optional social or linguistic practice when we speak of persons in this way. It is also worth emphasizing in this context, as we did in c­ hapter 3, that there is no tension at all between Hume’s taking for granted the existence of persons as complex webs of passions and ideas in Book II, and his denial that we even have an idea of simple, continuing entities such as souls that constitute our identities in Book I.  Hume neither argues in Book II that there are substantial independent selves, nor does he deny that there are socially and affectively constituted persons in Book I. On the contrary, in Book II he shows that persons are complex characters, socially and biologically constituted by constantly changing sets of perceptions, passions, and social relations. In Book I, on the other hand, he shows that they cannot be reduced to simple souls or selves that underlie and are distinct from these changing psychophysical and social

( 208 )  Books I and III

processes regarded as individuals by custom. These positions are hardly inconsistent; in fact, they are complementary, and together constitute a unitary account of selfless personhood.4 Once again, Hume takes his project to be the explanation of a custom, both psychological and social, and neither to be an exploration of fundamental ontology nor the justification of that custom. And once again, his approach will be skeptical, explaining that the assertion of identity is itself nothing but a custom, that the identity we ascribe depends upon that custom, not the other way around. Nonetheless, just as in previous deployments of this skeptical strategy, Hume never rejects that custom in virtue of the absence of any noncustomary foundation; he rather affirms—​in a now familiar pattern of argument—​that the custom needs no foundation at all. Hume begins by examining the notion of diachronic identity itself: We have a distinct idea of an object, that remains invariable and uninterrupted thro’ a suppos’d variation of time; and this idea we call that of identity or sameness. We have also a distinct idea of several different objects existing in succession, and connected together by a close relation; and this to an accurate view affords us a perfect a notion of diversity, as if there was no manner of relation

4. See Baier (1991) and Ainslie (2005) for discussion of this issue. Ainslie emphasizes correctly the difference between Hume’s project in Book I and that in Book II. In Book II he is characterizing the complex social and psychological dimensions of personal identity and of the experience of our own subjectivity. In Book I he is pointing out that that understanding of ourselves is not accidentally social, relational, and determined in part by our sympathies, interests, projects, and the customary institutions in which we participate; it is essentially so, simply because it could not arise just from the introspective discovery of a simple, continuing self that merely happens to participate in social relations, that merely happens to experience passions, and that merely happens to experience bonds of sympathy to others. The two accounts of the self are hence not only not in tension with one another; they are two sides of the same broader account. Hume’s distinction between the two accounts of personal identity, Penelhum writes, “was a wise one . . . For whatever the span of the life is taken to be, our emotional life . . . requires for its logical structure that the span of the self is given. . . . What I take to be myself determines what I can be proud or ashamed of, and not the reverse. The personal identity that concerns the passions has to be one and the same as the personal identity that concerns the thought or imagination. But it is the thought or imagination, not the passions, that have to determine its boundaries” (2000b, 87). If my analysis is correct, then while Penelhum is right to see that there is one self, not two, being characterized in Hume’s two accounts, he reverses the priority of the two accounts in Hume’s project. It is the passions that lead thought and imagination, not the other way around. Once again, the affinities to Candrakīrti are striking: Candrakīrti argues against the reality of the self as an enduring substratum of our experience, but argues for the empirical reality of a person constituted by psychological and social convention.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 209 )

among the objects. But tho’ these two ideas of identity, and a succession of related objects be in themselves perfectly distinct, and even contrary, yet ’tis certain, that in our common way of thinking they are generally confounded with each other. [1.4.6.6, 253]

That is, any idea of strict identity, by Leibniz’s law, is an idea of the persistence of all properties with no change over time. (Note that this is strictly speaking impossible, as age changes over time necessarily, but leave that issue aside for the moment.) On the other hand, we also have an idea of distinct things that are closely related to one another in virtue of sharing many of their properties. Hume will argue that identity over time is always specific, and never numerical identity; the relation that earlier and later stages of the continua of states and processes to which we refer as persons is more like that between Julian and Joaquin Castro than between Superman and Clark Kent. To be sure, the relations between successive stages of a person are mediated by causality as well as resemblance; nonetheless, they are relations of similarity, not genuine identity. So, any attribution of strict, or numerical identity over time always involves a kind of confusion, one he here attributes to a certain laziness of the imagination: That action of the imagination, by which we consider the uninterrupted and invariable object, and that by which we reflect on the succession of related objects, are almost the same to the feeling, nor is there much more effort of thought requir’d in the latter case than in the former. The relation facilitates the transition of the mind from one object to another, and renders its passage as smooth as if it contemplated one continu’d object. [1.4.6.6, 253–​54]

This laziness, or ease of “transition of the mind from one object to another,” is not innocent, though. It leads to a serious confusion. We begin by confusing specific and numerical identity, but then, in order to posit an entity to support that identity, we end up talking about souls or substances, using terms to which we can attach no idea whatsoever as a way of camouflaging the confusion: This resemblance is the cause of the confusion and mistake, and makes us substitute the notion of identity, instead of that of related objects.  .  .  . Our propensity to this mistake is so great from the resemblance above-​mention’d, that we fall into it before we are aware; and tho’ we incessantly correct ourselves by reflexion, and return to a more accurate method of thinking, yet we cannot long sustain our philosophy, or take off this biass from the imagination. Our last

( 210 )  Books I and III resource is to yield to it, and boldly assert that these different related objects are in effect the same, however interrupted and variable. In order to justify to ourselves this absurdity, we often feign some new and unintelligible principle, that connects the objects together, and prevents their interruption or variation. Thus we feign the continu’d existence of the perceptions of our senses, to remove the interruption; and run into the notion of a soul, and self, and substance, to disguise this variation. [1.4.6.6, 254]

It is worth noting the uniformity of Hume’s method here, and the nice symmetry between his treatment of the external world and the inner world, a symmetry that contrasts directly with the asymmetry in the thought of philosophers like Berkeley or Descartes.5 Hume here has offered exactly the same kind of account of our propensity to think of the subject as persisting through time as he offered in the section on skepticism with regard to the senses of our propensity to take external objects to persist through time. After pointing out [1.4.6.6, 254–​55 and 1.4.6.7, 255] that we ascribe identity to ourselves in the same way and for the same reasons that we ascribe identity over time to things like animals and vegetables, which we know to undergo constant change, Hume emphasizes the importance of this deep and instinctive confusion: Thus the controversy concerning identity is not merely a dispute of words. For when we attribute identity, in an improper sense, to variable or interrupted objects, our mistake is not confin’d to the expression, but is commonly attended with a fiction, either of something invariable and uninterrupted, or something mysterious and inexplicable, or at least with a propensity to such fictions. [1.4.6.7, 255]

This is an important aspect of Hume’s consistently skeptical methodology. The skeptic, as Sextus emphasizes, and as we have noted previously, begins with a debate between opposing parties, in this case, about whether persons exist over time. One asserts a positive position, grounded in some supposed metaphysical reality, in this case, a soul or personal substance; the other insists that since there is nothing that corresponds to that metaphysical posit, the negation of that position must be true, and so that there is no personal identity at all. 5.  Kant is aware of this distinction in method when, in the Refutation of Idealism in the second edition of the Critique of Pure Reason, he takes Descartes and Berkeley as the two representatives of idealism, which he characterizes not in terms of a denial of the existence of the external world, but rather in terms of a distinction between the way the inner and the outer worlds are taken to be known, or to exist. He does not regard Hume as an idealist, and although he does not say as much, it is clearly for this reason.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 211 )

These two parties appear to disagree in all respects; but, the skeptic sees, they agree about the only thing that makes any difference: the confusion that grounds the debate, namely, that the positive thesis is true if, and only if, the supposed metaphysical underpinning is correct. In this case, the proponent of personal identity over time asserts the existence of a soul in order to explain the obvious truth of personal endurance; the nihilist denies identity over time because of the incoherence of the doctrine of an immutable soul. Hume describes this soul as a “fiction,” something we construct to paper over the inconsistency between existence over time which necessarily involves change, and numerical identity, which requires the sharing of all properties, including age. (We will return to the importance of the device of a fiction below.) The skeptic suspends the debate, refusing either position, by denying exactly the biconditional to which both parties to the debate assent—​in this case, the idea that personal persistence makes sense if, and only if, there is a soul that remains numerically identical over time. Instead, the skeptic explains our conventions for talking as the proponent talks, and shows how those conventions themselves require no metaphysical grounding. A debate about the persistence of the soul can be replaced by an explanation and survey of the conventions we adopt regarding pronouns and other referring expressions. And this is precisely what Hume does here, following exactly the same Pyrrhonian pattern we saw in the case of skepticism with regard to reason and with regard to the senses.6 At 1.6.6.8 [256] Hume points out that very small changes in very large things do not undermine our assertion of identity as much as large changes, and that very gradual changes do not undermine our sense of identity as much as abrupt changes. This, he argues, explains our conventions regarding the continued use of the same referring term despite change over time: we do so when there is sufficient causal continuity and gradual enough changes of properties. He then [1.6.6.10, 257] uses the example of Theseus’s ship to show that a common purpose can also lead us to assert identity, as in the cases of institutions.7 He concludes: We may also consider the two following phaenomena, which are remarkable in their kind. The first is, that tho’ we commonly be able to distinguish pretty

6.  Candrakīrti, in Introduction to the Middle Way, also argues that there are no convention-​independent grounds for our conventions regarding the individuation of persons, but that, nonetheless, those conventions constitute persons as well as the conventional truth about persons. See Garfield (2015, chap. 4) for details. 7. Green (1968) also notes that Hume takes the unity of an object to consist in the identity of its temporal parts, where that identity is imputed on the grounds of similarity and causal relations.

( 212 )  Books I and III exactly betwixt numerical and specific identity, yet it sometimes happens, that we confound them, and in our thinking and reasoning employ the one for the other. Thus a man, who hears a noise, that is frequently interrupted and renew’d, says, it is the same noise; tho’ ’tis evident the sounds have only a specific identity or resemblance, and there is nothing numerically the same, but the cause, which produc’d them. [1.4.6.13, 257–​58]

The example that comes closest to the case of personal identity, and which draws all of these ideas together, is that of the church.8 Here we see precisely that it is only conventions that undergird the ascription of identity, not any underlying substantial continuity, and that those conventions are based upon gradualness of change, causal continuity, and community of purpose: In like manner it may be said without breach of the propriety of language, that such a church, which was formerly of brick, fell to ruin, and that the parish rebuilt the same church of free-​stone, and according to modern architecture. Here neither the form nor materials are the same, nor is there anything common to the two objects, but their relation to the inhabitants of the parish; and yet this alone is sufficient to make us denominate them the same. But we must observe, that in these cases the first object is in a manner annihilated before the second comes into existence; by which means, we are never presented in any one point of time with the idea of difference and multiplicity; and for that reason are less scrupulous in calling them the same. [1.4.6.13, 258]

With these analogies and explanations of the grounds of the conventions regarding reference in hand, Hume turns directly to the application to the case of personal identity: We now proceed to explain the nature of personal identity. . . . And here ’tis evident, the same method of reasoning must be continu’d, which has so successfully explain’d the identity of plants, and animals, and ships, and houses, and of all the compounded and changeable productions either of art or nature. The identity, which we ascribe to the mind of man, is only a fictitious one, and of a like kind with that which we ascribe to vegetables and animal bodies. It cannot, therefore, have a different origin, but must proceed from a like operation of the imagination upon like objects. [1.4.6.15, 259]

8. But note that Hume also invokes the same metaphor used so often in Buddhist philosophy—​that of the river, which is ascribed a unity despite the constant change in its composition and characteristics [1.4.6.13, 258].

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 213 )

Hume first brings the analogies to bear, emphasizing the conventional basis of all identity ascriptions, and the easy confusion between an identity we construct and an identity we discover. He then concludes that personal identity is of exactly the same kind as other instances of identity:  a specific identity we ascribe, confused with a numerical identity we take ourselves to have discovered. The critical term here, just as in the context of the conclusion of the discussion of skepticism regarding the senses, is fictitious, and we should treat it in the same way in both contexts. Fictions, as we emphasized in the context of skepticism regarding the senses, are entities or tales we create. They are not unreal, but are instead real creations. In this context, the important features of fictional entities are these: (1) fictions are created, not discovered; and (2) fictions nonetheless can determine a real distinction between truth and falsity. First, fictions are created by persons, not discovered in nature.9 Specific identities are like that. For two things to be called by the same term, we must contrive a convention for using that term that comprises those two things; for two distinct things to be regarded as identical, we must have a custom that singles out the properties they share as salient, as opposed to those that they do not, in virtue of which they are in fact distinct. These customs may themselves be artificial, or they may be innate cognitive customs; either way, they are customs that enable a fiction, not mechanisms for the discovery of hidden metaphysical facts. Second, there is a distinction between truth and falsity with respect to fictions. It is true that Ahab was captain of the Pequod, and false that Queequeg captained the ship. And Melville’s fiction grounds that truth and that falsity; nothing outside of the fiction could provide that ground. You wouldn’t go to Nantucket and search the whaling archives to verify that fact; you would read Moby Dick. Similarly, emphasizes Hume in using this term, there are truths and falsities about personal identity.10 It is true that I am the person who wrote Engaging Buddhism a few years ago, and false that I wrote A Treatise of Human Nature a few centuries ago. But what makes it true that I wrote Engaging Buddhism and not the Treatise is a set of conventions for ascribing identity over time, not some facts about my soul

9. The etymology of fiction takes us to the Latin fingere, to invent or to contrive. See Traiger (1987) for an insightful discussion of the variety of ways in which Hume uses the term fiction. 10. And in this context, it is also worth noting that truth and trust are cognate in English. As fictions create truth, they create entities in which we can trust.

( 214 )  Books I and III

and Hume’s. These are fictional truths.11 Hume emphasizes the fact that we create, rather than discover, identity (even though you or I may discover the identity that we created), and the role of custom in this creation in the following passage: But, as, notwithstanding this distinction and separability, we suppose the whole train of perceptions to be united by identity, a question naturally arises concerning this relation of identity; whether it be something that really binds our several perceptions together, or only associates their ideas in the imagination. That is, in other words, whether in pronouncing concerning the identity of a person, we observe some real bond among his perceptions, or only feel one among the ideas we form of them. This question we might easily decide, if we wou’d recollect what has been already prov’d at large, that the understanding never observes any real connexion among objects, and that even the union of cause and effect, when strictly examin’d, resolves itself into a customary association of ideas. For from thence it evidently follows, that identity is nothing really belonging to these different perceptions, and uniting them together; but is merely a quality, which we attribute to them, because of the union of their ideas in the imagination, when we reflect upon them. [1.4.6.16, 259–​60]

That is, Hume emphasizes, there are no metaphysical or otherwise custom-​ independent facts that could confirm or disconfirm personal identity, just as there are no records in Nantucket that could confirm or disconfirm Ahab’s 11. It is also true, in another fiction, that the fourteenth Dalai Lama is identical to the thirteenth Dalai Lama in the same sense that I am identical to the author of Engaging Buddhism. Many people take that fiction just as seriously as most readers of this book take the fictions regarding identity popular in our own culture. Facts about personal identity are more cultural than we often realize. One might reply that there must be more to it than this; not just any convention will do; our practices for identifying each other and everyday objects over time are not simply arbitrary. There is something to this objection, but less than meets the eye. So, for instance, one might argue that even though I am older now than I was when I was a child, my present stage shares the same DNA with that stage; perhaps that guarantees my identity in a sense that transcends custom, a guarantee that would distinguish the relation between my younger and my present self from that between the thirteenth and fourteenth Dalai Lamas. But this objection misses the mark. The physical or biological facts may be indisputable; but the link between them and judgements of identity require customs. The possibility of mind-​body transplants, for instance, calls into question our willingness to rely exclusively on DNA as a criterion of personal identity. As Hume’s example of the church [1.4.6.9, 258] shows, we may agree about all of the facts on which judgments of identity are grounded, but disagree on whether or not they warrant judgments of identity; our interests and customs matter, just as determinations of citizenship may in one country be grounded on blood, in another on birth; customs are needed to move from physical facts to identity claims.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 215 )

captaincy of the Pequod. And even causality—​one of the most important relations to which we appeal in judgments of identity or difference—​is itself a relation we recognize only in virtue of custom. Hume draws his series of analogies to personal identity to a close with that of the republic, illustrating once again not only that personal identity is constructed, not discovered, but that it is institutional and multidimensional: In this respect, I cannot compare the soul more properly to anything than to a republic or commonwealth, in which the several members are united by the reciprocal ties of government and subordination, and give rise to other persons, who propagate the same republic in the incessant changes of its parts. And as the same individual republic may not only change its members, but also its laws and constitution; in like manner the same person may vary his character and disposition, as well as his impressions and ideas, without losing his identity. Whatever changes he endures, his several parts are still connected by the relation of causation. And in this view our identity with regard to the passions serves to corroborate that with regard to the imagination, by the making our distant perceptions influence each other, and by giving us a present concern for our past or future pains or pleasures. [1.4.6.18, 261]

At the end of this passage, Hume connects this Book I  discussion of personal identity to that of Book II, which we discussed in ­chapter 3. The fact that causal relations are so central to our identity as persons provides the bridge. We initially come to regard ourselves as individuals in virtue of experiencing passions that require an object as well as a cause. To experience pride or shame, anger, resentment, or even affection, we need not only an external cause for these passions, but also an object to which to refer them, represented in relation to the causes of those passions. But those causes are often in the past or even, as anticipated, in the future (and, as McIntyre [1993] notes, rely on relations of sympathy between the successive stages of their subjects). To unite them and to make sense of our affective lives, as Penelhum (2000b) points out, the persons we posit must extend over time. For this reason, the passions are also important sources of our sense of diachronic personal identity. Hence, the same demand for the fiction that underlies our cognitive lives underlies our affective lives.12 The entire discussion is brought to a

12. I also note that the way Hume writes this passage indicates that he already had the view of the self in relation to the passions in mind when he wrote this part of Book I, confirming the thought that Book II is in fact the conceptual foundation of the Treatise.

( 216 )  Books I and III

close with this lovely summary that draws together the skepticism, the nominalism, the naturalism, and the concern for custom that characterize the methodology that runs through the Treatise. Notwithstanding Hume’s earlier remark that grammatical errors lead to metaphysical folly, the essentially grammatical character of this debate remains central to its diagnosis and resolution: The whole of this doctrine leads us to a conclusion, which is of great importance to the present affair, viz. that all the nice and subtile questions concerning personal identity can never possibly be decided, and are to be regarded rather as grammatical than as philosophical difficulties. [1.4.6.20, 262]

2.  PHILOSOPHICAL METHOD Hume concludes Book I with methodological reflections, and it is time for us to conclude this part of our investigation by following those reflections. As we have seen, Hume’s approach to reason, to the existence of the external world, and to the status of the person has been resolutely Pyrrhonian. One of the complicating features of a Pyrrhonian—​as opposed to an Academic—​ skepticism is its reflexivity:  like the laxative that purges itself as well as the material it is designed to purge, the skeptical arguments are also self-​ undermining. The very considerations they adduce to undermine the confidence we ought to have in reason, in the senses, and in our own existence undermine the confidence we ought to have in those arguments themselves. What attitude, then, ought Hume have to his own investigations and their results? Hume expresses these worries as follows (and the very fact that he expresses these particular worries is further evidence for his Pyrrhonism): For with what confidence can I venture upon such bold enterprizes, when beside those numberless infirmities peculiar to myself, I find so many which are common to human nature? Can I be sure, that in leaving all establish’d opinions I am following truth; and by what criterion shall I distinguish her, even if fortune shou’d at last guide me on her foot-​steps? . . . Experience is a principle, which instructs me in the several conjunctions of objects for the past. Habit is another principle, which determines me to expect the same for the future; and both of them conspiring to operate upon the imagination, make me form certain ideas in a more intense and lively manner, than others, which are not attended with the same advantages. Without this quality, by which the mind enlivens some ideas beyond others (which seemingly is so trivial, and so little founded on reason) we cou’d never assent to any argument, nor carry our view

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 217 )

beyond those few objects, which are present to our senses. Nay, even to these objects we cou’d never attribute any existence, but what was dependent on the senses; and must comprehend them entirely in that succession of perceptions, which constitutes our self or person. Nay farther, even with the relation to that succession, we cou’d only admit of those perceptions, which are immediately present to our consciousness, nor cou’d those lively images, with which the memory presents us, be ever receiv’d as true pictures of past perceptions. The memory, senses, and understanding are, therefore, all of them founded on the imagination, or the vivacity of our ideas. [1.4.7.3, 265]

This is exactly what should bother Hume at this point. His arguments have undermined any reason to have confidence in reason, or even in those arguments themselves. He has responded to those arguments with an account of human nature—​by instituting habit or custom as the principle that takes us from premises to conclusions, from causes to effects, from sensations to confidence in the external world, and from our passions to our sense of self. All well and good; but even if these naturalistic explanations are correct replacements for rational proofs of the veridicality of reason, of the existence of the external world, of the reliability of our senses, and of the existence of the self, these explanations are themselves justified by arguments whose probative force they themselves undermine. So, whatever their merits, why believe them? This is the kind of issue that only a true Pyrrhonian can raise, an issue that is as deep, troubling, and vertiginous as those raised in the context of skepticism regarding the senses, and Hume raises them in a specifically naturalistic voice. Hume concludes by noting once again that on his view, the faculty responsible for our confidence in reason, our senses, and our own identity over time, is neither reason nor the senses, but the imagination. But this only raises a further problem, as the imagination itself has no claim to represent the truth. This is poignant in the present reflective context because the only confidence Hume can have in his own reasoning is, on his own reasoning, due to the operation of the imagination: ’Tis this principle, which makes us reason from causes and effects; and ’tis the same principle, which convinces us of the continu’d existence of external objects, when absent from the senses. But tho’ these two operations be equally natural and necessary in the human mind, yet in some circumstances they are directly contrary, nor is it possible for us to reason justly and regularly from causes and effects, and at the same time believe the continu’d existence of matter. How then shall we adjust those principles together? Which of them shall we prefer? [1.4.7.4, 266]

( 218 )  Books I and III

As Hume notes here, things get even worse as we follow this train of thought:  the imagination leads us both in the direction of affirming reason and in the direction of faith in our senses. But, as he concluded at the end of 1.4.4, in his discussion of skepticism, “There is a direct and total opposition betwixt our reason and our senses; or more properly speaking, betwixt those conclusions we form from cause and effect, and those that persuade us of the continu’d and independent existence of body. When we reason from cause and effect, we conclude, that neither colour, sound, taste, nor smell have a continu’d and independent existence” [1.4.4.15, 231]. The problem is that the imagination is both the foundation of our confidence in reason, and the foundation of our confidence in the existence of the external world. Reason, however, tells us that we ought to have no faith at all in our senses, and no confidence in the existence of the external world; and certainly, we might add, in a faculty so free in its operations as the imagination, which independently leads us to believe in the external world. So even if we were to provide some basis for confidence in the imagination, that confidence would soon undermine itself. This, Hume argues [1.4.7.5, 266–​67], leads us to a “dangerous dilemma”: we can either follow the imagination blindly and risk being misled by its unconstrained ruminations, or reject it as a source of cognitive confidence. But, if we follow the latter course, we end up following reason, which, as we have already seen, necessarily subverts itself. Hume asks: Shall we, then, establish it for a general maxim, that no refin’d or elaborate reasoning is ever to be receiv’d? Consider well the consequences of such a principle. By this means you cut off entirely all science and philosophy: You proceed upon one singular quality of the imagination, and by a parity of reason must embrace all of them:  And you expressly contradict yourself; since this maxim must be built on the preceding reasoning, which will be allow’d to be sufficiently refin’d and metaphysical. What party, then, shall we choose among these difficulties? If we embrace this principle, and condemn all refin’d reasoning, we run into the most manifest absurdities. If we reject it in favour of these reasonings, we subvert entirely the human understanding. We have, therefore, no choice left but betwixt a false reason and none at all. For my part, I know not what ought to be done in the present case. I can only observe what is commonly done; which is, that this difficulty is seldom or never thought of; and even where it has once been present to the mind, is quickly forgot, and leaves but a small impression behind it. Very refin’d reflections have little or no influence upon us; and yet we do not, and cannot establish it for a rule, that they ought not to have any influence; which implies a manifest contradiction. [1.4.7.7, 268]

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 219 )

This is why Hume opens this section with dramatic paragraphs in which he compares himself to “a man, who having struck on many shoals, and having narrowly escap’d ship-​wreck in passing a small frith, has yet the temerity to put out to sea in the same leaky, weather-​beaten vessel” [1.4.7.1, 263]. He describes his state of mind as “affrighted and confounded with that forelorn solitude” [1.4.7.2, 264]. He points out that his skeptical arguments and the conundrum in which they seem to have landed him inevitably expose him “to the enmity of all metaphysicians, logicians, mathematicians, and even theologians” [1.4.7.2,  264]. And while Hume refers to these attitudes as the symptoms of “spleen and indolence” [1.4.7.11, 270], he only uses this hyperbole to indicate the depth of the abyss that confronts the Pyrrhonian. The task of the Treatise, after all, is to ground a science of human nature, and the quest for the foundations of that science at least appear to have rendered that science impossible, in virtue of the fact that no cognitive instruments of scientific inquiry could be adequate to the task. The enmity Hume anticipates is easy to understand. In establishing an empirical science of human nature grounded in an account of custom, and thereby a naturalized epistemology, Hume is criticizing the foundationalism and the commitment to an Archimedean transcendental approach to epistemology shared by those as disparate as Descartes, Berkeley, Locke, and Leibniz. They would seek a warrant for knowledge or for obligation independent of custom that might, among other things, warrant and justify the epistemic and moral customs we adopt. This is how Descartes’s account of the authority of clear and distinct perception works in the domain of knowledge, and Locke’s contract theory functions in the political domain. Such an account would ground the certainty of metaphysical reasoning, of ethics, and of mathematics; such an account might even allow us to prove the existence of a deity. Hume’s skeptical method puts paid to all of this. On his account, the only source of warrant is custom itself. There the spade is turned. But, as he now notes, these arguments can be suicidal, turning on themselves, and undermining their own claims to cogency as well. In the introduction to the Treatise, Hume asserts that “nothing but the most determined scepticism, along with a great degree of indolence, can justify this aversion to metaphysics” [introduction 3, xiv]. In that context, he expresses confidence in the prospects for his own philosophical project. Now, however, indolence and skepticism seem to have taken hold. One could imagine that, at this point, Hume might mitigate his own skepticism, to find a way out through the discovery of some transcendental epistemic principle that would confer warrant on at least some of

( 220 )  Books I and III

his arguments. Garrett (1997, 2006) argues that he does so through the adoption of what he calls the “title principle” (again, not to be confused with my Cover Principle!). That principle, he argues, is expressed in Hume’s statement that “where reason is lively, and mixes itself with some propensity, it ought to be assented to. Where it does not, it never can have any title to operate upon us” [1.4.7.11, 270]. But, were Hume to do so, this would be a straightforward betrayal of his own naturalism and skepticism. Hume is, as I have been emphasizing throughout this study, explaining how it is that we come to believe. To the extent that that account is not only genetic, but warranting, the warrant comes not from a transcendental principle regarding epistemic warrant, but from the power of custom to provide warrant for our practices. Moreover, even if we were to suppose his willingness to betray these explicit methodological commitments that are so obviously at work throughout the Treatise, the betrayal itself would be in vain, and Hume would be cognizant of its vanity:  such a principle, were it to be understood as anything but a description of our psychological dispositions, could never be an escape from the skepticism to which it is thought to be a response. In that case, it would be subject to the same “dangerous dilemma” undermining all of our conclusions. Instead of trying to climb out from his skepticism, Hume therefore escapes his “spleen and indolence” and his despair by diving even deeper. That is, Hume here offers a thoroughly Pyrrhonian reply to these new Pyrrhonian worries. These worries, as I  noted earlier, are somewhat different from those Hume put to rest in the discussion of skepticism with regard to reason. Those concerned our faith only in reason, and we saw that it is custom that provides the confidence we indeed have in reason, and grounds its normative force. But, in these new reflections, Hume has shown that that very confidence is undermined by reason itself when it reflects on its own operations, using the very processes in which we have customary confidence: We “have . . . no choice left but betwixt a false reason and none at all.” As Baier puts it, “Reason subverted by its own unrestrained and reflexive application is ultimately ‘no reason at all,’ whereas the false reason is the prophetically Kantian understanding which uses forbidden reasoning to arrive at that very prohibition, which thinks beyond the limits of the understanding to discern those limits. It is false to itself” (Baier 1991, 14). Or, to put it another way, for reason to be false is for reason to operate in a way that reason itself shows to be irrational; for it to be shown to be no reason at all, is for the rational skeptical reflection on reason to show that reason taken by itself cannot be probative.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 221 )

The only solution at this point is to “observe what is commonly done,” that is, as Sextus would have put it, to observe the fourfold skeptical prescription to follow the passions, the guidance of nature, custom, and the instruction of the arts, including the art of philosophy, a practice that allows us to avoid thinking of this difficulty, or if we do, to forget it quickly. This is what Hume does here and throughout the Treatise. And it is all that one can do given the force of skeptical arguments. After all, Hume points out that he does not know what “ought to be done.” Reason and norms that transcend custom have nothing to do with this; the very warrant of reason, like all norms, depends entirely on custom. His description of what he does is just that: a description, not a prescription—​albeit a description that entails the probity of prescriptions in ordinary life—​once again, this time reflexively, following the Cover Principle, and even in his metaphilosophy, hewing to the method that guides all of his philosophical reflections.13 This is not a refusal, or even a mitigation of his skepticism, but a deepening of it: In all the incidents of life we ought still to preserve our scepticism. If we believe, that fire warms, or water refreshes, ’tis only because it costs us too much pains to think otherwise. Nay if we are philosophers, it ought only to be upon sceptical principles, and from an inclination, which we feel to the employing ourselves after that manner. [1.4.7.11, 270]

This is the context for Garrett’s “Title Principle,” which read in this context can be nothing but a description of an epistemic inclination. The positive turn comes when Hume takes “a solitary walk by a river-​side” and discovers that he is “naturally inclin’d to carry [his] view into all those subjects,” subjects that include “the principles of moral good and evil, the nature and

13. Allison sees this differently. He writes that “Hume’s task is not to find some way to avoid the dilemma of having to choose between a false reason and no reason at all, but to vindicate his continued reliance upon the former” (Allison 2008, 318). He argues that “he needs some sort of justification to set aside these worries and to proceed ‘carelessly’ on his voyage. Or, paradoxically expressed, he needs something like warrant to carry on without the kind of warrant that epistemologists traditionally require for belief” (318). The first comment is off base: Hume does not seek to rely upon “false reason,” but to show that belief is not warranted by reason. So Allison gets it right in the end, but does not see that the second remark in fact undermines, rather than explains, the first. It is the Pyrrhonian fourfold prescription that supplies the requisite warrant. This also shows why Allison is wrong when he writes that “it is evident that the skepticism that Hume is here advocating should not be equated with the radical or Pyrrhonian variety, since the whole point of Hume’s critique . . . is that it cannot be preserved in common life” (2008, 320). This simply misses the structure of Pyrrhonism. And this is the same problem that undermines Garrett’s appeal to the title principle.

( 222 )  Books I and III

foundation of government, and the cause of those several passions and inclinations, which actuate and govern [him]” [1.4.7.12, 270–​71]. That is, he allows his inclinations to guide his behavior, not because it is an independent warrant, not because it is reliable, but because that is how custom manifests itself; that is how we are wired. Hume pursues philosophy because he is inclined to; others such as the “honest gentlemen of England” [1.4.7.14, 272] with no such disposition, are welcome not to do so. When Hume determines to leave the study to enjoy a glass of wine, the company of friends, and a game of backgammon, he is happily participating in the customs of his country, participating in social conventions with the same gusto with which he participates in epistemic conventions, granting neither any special transcendental status.14 Skepticism, when—​and only when—​pursued to the end, is hence no real cause for anxiety. Recall the stories of Pyrrho and his happiness, and Sextus’s own description of the goal of the skeptic, ataraxia. Sextus avers that there is no reason that the skeptic can endorse to believe that the skeptical method will lead to that end. Like Apelles, who after trying everything else, flung his sponge at the painting and by chance got the foam on the horse just right, Sextus argues that skeptical arguments indeed lead to ataraxia. Hume makes the same point: Most fortunately it happens, that since reason is incapable of dispelling these clouds, nature herself suffices to that purpose, and cures me of this philosophical melancholy and delirium, either by relaxing this bent of mind, or by some avocation, and lively impression of my senses, which obliterate all these chimeras. I dine, I play a game of back-​gammon, I converse, and am merry with my friends; and when after three or four hour’s amusement, I  wou’d return to these speculations, they appear so cold, and strain’d, and ridiculous, that I cannot find it in my heart to enter into them any farther. [1.4.7.9, 269]

So, Hume solves the skeptical predicament by relying on human nature. In doing so, he follows the skeptical method of yielding to nature and to custom for guidance in life, including guidance in philosophy. And it is here that Hume’s allegiance to Pyrrhonian skepticism is most completely revealed. This yielding is not and cannot be rational assent, but conformity to custom, where, once again, custom is understood in both its habitual and conventional senses, and comprises not only the custom of playing 14. Ainslie (2015) offers a similar reading, but he does not note the central role of custom in Hume’s argument for this conclusion or in Hume’s understanding of that in which this position consists.

P e r s o n a l I de n t i t y a n d P hi l o s ophi c a l  Me t h o d  

( 223 )

backgammon, but also that of doing philosophy, of reflecting, even when one acknowledges the limitations and even dangers of that reflection:15 Here then I find myself absolutely and necessarily determin’d to live, and talk, and act like other people in the common affairs of life. . . . I may, nay I must yield to the current of nature, in submitting to my senses and understanding; and in this blind submission I shew most perfectly my sceptical disposition and principles. But does it follow, that I must strive against the current of nature . . . that I must torture my brain with subtilities and sophistries. . . . If I must be a fool, as all those who reason or believe any thing certainly are, my follies shall at least be natural and agreeable. [1.4.7.10, 269]

Popkin puts the point nicely: The skeptical view that we ought not and do not hold any opinions is false. We must hold opinions since nature forces us to. It is not really a question of what we should do, but rather a question of what we have to do. We are required to judge about (i) factual, (ii) moral, and (iii) rational matters, regardless of the legitimate evidence that we have at our disposal. As Hume explained at great length, our belief in judgment about matters of fact is not due to rational evidence, but is the result of a custom or habit, which produces a strong feeling in us regarding certain ideas usually conjoined with an impression now present to the senses or memory. The judgments that we believe concerning such matters impose themselves upon us because of our mental constitution. Thus, it is nature and not logical reasoning that leads us to make all causal inferences. It is this type of inference which “peoples the world” and allows us to talk of matters not immediately present to sense or memory. (Popkin 1980, 116)

Popkin points out in what follows that the same is true of moral judgments, as we shall see in the next chapter, and indeed, of our reflective judgments about our own psychological processes and epistemic practices, completing the Pyrrhonian circle. 15. Baier (1991) points out that in taking this Pyrrhonian course, and following skepticism not only in the negative, but also in the positive phase of its dialectic, Hume is acting in good faith, avoiding the kind of inconsistency that would follow from any attempt to save reason from itself. This explains why Hume feels that he can return to philosophy after all: philosophy is one more human practice or custom, one from which the mind curious about human nature or even the role of custom in our epistemic practices cannot recuse itself. But in returning to philosophy in this skeptical temper, we return to a practice that we can no longer take as foundational to the remainder of our knowledge, but rather part and parcel of an entire set of epistemic and other human practices.

( 224 )  Books I and III

So Hume concludes Book I, and we conclude our consideration of that portion of the Treatise. I hope to have shown in these five chapters that while Book I  precedes Book II in publication, it rests upon Book II conceptually. It is in Book II that Hume sets out the account of the passions that rule our nature, and of the custom that governs our behavior. Without that account of the fundamental individual and social psychology that constitutes human nature, and of the central role of the imagination in our cognitive lives, the constructive skepticism developed in Book I would be impossible, and the uniformity of method Hume deploys would not be apparent. We now turn to Hume’s account of morality in Book III, which we will see is identical in form to the account of our cognitive lives in Book I. Once again, Hume eschews a prescriptive approach for a descriptive one, focusing on human nature. And once again, he deploys a skeptical analysis grounded in an analysis of the role of the imagination and custom in modulating our passions, and making us fit for the society in which we become fully human.

C H A P T E R  1 1

w

Ethics

M

uch has been made of Hume’s debt to Hutcheson and to Shaftesbury, and properly so. His account of the moral sentiments derives directly from their work, as Hume himself acknowledges. In emphasizing the role Hume assigns to the sentiments in ethics, it is, however, easy to lose sight of other features of his approach to morality that are both more distinctive to Hume himself, and more in harmony with the remainder of the Treatise. We will see that when we read Book III with the interpretative principles I have proposed for the Treatise as a whole, the analysis he deploys is far more important to Hume’s project than the sentimentalism he inherits from his predecessors.1

1. Hume is explicit about this intention to maintain his naturalism in the study of ethics in his letter to Cheyne in 1734: Having now Time & Leisure to cool my inflam’d Imaginations, I began to consider seriously, how I shou’d proceed in my Philosophical Enquiries. I found that the moral Philosophy transmitted to us by Antiquity, labor’d under the same Inconvenience that has been found in their natural Philosophy, of being entirely Hypothetical, & depending more upon Invention than Experience. Every one consulted his Fancy in erecting Schemes of Virtue & of Happiness, without regarding human Nature, upon which every moral conclusion must depend. This therefore I resolved to make my principal Study, & the Source from which I wou’d derive every Truth in Criticism as well as Morality. I believe ’tis a certain Fact that most of the Philosophers who have gone before us, have been overthrown by the Greatness of their Genius, & that little more is requir’d to make a man succeed in this Study than to throw off all Prejudices either for his own Opinions or for this of others. At least this is all I have to depend on for the Truth of my Reasonings, which I have multiply’d to such a degree, that within these three Years, I find I have scribled many a Quire of Paper, in which there is nothing contain’d but my own Inventions. (Letter 3, to George Cheyne, March or April 1734, in Hume 1932, vol. 1, 16) ( 225 )

( 226 )  Books I and III

Baier, reflecting on Hume’s descriptive approach to ethics, writes: What I look for in a moral theory is a demystifying account of the deontological component in morality as decent people recognize it, an account which does not subordinate the gentler tones of that morality to its sterner deontological voice, along with a plausible explanation of our persistent tendency to mystify moral matters. Hume’s account satisfies these demands. (2010b, 124)

Needless to say, I agree, and I note that, here, we see an endorsement of the Principle of Uniformity of Method in Hume’s interpretation. In the moral domain, as in the domain of epistemology, Hume’s task is to provide a naturalistic explanation of our moral conventions, as well as to explain why we so often deeply misunderstand the structure of our own moral thought, taking ourselves to have ideas and motives that we never could have. He aims to explain how custom grounds, and is not grounded by, the moral universe we inhabit, just as he takes custom to ground, and not to be grounded by, the force of law. Although many of the components of Hume’s account of morality derive from Hutcheson, Shaftesbury, and Mandeville, the structure into which he places these components—​constituted by his naturalism and Pyrrhonism—​together with the understanding of custom based in legal theory, gives rise to something entirely new. Hume offers an account of ethics grounded in the union of natural sentiment and a

And he frankly acknowledges that this approach departs from that of a moralist in a letter to Hutcheson of about the same time: What affected me most in your Remarks is your observing, that there wants a certain Warmth in the Cause of Virtue, which, you think, all good Men wou’d relish, & cou’d not displease amidst abstract Enquirys. I must own, this has not happen’d by Chance, but it is the Effect of a Reasoning either good or bad. There are different ways of examining the Mind as well as the Body. One may consider its most secret Springs & Principles or to describe the Grace & Beauty of its Actions. I imagine it impossible to conjoin these two Views. Where you pull off the Skin, & display all the minute Parts, there appears something trivial, even in the noblest Attitudes & most vigorous Actions: Nor can you ever render the Object graceful or engaging but by cloathing the Parts again with Skin & Flesh, & presenting only their bare Outside. An Anatomist, however, can give very good Advice to a Painter or Statuary: And in like manner, I am perswaded, that a Metaphysician may be very helpful to a Moralist; tho’ I cannot easily conceive these two Characters united in the same Work. Any warm Sentiment of Morals, I am afraid, wou’d have the Air of Declamation amidst abstract Reasonings, & wou’d be esteem’d contrary to good Taste. And tho’ I  am much more ambitious of being esteem’d a Friend to Virtue, than a Writer of Taste; yet I must always carry the latter in my Eye, otherwise I must despair of ever being servicable to Virtue. I hope these Reasons will satisfy you; tho at the same time, I intend to make a new Tryal, if it be possible to make the Moralist & Metaphysician agree a little better. (Letter 13, to Francis Hutcheson, in Hume 1932, 32–​33)

E t hi c s  

( 227 )

natural propensity to artifice, reflecting the seamless integration of distinct instances of custom as it manifests both as habit and as convention, as a foundation for the normativity we seek in morality.2 In particular, we will see that rather than provide an account of ethics, per se—​that is, either an account of our duties, or of ethical facts (even facts constituted by our moral sentiments and customs)—​Hume will instead provide an account of our propensities to constitute moral conventions and to make moral judgments; that is, he will be demonstrating how our moral practices both emerge from and shape our human nature, as per the Cover Principle.3 To put this another way, Hume is neither concerned to defend nor to attack particular moral principles; nor is he concerned to assess customs themselves morally. Instead, he is concerned with the origins of normativity per se, and is committed to providing a naturalistic explanation of how we come to respect and to enforce norms. We have already seen this commitment at work in Hume’s discussion of epistemology: Hume, in his discussion of the warrant for causal inference, of the warrant of mathematical knowledge, or of the warrant of our beliefs in the external world, shows that we are neither justified nor unjustified in particular beliefs or practices; instead, he explains the social and psychological origins of our normative practices in those domains. In conformity to the Principle of the Uniformity of Method, we will see the same approach in the case of moral norms. As Cohon puts it, “Hume likewise wishes to give a naturalistic and empiricist account of valuing, in particular valuing of persons, and one that does not presuppose any independent conception of the good beyond natural good (pleasure)” (2011, 174).4 We will also see Hume’s Pyrrhonism at work: Hume neither grounds our moral practices on an underlying moral reality nor infers from the absence of any such foundation that our moral conventions are pointless; rather, 2.  See Baier (2010b) for an extended discussion of the respects in which Hume’s moral theory can be said to be original, and Cohon (2006, 2008)  for a very helpful discussion of how Hume’s moral theory can be simultaneously naturalistic and normative. But also see Shaver (1995) and Qu (2014) for a contrary reading, according to which Hume is concerned not only to explain, but also to justify, more transcendentally than naturalistically, our moral norms. 3.  Laird gets this wrong. He writes, “Hume’s ethical theory revolved round two principal questions, viz., the nature of approval and the problem of what we approve” (1932, 213). Now, as Laird correctly observes, Hume argues that approval is a matter of moral sense, and that what we approve is what we find agreeable. But that is what Hume inherits from Hutcheson and Shaftesbury, and he takes it more or less taken for granted. The questions that animate Hume’s moral theory concern how the interplay of individual nature and social forces constitute our moral sensibility. 4. See also Cohon (2008).

( 228 )  Books I and III

he demonstrates that our conventions themselves constitute the moral world. That account appeals to custom in both of Hume’s senses—​to our sentiments or passions, and to our collective practices. Implicit in all of this is Hume’s nominalism: we don’t begin with moral ideas and develop a moral discourse; rather, our moral discourse generates the moral ideas we in fact have. In short, Hume’s account of morals has exactly the same structure as his account of every other human phenomenon: Book III, like Book I, presupposes the account of the passions developed in Book II.

1.  THE ENDS OF MORALITY Hume begins by noting that morality is generally acknowledged to have a specific purpose, viz., “the peace of society”: Morality is a subject that interests us above all others: We fancy the peace of society to be at stake in every decision concerning it; and ’tis evident, that this concern must make our speculations appear more real and solid, than where the subject is, in a great measure, indifferent to us. [3.1.1.1, 455]

Not only does morality have a specific purpose, but that purpose generates a particular kind of illusion: it makes “our speculations appear more real and solid.” The central role that morality plays in our lives can blind us to its conventional character, and can convince us that the moral facts we construct are instead phenomena that we discover. This is why attacks on conventionalism or skepticism often use the case of values as the supposedly decisive reductio: “You may be willing to countenance conventionalism about meaning, or about scientific justification, but surely you can’t believe that our moral principles are merely conventional! How could you engage in serious moral criticism if you believe that? What about slavery, etc.?” And this is so despite the fact that it is in the domain of morals that diversity in judgment is most salient. Hume’s ethical theory disarms this reductio. It is essential to Hume’s project to demonstrate how we construct the moral domain, and how that domain, in turn, constructs us as persons, as well as to show why we are so blind to this conventional status of our moral sentiments. Hume begins his investigation by noting that morality has at least an influence on the passions (later he will also argue for its origins in the passions as well), and so, he argues, cannot derive from reason: If morality had naturally no influence on human passions and actions, ’twere in vain to take such pains to inculcate it; and nothing wou’d be more fruitless

E t hi c s  

( 229 )

than the multitude of rules and precepts, with which all moralists abound. . . . morality is . . . supposed to influence our passions and actions, and to go beyond the calm and indolent judgments of the understanding. And this is confirm’d by common experience, which informs us, that men are often govern’d by their duties, and are deter’d from some actions by the opinion of injustice, and impell’d to others by that of obligation. [3.1.1.5, 457]

And near the conclusion of his account of ethics and political philosophy, Hume writes: Here then is the origin of civil government and allegiance. Men are not able radically to cure, either in themselves or others, that narrowness of soul, which makes them prefer the present to the remote. They cannot change their natures. All they can do is to change their situation, and render the observance of justice the immediate interest of some particular persons, and its violation their more remote. These persons, then, are not only induc’d to observe those rules in their own conduct, but to constrain others to a like regularity, and enforce the dictates of equity thro’ the whole society. [3.2.7.6, 537]

Together, we construct, institute, obey, and come to propagate and enforce customs that shape our behaviors and our attitudes. Once those customs are in place, it is in each of our interests to observe them, and to ensure that others do as well. Moreover, the stability of this system of custom is in our individual and joint interest as well. It is what makes society possible, and so the values that emerge from, sustain, and advance our social being. One could not hope for a more prescient anticipation of David Lewis’s own (1969) analysis of conventions, their values, and the mechanisms of their institutions. And here, we see reflections on the role of custom in British law in Hume’s thought about its role in morality. In each case, it is custom that provides the fuel for normative force; neither in moral nor in legal principles is there primitive normative force that grounds custom. When we put together these two passages that bookend Hume’s exposition of moral theory, we see the lineaments of an account that is strikingly similar in form to those we have encountered thus far in Books I and II, and one that, like those, takes the passions and custom as the basis. The structure of the account, like those we encountered in Books I and II, is also thoroughly Pyrrhonian. Hume does not argue for the normative force of ethical principles on the ground that they can be justified independently of our human and social nature. Nor does he argue that because they have no such independent ground that they are bereft of normative force. Instead, he grounds morality in our biological nature, in our conventions, in the

( 230 )  Books I and III

arts of politics; in short, he grounds morality in the positive side of the Pyrrhonian enterprise, with the understanding that this basis in custom is the only kind of basis that norms could have. Rather than justify our ethical practices by appeal to independent principles, Hume justifies our principles by appeal to custom, in both the biological and the social senses of that term. Moral sensibility has its roots in the passions, but to engender the society that is necessary to extend it to a full sense of justice, obligation, and respect for the law and the rights of others requires the establishment of complex sets of conventions. Those conventions are indispensable for enforcing its dictates, but also for inculcating the sensibility that makes moral life more than mere obedience to the law. Most importantly, they enable a way of taking up the world that reflects a cultivated “progress of sentiments” [3.2.2.23, 500]. Custom, as many of the passages in Book I we have encountered suggest, can indeed be a conservative force, inculcating and reflexively validating individual or collective habits. But, even though custom can be a conservative force, it need not be. We should recall that customs—​ prominently including those that govern reasoning and social life—​as we saw in ­chapter 2, also include practices of reflective criticism and responsiveness to others, and in this guise provide the foundation for both intellectual and moral progress. In this context, it is worth reflecting once again on the legal roots of these ideas, which illuminate both the conservative and the progressive aspects of custom. English and American jurisprudence rely on the principle of stare decisis, and for good reason. While the respect for precedent may indeed be a conservative force in legal reasoning, it also allows us to predict the consequences of our actions, to understand our rights and responsibilities, and to order our lives. Previous decisions warrant successive decisions, judgments, policies, and courses of action, and this is precisely the way that custom operates, warranting our behavior and our normative assessments, just as the past warrants our confidence in the course of the future, albeit defeasibly. Indeed, it is important that stare decisis is defeasible as well:  our customs regarding jurisprudence include customs of reasoning capable of overturning and rejecting precedent. Constitutions are similarly conserv­ ative, always constructed so as to be difficult to amend, but always with constitutional procedures for their own amendment—​customary ways of transforming custom. In each case, the conservative force functions more as ballast than as an anchor, conferring stability, but permitting progress. Once again, we must bear in mind that Hume’s purpose is neither to defend

E t hi c s  

( 231 )

nor to criticize particular customs, but to show how custom functions to constitute normative practices, and to demonstrate the richness of custom as a resource for explaining this dimension of our lives.5 Those sentiments, in turn, come to structure our intuitions, our interactions, and, reflexively, the institutions and customs that institute those conventions and maintain those sentiments in the first place. Moral customs, Hume argues, while grounded in original facts about human nature, exceed those facts, albeit in natural ways, as we construct our social units; nonetheless, they do not reflect antecedent moral facts of which we are aware and which justify them, but constitute those facts, in yet another instance of the skeptical inversion.6

5. I thank Kazunori Sawada for emphasizing the importance of this issue. Morality is also concerned with the restraint of some of our more self-​interested or nepotistic sentiments, and the cultivation of more socially advantageous ones. Even in these contexts, however, Hume regards our passions not as individualistic, but as social; passions are not to be subdued, but shaped. Miller (1980) puts the point nicely: [Hume’s] assumptions about human nature are not those of “possessive individualism.” At the same time it does not diminish the weight of those remarks; the love of gain remains a powerful and in extinguishable force in human affairs, a useful and yet dangerous passion which needs to be regulated properly if social order is to be preserved. (112) 6.  Once again, the structural affinities to the Madhyamaka Buddhist project are worth emphasizing: in each case a central diagnosis of common confusion about the human condition is the tendency to take that which we construct to be something we discovered; in each case, convention operates reciprocally, both as something we create and institute and as something that creates us as creatures capable of creating conventions. Korsgaard (1996, 60) also notes that Hume concludes in Book III that moral sensibility is self-​reinforcing: the moral sense approves of moral sensibility: It requires but very little knowledge of human affairs to perceive, that a sense of morals is a principle inherent in the soul. . . . But this sense must certainly acquire new force, when reflecting on itself, it approves of those principles, from whence it is deriv’d, and finds nothing but what is great and good in its rise and origin . . . not only virtue must be approv’d of, but also the sense of virtue:  And not only that sense, but also the principles, from whence it is deriv’d. [3.3.6.3, 619] Korsgaard contrasts this with the situation in which Hume regards the understanding at the conclusion of Book I. She notes correctly that unlike the moral sense, it cannot bear its own survey, and, by similar reflection on its own operations, concludes that it cannot endorse its own authority. I agree with Korsgaard that there is this asymmetry in Hume’s attitude toward these two faculties. But the asymmetry is situated in the context of a larger symmetry on which Korsgaard does not remark:  Hume does not conclude at the end of Book I  that we ought not reason; not only do both faculties naturally reflect on themselves, but in each case the ultimate ground of the warrant of each faculty lies in custom, and the psychological mechanisms that enable the reliance on each are located in the imagination.

( 232 )  Books I and III

2.  MORALITY AND THE SENTIMENTS Hume now turns to the relation between morality and reason on the one hand and the sentiments on the other. This is another point at which his Pyrrhonism comes to the fore, and where we see the homology between Hume’s treatment of reasoning and his treatment of morality, despite the fact that he argues that morality is not derived from reason: Since morals, therefore, have an influence on the actions and affections, it follows, that they cannot be deriv’d from reason; and that because reason alone, as we have already prov’d, can never have any such influence. Morals excite passions, and produce or prevent actions. Reason of itself is utterly impotent in this particular. The rules of morality, therefore, are not conclusions of our reason. [3.1.1.6, 457]

Hume offers several well-​known arguments to confirm this inference. First, he argues [3.1.1.12, 458] that while the faculty of reason issues in judgments that can be true or false, passions, volitions, and actions cannot be either true or false, or even reasonable or unreasonable, insofar as they are “original facts and realities, compleat in themselves.” They are therefore no part of reason (although they may be grounded in judgments which themselves might be true or false). Second, he argues [3.1.1.12, 459–​60] that in judging actions, errors regarding matters of fact may lead to foolish or ineffective action, but they do not lead to actions being regarded as immoral. Failings of morality, then, are not failings of reason, and so morality does not derive from reason. Finally, he argues [3.1.1.14, 460], if morality were grounded in reason, then judgments themselves would be the objects of moral approval or disapproval, as opposed to our actions or motives, and that since the failings of reason are represented in errors regarding truth or falsity, neither the subject matter of these judgments nor their importance in our affairs should be relevant to moral assessment: error would be vice, and truth, virtue.7 7. It is therefore hard to agree with Norton (1982) when he writes, “Hume consistently argues that moral evaluation requires the prior activity of reason, and Hume’s ideal of the morally acute individual is that of a person who can correct for certain normal and widely occurring biases, and thus judge matters, if not sub specie aeternitatis, then at least at very long range and according to rationally derived general rules” (95). Norton argues that Hume follows Hutcheson in giving a central role to reason in moral judgment; I think that it is in fact plain that this is one of the respects in which Hume departs from Hutcheson. He also departs from Hutcheson in regarding much of the moral landscape as constituted by our moral sensibility, as opposed to being tracked by it. (See also p.  150, where Norton argues that Hume asserts that we have moral knowledge of an independent domain of moral facts, and hence responds directly to moral skepticism.)

E t hi c s  

( 233 )

Now, it is not my purpose to defend any of these arguments. Indeed, none of them are very convincing.8 The important point for our purposes is that Hume does not regard morality as tracking independent matters of fact. This connects directly to his Pyrrhonism.9 Debates about moral facts will always require suspension, since no claim about convention-​ independent moral facts can be true. That does not, however, undermine the reality of morality as a phenomenon, or the point of moral customs. Moral customs, on Hume’s view, themselves constitute morality and require no grounding in antecedent moral facts. Hume makes this point in the following remark: Shou’d it be pretended, that tho’ a mistake of fact be not criminal, yet a mistake of right often is; and that this may be the source of immorality: I would answer, that ’tis impossible such a mistake can ever be the original source of immorality, since it supposes a real right and wrong; that is, a real distinction in morals, independent of these judgments. A mistake, therefore, of right may become a species of immorality; but ’tis only a secondary one, and is founded on some other, antecedent to it. [3.1.1.14, 460]

This is a nice point. Hume anticipates the reply that there are errors that count as instances of immorality:  errors regarding what is right.10 But, Hume points out in reply, a mistake regarding the right is not a mistake It is also worth noting that Hume reaffirms his commitment to this principle in the Dissertation: “It seems evident, that reason, in a strict sense, as meaning the judgment of truth and falsehood, can never, of itself, be any motive to the will, and can have no influence but so far as it touches some passion or affection” (5:1, 24). In this regard Hume departs from Shaftesbury, who argues that “’tis the known Province of Philosophy . . . to regulate our governing Fancys, Passions and Humours” (1964, 1:176). 8.  See Korsgaard (2009, 59–​61) for a critique of this view of the role of reason in normativity. Korsgaard does not, however, consider Hume’s broader positive account of the sources of normativity in this context, relying on a narrow, psychologistic understanding quite different from that I offer here. 9.  Norton (1982, 43–​54) argues that Hume’s ethics, unlike his epistemology, is antiskeptical, responding to a crisis of moral skepticism occasioned by the naturalism of Hobbes and Mandeville, and defending Hutcheson’s moral realism. I think that not only does this reading of Hume miss the uniformity of his method I  am concerned to identify in this volume, but it also mistakes Hume’s relationship to skepticism on both sides of the distinction Norton wishes to draw. On the epistemological side, Norton holds Hume to be attacking, rather than adopting, Pyrrhonism, relying both on Hume’s own misleading terminology and on Norton’s own rather nihilistic reading of Pyrrhonism (Norton 1982, 255–​69), ignoring the positive side of that approach to skepticism. On the ethical side, it misses seeing that to take ethical distinctions as arising from custom, as Hume does, is also a Pyrrhonian attitude. 10.  Compare Aristotle on the distinction between errors regarding the particular, which excuse an otherwise blameworthy act on the grounds of ignorance vs. errors regarding the universal, which count as vice, and cannot be cited as excuses. I can attempt

( 234 )  Books I and III

regarding facts that obtain independent of our moral judgments. While mistakes regarding the right are instances of immorality, they presuppose antecedent conventions consisting in collective judgments that themselves determine the moral domain. There are no primitive, convention-​ independent moral facts to which our judgments are hostage. To put the point in a more epistemological register, anticipating Hume’s next series of arguments, there are no custom-​independent facts in the world to which we are sensitive that, by themselves, constitute evidence for moral claims. Moral claims are, at bottom, claims about custom, just as the truthmakers for legal claims about rights or obligations are always in the end our customary practices. Hume offers a number of arguments to reinforce this conclusion. Again, they are not all entirely compelling, but for our purposes, the most important point is the conclusion Hume is after: that morality is not responsive to facts extrinsic to the moral framework itself, and that our moral framework is grounded in our passions, not in reason, and in the customs constituted by our response to those passions.11 He argues first [3.1.1.19, 464–​65] that the object of reason is always the relations between phenomena, and that relations neither between internal phenomena nor between external phenomena could serve as the objects of moral judgment. Internal phenomena fail because we cannot be judged morally on the grounds of our actions with regard to ourselves; external ones fail because then inanimate objects would be subject to moral appraisal on the grounds of the relations in which they stand. Hume draws these reflections together in the following remark: Morality, therefore, 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. [3.1.2.1, 470]

It is tempting to see this as a special observation about morality, and it is remarks like these that draw attention to the connection between Hume and Hutcheson.12 But we should not be too hasty. Remember Hume’s conclusion at the end of his first discussion of skepticism with regard to reason, “that all our reasonings concerning causes and effects are deriv’d from nothing but custom” [1.4.1.7, 183]. So we should be careful in characterizing to excuse my homicide on the grounds that I thought that his head was a pumpkin, but not on the grounds that I thought that splitting someone’s head is permissible. 11. See Cohon (2008, 20) for a perceptive discussion. 12.  But also note the impact of Shaftesbury, who also argues for the importance of the passions as motivators, and indeed as moral motivators that run counter to self-​interest:

E t hi c s  

( 235 )

Hume’s reference to the passions and to custom in the context of morality as a mere repetition of the moral sentimentalism of Shaftesbury and Hutcheson, despite its obvious debt to their ideas. They distinguished moral knowledge and reasoning from reasoning and knowledge of matters of fact on the grounds that, whereas the latter depend on the senses and reason, moral knowledge and judgment are grounded in sentiment. Hume assimilates moral reasoning to reasoning about matters of fact:  both are grounded in custom and the “sensitive . . . part of our natures.” The only difference that Hume sees is that reason is concerned with the discovery of matters of empirical fact and with the relations among ideas; morality is not. Their domains differ, but their structures are identical.13 There is a very deep connection indeed between the passions and moral sentiments. Not only does sympathy act on us to stimulate the moral sentiments as a passion, but the moral sentiments themselves, when seen at this structural level, are in fact species of passions; they must be, as they are not only motivators, but very strong motivators of behavior.14 And while Hutcheson would agree in this assimilation, it is Hume’s much more complex understanding of the nature of the passions themselves and of their intentional structure that makes this plausible in the case of moral attitudes. Hume then turns to the question of the nature of virtue and vice. You have heard it (my Friend!) as a common Saying, that Interest governs the World. But, I believe, whoever looks narrowly into the Affairs of it, will find that Passion, Humour, Caprice, Zeal, Faction, and a thousand other Springs, which are counter to Self-​Interest, have as considerable a part in the Movements of this Machine. (1964, 1:72) 13.  There is another difference:  as we noted above, on Hume’s account, reasoning about matters of fact is undertaken by a faculty which, were it to follow its own normative canon, would subvert that very canon, while moral reasoning is undertaken by a faculty that reflexively approves its own normative canon. Once again, though, once we consider the larger context in which both of these accounts are set, that difference appears to be a matter of (interesting) detail; not a structural difference, inasmuch in each case the foundations of the relevant normativity—​a normativity that we end up taking as regulative—​lie in custom mediated by the imagination. It is also important to bear in mind that this does not amount to moral irrealism (although, as I will argue in the final chapter, it does amount to moral fictionalism in an important sense). Hume is not denying that there are moral facts about which we can be right or wrong, but is instead providing an account of the nature and origin of those facts—​in custom, not in custom-​independent nature. See Cohon (2008) for a similar analysis. 14.  Cohon (2011) argues for an important structural relation between the moral sentiments and the indirect passions. Like the indirect passions, she argues, the moral sentiments each have both an object and a cause, and these are always different from one another. The cause of a moral sentiment is the feeling I experience when sympathy (whether natural or cultivated) conveys to me the feelings of others. The object of the moral sentiment is the character, act, or institution of which I approve or disapprove. She offers the following example:

( 236 )  Books I and III

3.  VIRTUE AND VICE So virtue and vice are determined by the passions and custom. We have yet to see how that goes in detail. The first task is to characterize virtue and vice, per se, and the second is to explore the mechanisms by which they are constituted. Here is how Hume introduces the analysis of virtue and vice. An action, or sentiment, or character is virtuous or vicious; why? because its view causes a pleasure or uneasiness of a particular kind. In giving a reason, therefore, for the pleasure or uneasiness, we sufficiently explain the vice or virtue. To have a sense of virtue is nothing but to feel a satisfaction of a particular kind from the contemplation of a character. The very feeling constitutes our praise or admiration. . . . We do not infer a character to be virtuous, because it pleases: But in feeling that it pleases after such a particular manner, we in effect feel that it is virtuous. The case is the same as in our judgments concerning all kinds of beauty, and tastes, and sensations. Our approbation is imply’d in the immediate pleasure they convey to us. [3.1.2.3, 471]

There is a lot going on in this short passage. The first two sentences concern the ground for the attribution of virtue or vice to a character, motive, or action. We call a person or an action virtuous when she or it gives us a certain kind of pleasure (we have yet to specify what kind); vicious when she or it causes a certain kind of uneasiness (again, yet to be specified). This comes directly from Hutcheson. But this is simply a claim about the epistemology of virtue—​about what induces us to judge that someone or something is virtuous or vicious. Now let us see what Hume does with this. What causes me to approve Mother Theresa’s character is the feelings of relief and joy experienced by the poor of Calcutta as a result of her compassion, transmitted to me via corrected sympathy; but the intentional object of my feeling of approval is not the poor of Calcutta or their feelings, but Mother Theresa’s character. (2011, 177) Moreover, she notes, the moral sentiments are even more indirect than the indirect passions, inasmuch as they typically involve a triadic, as opposed to a dyadic, relation between the subject and object. She continues: Where love, for example, is caused in me by some pleasing trait or possession of another person but is redirected onto that very person who possesses it, moral approval is caused by a pleasure I get from sympathizing with one person but very often is redirected onto some other person whose character caused the pleasure. . . . Thus while pride typically involves one person (oneself), and love typically involves two people (the lover and the beloved), moral approval in many cases requires three: the observer, the person whose trait is evaluated, and the beneficiary of that trait. (2011, 177) I am sympathetic with this reading, but my own analysis is independent of it.

E t hi c s  

( 237 )

He first notes that the praise or blame—​that is, the attribution of moral value—​just is the feeling of pleasure or pain, thus locating praise or blame firmly in the passions. He emphasizes that “we do not infer a character to be virtuous because it pleases . . . we in effect feel that it is virtuous.” Virtue and vice, then, are simply whatever properties that we are tuned to feel in particular ways: a virtue is a property our passions tune us to approve; a vice is one our passions tune us to disapprove; our passions, that is, are not originally tuned to virtue and vice as independent phenomena, as Hutcheson would have it; instead, what counts as a virtue or a vice is determined by the particular passions that quality excites.15 This is yet another example of a skeptical inversion of the order of explanation, and entails that we can’t ask about the accuracy or inaccuracy of these passions as original aspects of our nature, but only about what they deliver. Hume then turns to the question about the precise nature of that property and its relation to sentiment: It may now be ask’d in general, concerning this pain or pleasure, that distinguishes good and evil, From what principles is it derived, and whence does it arise in the human mind? To this I reply, first, that ’tis absurd to imagine, that in every particular instance, these sentiments are produc’d by an original quality and primary constitution. For as the number of our duties is, in a manner, infinite, ’tis impossible that our original instincts should extend to each of them, and from our very first infancy impress on the human mind all that multitude of precepts, which are contain’d in the compleatest system of ethics. [3.1.2.6, 473]

The first thing to note is that Hume recuses himself from producing a precise account of the nature of the pleasure and unease responsible for moral judgments, and this is fair enough. After all, the nature of the pleasure we experience when we hear good music is different from that we experience when we contemplate the success of our children, and different again from that we experience lying on a beach in the summer sunshine. We might

15. But this property must be a standing trait of the individual, as Hume writes to Hutcheson: I cannot forebear recommending another thing to your Consideration. Actions are not virtuous nor vicious; but only so far as they are proofs of certain Qualitys or durable Principles in the Mind. This is a Point I shou’d have establish’d more expressly than I have done. Now I desire you to consider, if there be any Quality, that is virtuous, without having a Tendency either to the public Good or to the Good of the Person, who possesses it. If there be none without those Tendencys, we may conclude, that their Merit is derivd from Sympathy. (Letter 13, to Francis Hutcheson, in Hume 1932, 32)

( 238 )  Books I and III

even say that all that these pleasures have in common with one another is that they are pleasures. So, Hume is surely entitled to the view that moral pleasure is another “original existence,” another kind of pleasure primitively described as that which derives from the contemplation of character, motive, or action, and mutatis mutandis for the relevant unease. And he is also entitled to the view that it is a brute fact about human nature that just as we take pleasure in art, sunshine, and the accomplishments of our children, we take pleasure in character, motives, and actions. We could even tell a plausible evolutionary psychological just-​so story about selection in favor of this kind of pleasure and unease on Hume’s behalf: those who are naturally inclined to participate in social conventions, and who tend to experience pleasure or pain in response to the same kinds of character traits and human circumstances as their fellows will be more attractive, better-​liked, more successful, and so will enjoy selective reproductive advantages, passing on this tendency to affective and behavioral convergence to their offspring. The power of custom can thus be enhanced both socially and biologically, with each dimension supporting the other. The second point in this passage is equally important, and takes us to the next principal topic for this chapter. Hume points out that it is impossible—​even if susceptibility to pleasures and states of uneasiness in response to the contemplation of character, motive and action is natural—​ for the relation between each type of stimulus and the corresponding moral response to be original, that is, to be innately determined, as, for instance, Hutcheson thought. There are simply far too many motives, duties, characters, and circumstances—​and, we might add, there is far too much intercultural variation in moral assessment—​for these to be original connections. It is far more likely, he thinks, that while there may be some original dispositions of this kind, there is also an original disposition to learn new ones, and that artifice and convention intervene to yield a complete “system of ethics.”16 It is to that interaction that we now turn.

4.  NATURE AND ARTIFICE Hume’s clearest statements about the relation between nature and artifice and concerning the role of social convention in human nature occur in 16. Once again, compare this to the case of language. An original faculty of our mind determines that we will (in the course of normal development) acquire a human natural language—​that is, custom in the individual sense; but what language we will learn depends entirely upon the community in which we are raised, and the collective customs regarding syntax, semantics, phonology, and pragmatics enforced in that community.

E t hi c s  

( 239 )

the context of the articulation of his ethical theory. It is quite common to draw a distinction between the natural and the artificial, and to draw that distinction on the grounds of whether the phenomenon in question owes its origin to human activity. But Hume properly calls this distinction into question, pointing out twice in Book III that it is natural for human beings to engage in a wide variety of artifice. Simply because a behavior or an institution is not original in Hume’s sense is no reason to take it to be less than natural: But nature may also be opposed to artifice  .  .  .  and in this sense it may be disputed, whether the notions of virtue be natural or not. We readily forget, that the designs, and projects, and views of men are principles as necessary in their operation as heat and cold, moist and dry. [3.1.2.9, 474]

With this distinction in hand, Hume can assert that our sense of virtue and vice is entirely natural (here, again, he is in agreement with Hutcheson), but also assert that natural artifice, enabled by our natural proclivity for such artifice and necessitated by the circumstances of life, leads to an enormous variety in the specific instances of virtues and vices that we recognize (again compare language). Custom in the sense of innate habit—​the habit in this case of producing artifices that extend the moral sentiments that are in us by custom in this very sense—​and custom in the sense of convention hence operate inseparably to generate our moral sense.17 The most important of the acquired virtues for Hume is the virtue of justice, in part because it so richly exemplifies this interplay of custom, and in part because it is the virtue that underpins the stability of society, and therefore the very context in which the conventions that constitute us as persons can arise.18 In this case, too, original capacities, artifice, and convention conspire to produce the norms that in turn induce a confluence of behavior, judgments, and social practices. 17. But artifice can get it wrong. In an amusing comment on this possibility, Hume writes to the Marquise de Barentane: I saw today an Italian abbé, and talked to him about the Court of Florence. He says that it has become a very disagreeable place, and that nothing can exceed the narrowness of mind in the Archduke and Archduchess. The princess rubs off, with her own hands, the paint and patches from the faces of the ladies, and makes them presents of tippets to cover their bosoms. I beseech you, never go to a place where you must become virtuous by constraint, lest you should take an inclination to become otherwise. You may, perhaps, be allowed, as Ambassadress to France, to keep your rouge; but are you sure that you may not take quarrel with virtue, when you see it accompanied by so much folly and ridicule? (Letter 349, August 29, 1776, in Hume 1932, vol. 2, 84–​86) 18.  As Baier (1997) notes, on Hume’s account our collective sense of justice is a constructed, or as Baier puts it, a “contrived commonality” (47). There is hence a

( 240 )  Books I and III

While justice is not an original virtue, it is a virtue absolutely essential to the processes that enable the vast range of artificial virtues human intercourse requires. After all, without justice, we have no commerce and no civic life; without commerce and civic life no arts, education, or industry; without these, no recognizable human culture or progress, and without these, no life we could recognize as human.19 We therefore bring ourselves into existence as persons through artifice, and might indeed recognize ourselves as natural artifacts: I have already hinted, that our sense of every kind of virtue is not natural; but that there are some virtues, that produce pleasure and approbation by means of an artifice or contrivance, which arises from the circumstances and necessities of mankind. Of this kind I assert justice to be; and shall endeavor to defend this opinion. [3.2.1.1, 477]

And again: To avoid giving offence, I must here observe, that when I deny justice to be a natural virtue, I make use of the word, natural, only as oppos’d to artificial. In another sense of the word; as no principle of the human mind is more natural than a sense of virtue; so no virtue is more natural than justice. Mankind is an inventive species; and where an invention is obvious and absolutely necessary, it may as properly be said to be natural as any thing that proceeds immediately from original principles, without the intervention of thought or reflexion. Tho’ the rules of justice be artificial, they are not arbitrary. Nor is the expression improper to call them Laws of Nature; if by natural we understand what is common complex feedback loop in the account of the role of custom in our construction of our moral life. Through childrearing practices, collective social discourse, and other discursive practices, we construct a shared account of virtue and vice, and then introduce customs of approbation and disapprobation that reflect and reinforce these customs. Without these practices that institute and maintain conventions, morality itself would be impossible, as would society and human life as we know it. Fortunately, as Hume also emphasizes, part of our human nature is a natural tendency to institute and maintain these complex networks of custom. As Baier also points out in the same context (1997, 48), part of the psychological foundation of these practices is our natural sympathy, our tendency to perceive and to react sympathetically to the affective states of others. And these sympathetic reactions are powerful enough to inspire action. It is for this reason that, pace Darwall (2002), Hume does have a morality that concerns our actions and responsibilities to one another—​a morality of participation—​and not merely a morality of spectatorship, or of moral connoisseurship. 19. Hume emphasizes these points as well in EPM 3.13 (SBN 188). His discussion of justice in the Enquiry Concerning the Principles of Morals does not differ from that in the Treatise in content or in method.

E t hi c s  

( 241 )

to any species, or even if we confine it to mean what is inseparable from the species. [3.2.1.19 484]

Cohon’s reading is for the most part consistent with my own. She writes: The commonsense conception of virtue is rooted in human nature and based on our natural proclivities to prefer and to approve (and thence to judge good), so we may call it the natural conception of virtue. It fits the character traits that Hume labels “natural virtues.”  .  .  .  Successful social life, however, requires us to invent and acquire other character traits that are in tension with these natural proclivities, traits that involve impartiality and enable us to cooperate with people to whom we have no affectionate ties. Among these are honesty in regard to property, faithfulness to promises, and respect for the civil authorities. (2008, 3)

I agree with Cohon entirely regarding the distinction between the natural and the artificial virtues, and regarding her claim that on Hume’s view (and in reality as well), it is the demands of social life that drive the (natural) construction, cultivation, and mechanisms for approval of the artificial virtues. I disagree only with her remark that these artificial virtues “are in tension with” the natural virtues. While they may, in some sense, mitigate some aspects of the natural virtues, such as partiality (and this is why they are so important in social life), on Hume’s account this does not reflect a tension between these sets of virtues so much as a modulation of the natural virtues in order to create a set of artificial virtues that augment and extend them.20 After pointing out [3.2.1.8–​9, 479] the necessity of a powerful sense of justice for social and commercial cohesion, Hume notes that there is no

20. Winch (1996) connects this point nicely with Hume’s reaction to Mandeville, and to his reliance on custom as a model for the artificial virtues: In their philosophical and historical treatments of law and government, and in their accounts of how morals were shaped by society, Hume and Smith rejected [Mandeville’s] doctrine as antagonistic to their respective moral psychologies based on the primacy of the social passions. For Hume, perhaps, the task of differentiating his position from that of Mandeville was more acute, partly because he had advertised his pursuit of a similarly naturalistic anatomy of morals in his Treatise of Human Nature. . . . Hume's solution, in the case of justice, was to have recourse to “artificial” virtues based on self-​interest and regard for public utility. . . . But if justice was . . . an “artificial” virtue, this implied that it was the product of habit and convention, entailing a learning process that was capable of surmounting the limitations of our benevolence in a world of scarcity and competition between “mine” and “thine.” (67–​68)

( 242 )  Books I and III

original disposition toward justice, or even toward the public interest that we can recognize justice to serve. And the natural affections we feel toward one another are far too weak and limited to support justice [3.2.1.10–​ 14, 480–​82]. Justice must, Hume concludes, arise as an artifice.21 It then behooves us to understand the nature of that artifice. Hume observes that we do have an original disposition to affection for one another, but as a good Newtonian, he appeals to the inverse square law that governs its operation, emphasizing that it declines in power rapidly with distance from ourselves: A man naturally loves his children better than his nephews, his nephews better than his cousins, his cousins better than strangers. . . . Hence arise our common measures of duty, in preferring the one to the other. Our sense of duty always follows the common and natural course of our passions. [3.2.1.18, 483–​84]

Nonetheless, as he notes here, despite its partiality, that natural affection is linked closely to our sense of morality. We take ourselves to have stronger obligations to our children than our nephews and nieces, and stronger obligations to them than to distant strangers. Agent-​neutral conceptions of morality, while they appeal to reason, rarely track our moral intuitions or judgments about one another’s conduct. Therefore, as Hume points out (closely following Mandeville in this regard), just as self-​interest, however much it may be moderated by concern for others, is a powerful motivator, and so requires restraint if civil society is to be possible, our natural concern for others, insofar as it is directed most powerfully to those close to us, requires restraint. And whether these passions are taken as virtuous or vicious, the genius of conventions of justice is that they can be regulated so as to serve the general good, and to be in the interests of all who participate in the social institutions they enable [3.2.2.13, 492ff.]. 21. In the Enquiry Concerning the Principles of Morals, however, Hume suggests that this artifice is not all that difficult to achieve, at least in the case of those of us who are reasonable. He notes: It has often been asserted, that, as every man has a strong connexion with society, and perceives the impossibility of his solitary subsistence, he becomes, on that account, favourable to all those habits or principles, which promote order in society, and insure to him the quiet possession of so inestimable a blessing. As much as we value our own happiness and welfare, as much must we applaud the practice of justice and humanity, by which alone the social confederacy can be maintained, and every man reap the fruits of mutual protection and assistance. (EPM 5.5; SBN 215) So Hume argues in this context that while justice may require artifice, it is an artifice that is made easier by its obvious attractiveness. He certainly makes eighteenth-​ century Edinburgh appear a reasonable place indeed in the Enquiry. The Treatise is not quite so sanguine in this matter.

E t hi c s  

( 243 )

In an extended discussion [3.2.2.3–​17, 485–​95], Hume considers and rejects social contract theory on the grounds that it fallaciously attempts to ground justice and social obligation in promises, because social contract theorists fail to note that promises require antecedent customs in order to be binding. The contract theory, he argues, thus attempts to explain our customs based on antecedent and inexplicable norms, rather than properly explaining norms by grounding them in the binding force of custom. Instead, he points out that justice emerges as a virtue from a kind of bootstrapping:  we recognize the benefits to be derived from justice, and the obstacles to its cultivation. These benefits can be recognized only in civil society, when a certain modicum of trust and mutual affection can be presumed, at least among members of particular groups, and in which discourse and agreements are possible.22 Recognizing the benefits of justice, we collaborate to construct the social, political, and judicial institutions as well as the discursive and pedagogical practices that encourage, facilitate, and enforce it. The more benefits we reap, the more admiration we have for those who exhibit justice, and the more value we see in encouraging and enforcing it. Justice eventually becomes a widely shared and self-​reinforcing source of pleasure, an artifice whose value we all recognize, a constant motive to action, and an artifice which we all promote because of its value to each of us.23 That value exceeds our own immediate interest and situation and consists in part in the infrastructure of civil life it makes possible, an infrastructure always visible to us: 22. Note once again the uniformity of method. Hume is always concerned to reject any Archimedean position outside of custom on which custom might be grounded, or from which it might be assessed (while insisting on the power of one custom to ground or provide the grounds for the assessment of another). Locke is in Hume’s sights here precisely because Locke’s project in the Second Treatise is in one sense the same as Hume’s—​to provide an account of the origins of obligations and rights. On Hume’s view, though, Locke’s explanation simply begs the question: by presupposing that a contract could be binding antecedent to any social institutions, he seeks just such an Archimedean fulcrum and proposes what in the end must be a nonnatural account of the foundations of value. Hume, on the other hand, insists that the explanation of the origin of obligation can only be found in the natural propensity we have to constitute, to enforce, and to conform to social customs; those customs institute the sense of obligation and warrant our moral judgments; they thus make possible contracts and promises, and cannot rest on an antecedent institution of contractual obligation. 23.  Indeed, as Baier (2010b) points out, without a considerable web of trust, the societies that constitute the foundation not only of our material well-​being, but also of all of the other conventions that make human life possible, would not be possible. This web of trust that binds us, as Hume sees, extends well beyond the political domain to the epistemic domain, in which so much of what we know relies both on epistemic conventions into which we trustfully enter and on our abilities to trust one another as sources of knowledge. (See also Korsgaard 1996 and Townley 2003.)

( 244 )  Books I and III Thus bridges are built, harbours open’d; ramparts rais’d; canals form’d; fleets equip’d; and armies disciplin’d; every where, by the care of government, which, tho’ compos’d of men subject to all human infirmities, becomes, by one of the finest and most subtle inventions imaginable, a composition, that is, in some measure, exempted from all these infirmities. [3.2.7.8, 539]

The subtle invention is not government per se and not justice alone, but the entire web of cultivated moral sentiments, extended and maintained by custom that constitute our collective lives and by the recursivity of that custom and its effects that in turn makes possible the trust that is the necessary condition of civil society. The idea here of their impact on each of us through their impact on the whole derives directly from Mandeville. But the conviction that in the institution of justice, self-​interest and public interest coincide—​that we come to have a direct regard for the public interest in part in virtue of this coincidence and in part because of the cultivation of the virtue of justice that leads to our taking it as an independent motive—​ has antecedents in Shaftesbury, who writes: If the Affection towards private or Self-​good, however selfish it may be esteem’d, is in reality not only consistent with publick Good, but in some measure contributing to it; if it be such, perhaps, as for the good of the Species in gen­ eral, every Individual ought to share; ’tis so far from being ill, or blameable in any sense, that it must be acknowledg’d absolutely necessary to constitute a Creature Good. (1964, 2:13)

Hume unites these two strains of thought. Nonetheless, the important difference between the original and the artificial virtue raises a serious problem that Hume must solve: while natural virtues—​in normal cases—​are immediate, prima facie motivators to virtuous action, artificial virtues, such as justice, often fail to motivate us without a good deal of additional social apparatus. We don’t require a lot of civics education or legal apparatus to get us to care for our children or to be kind to our friends; we do need it to prevent fraud, theft, and other antisocial criminal activity. (See Baier 2010b.) In the next section, Hume works out in detail how this natural artifice arises and is maintained.

5.  THE DOUBLE ROLE OF CUSTOM Hume sums up the results of his ruminations on the purposes of and obstacles to the virtue of justice as follows:

E t hi c s  

( 245 )

Here then is a proposition, which, I think, may be regarded as certain, that ’tis only from the selfishness and confin’d generosity of men, along with the scanty provision nature has made for his wants, that justice derives its origin. [3.2.2.18, 495]

Moreover, he notes, not only does justice arise from the exigencies that scarcity and limited generosity impose, but it is not original, and must arise from some artifice. Fortunately, however, humans are natural artisans of convention. Custom hence enters the picture in its sense as convention: We may farther confirm the foregoing proposition, that those impressions, which give rise to this sense of justice, are not natural to the mind of man, but arise from artifice and human conventions. [3.2.2.21, 496]

Hume concedes [3.2.2.22–​23, 496–​97] that not every act of justice serves a positive purpose. Justice may sometimes run against both individual and collective interest, as when I repay a loan to a drug addict, who will only spend the money on heroin. Nonetheless, Hume points out, as an institution, justice is undeniably not only beneficial, but also absolutely necessary to society, and its manifold benefits are apparent to anyone who bothers to consider the issue:24 However single acts of justice may be contrary, either to public or private interest, ’tis certain, that the whole plan or scheme is highly conducive, or indeed absolutely requisite, both to the support of society, and the well-​being of every individual. . . . Property must be stable, and must be fix’d by general rules. Tho’ in one instance the public be a sufferer, this momentary ill is amply compensated by the steady prosecution of the rule, and by the peace and order, which it establishes in society. And even every individual person must find himself a gainer, on ballancing the account; since, without justice, society must immediately dissolve, and every one must fall into that savage and solitary condition, which is infinitely worse than the worst situation that can possibly be suppos’d in society. [3.2.2.22, 497]

But, how are we to inculcate not only the understanding of the benefits of, but also the impulse to, and the feeling of approbation of justice that are

24. Baier (2010b) argues that Hume bids fair to be the first to deploy something like game theory to understand social conventions, with his account of justice as a demonstration to show that the adoption of and adherence to a cooperative convention beats unregulated competition. See also Korsgaard’s discussion of Hume on the “sensible knave” and the free rider problem (1996, 56–​58).

( 246 )  Books I and III

necessary for it to be more than a good idea, but a true virtue? Here, Hume turns to custom in its second guise, to human nature in the story of a “progress of sentiments,” a progress that essentially involves the operations of the imagination. Original human nature provides us with the raw material in natural sympathy, a powerful force that is both innate and that we innately approve of in the manner appropriate to virtue.25 Nonetheless, since that force falls off rapidly with perceived interpersonal distance, it can hardly be expected to bind us to act for the benefit of casual acquaintances or strangers when our own interests conflict with theirs, let  alone to an abstraction such as society at large. Nor could we originally ever find ourselves approving of acts of justice that require impartiality or the setting of the interests of the distant above those of the proximate. Here is where the imagination and artifice enter the picture. The key to the extension of natural sentiment into a concern for the well-​being of all, including the distant and the society at large, is to come to see others as close, as related to oneself. This requires a kind of moral education, sthe cultivation of a kind of moral vision in which we imagine ourselves as closely related to others, hence taking their interests into consideration as we would those of our own families or associates.26 This cultivation of moral vision through imaginative exercise is familiar to us. We see it whenever we are solicited for charitable contributions. The use of images, of biographies, and other devices to create a sense of relatedness or proximity is what opens our natural sympathy to those who otherwise would be beyond its bounds. This generates a willingness to take the well-​being of those who, although distant, are now perceived as proximate, as a motive for action, and equally importantly, leads us to approve of that

25. As Schneewind (1983) puts it, Hume rejects Hobbes’s claim that persons are universally egoistic in their motivation, and insists “that we are not wholly self-​seeking, that we can take immediate pleasure in the flourishing of others” (6). 26.  The affinities of this approach to ethical development as centrally involving a cultivation of a kind of moral vision is strikingly akin to that we see in Buddhist philosophers such as Āryadeva and especially Śāntideva. See Cowherds (2015) and Garfield (2010, 2012, 2015) for discussion of this issue. And the emphasis on the role of the imagination echoes Buddhaghosa’s, as well as Śāntideva’s accounts of the necessity of imagining, for instance, that other sentient beings have each been our mothers in previous lives, as well as our mortal enemies, as methods for cultivating a sense of impartial concern for others. This affinity is even more striking if we consider this passage in the Dissertation: “Compassion frequently arises, where there is no preceding esteem or friendship; and compassion is an uneasiness in the sufferings of another. It seems to spring from the intimate and strong conception of his sufferings; and our imagination proceeds by degrees, from the lively idea to the real feeling of another’s misery” (2007, 3.7). Hume could be writing a Buddhist meditation manual at this point.

E t hi c s  

( 247 )

motivation as virtuous. Justice is born.27 And justice is born in Hume’s account not as an abstract moral fact that we come to recognize, but as the natural outcome of our human nature given the boundary conditions of our collective life, an outcome that is nonetheless morally binding, since conventions of this kind are exactly that in which morality consists, and are precisely the kinds of institutions that—​as his reflection on customary law would have taught him—​bring rights and duties into existence.28 So, just as we not only feign, but believe in the external world; and just as we not only talk about, but rely on causal reasoning; we not only theorize, but ground our affective life on a sense of our own identity and our place in the social matrix that constitutes that identity. These facts and practices, Hume argues, are in turn grounded not in reason or the senses, but in the imagination and in the blind passions. That is, we are simply wired to react this way when in appropriate social circumstances, in virtue of our fundamentally social nature and innate responsiveness to our conspecifics. And this is in part why custom—​in the conventional sense—​can bind us in the way that it does. We might say that we have evolved—​as social creatures—​to take what is normal to be normative. Hume’s naturalization of normativity (and, as we noted above, not only in the moral domain, but in the epistemic as well) is accomplished by noting that as a matter of psychological and sociological fact, we are responsive to custom and naturally constitute customs that regulate our collective lives, even as we instinctively take the deliverances of those customs to have some kind of transcendental force they never could have. And, as I emphasized earlier, this is not to commit some kind of “naturalistic fallacy.” The idea that normativity is built into the very idea of custom was taken for granted by Hume and his contemporaries and is a bedrock idea in the Treatise. While customs might appear to be mere regularities, they are regularities with real normative force.29 27. And here we see a further affinity of Hume’s account to Buddhist ethics: the remarkable similarity of his account of justice to Buddhist accounts of the four divine attributes of karuṇā, metta, upekṣa, and muditā, or care, affection, nonegocentrism, and pleasure in the virtue of others. Just as in the Buddhist context, these attributes are cultivated by the imaginative extension of original affective impulses, in a process that constitutes us as morally mature individuals, justice is grounded in original impulses that are imaginatively extended to constitute us as citizens. 28. See also Cohon (2008) for a discussion of the role of parents and politicians—​ our networks of friends and family, and the official ideology of our communities—​ in achieving the “progress of sentiments” that Hume takes to constitute the path to moral maturity, a metaphor again strikingly reminiscent of the account of the development of the affective virtues on the path to awakening characterized by Śāntideva. 29. See Rorty (1982, 174–​175) for a clear exposition of the role of sympathy in the constituting the normative force of the moral sentiments, and a demonstration that

( 248 )  Books I and III

This is how custom and the imagination enable us to extend our domain of concern. Moreover, our approval of that extension is also grounded in the operation of the imaginations and in our passions; this is why morality is so self-​confirming. Hume is absolutely explicit about this progress. It is natural, inasmuch as we have both an original natural sympathy and an original capacity to extend that sympathy through the imagination. This is custom in its sense as habit. But is also natural—​but not original—​in that that sympathy is extended by that most natural of human activities, the artifice of language, convention, and collective action.30 Our very social institutions and our public discourse are shaped to accomplish this artifice: Tho’ this progress of the sentiments be natural, and even necessary, ’tis certain, that it is here forwarded by the artifice of politicians, who, in order to govern men more easily, and preserve peace in human society, have endeavour’d to produce an esteem for justice, and an abhorrence of injustice. . . . The utmost politicians can perform, is, to extend the natural sentiments beyond their original bounds; but still nature must furnish the materials. [3.2.2.25, 500]

And in the end, Hume, like Aristotle, sets another important natural human activity, childrearing, at the basis of the establishment of this invaluable custom: As publick praise and blame encrease our esteem for justice; so private education and instruction contribute to the same effect. For as parents easily observe, that a man is the more useful, both to himself and others, the greater degree of probity and honour he is endow’d with; and that those principles have greater force, when custom and education assist interest and reflexion: For these reasons they are induc’d to inculcate on their children, from their earliest infancy, the principles of probity, and teach them to regard the observance of those rules, by which society is maintain’d, as worthy and honourable, and their violation as base and infamous. By this means the sentiments of honour may take root in their tender minds, and acquire such firmness and solidity, that they may fall little short of those principles, which are the most essential to our natures, and the most deeply radicated in our internal constitution. [3.2.2.26, 500–​501]

Hume does not so much argue that ethical judgment is nonrational, as that its rationality is grounded in the passions and natural sympathy, together with the demands of social life. 30.  And, as Abramson (2015) argues—​defending Hume against Darwall’s (2002) charge that Hume’s moral theory is excessively detached and aesthetic, a theory for assessment, but not one that engages action or responsibility—​many of the sentiments to which Hume refers arise and motivate us not in disinterested contemplation, but in the course of social interaction.

E t hi c s  

( 249 )

The Humean image of virtue is very much that of something “deeply radicated” in our nature, but for all of that, it is natural, because we are the kinds of beings who create our own nature through the artifice that is natural to us.31 We come to be who we are in the social order because it is only in the social order that the conventions that are necessary to our constitution can be instituted. And it is the most natural thing of all for us to live in society, that domain in which the distinction between nature and artifice vanishes, and custom seamlessly traverses the boundary between the individual and the collective.32 Moreover, as Baier (2010, 67) points out, we come thereby to see that it is prudent to cultivate and to observe artificial virtues like justice. In becoming just or honest, we not only fulfill an obligation we have come

31. Garrett (2007, 11) puts this point well: “On Hume’s view, then all moral values, while real, are entirely within nature. Moral properties are discovered, as such, through the natural operations of a moral sense whose operation is refined in the course of experience and reflection. The normative character of these properties is determined by the role that, so detected, they play in human affairs; and their ability to do so is explained by the science of man.” I would only add that that “science of man” is as much social as it is narrowly psychological. Capaldi (1966, 137), on the other hand, argues, “In place of a normative conception, Hume holds the view that ethics is an empirical science. If ethics is an empirical science, then no ethical theorist has to explain the particular ethical views of any person or community.” I  certainly agree with the first sentence, but the conditional that follows is surely false: as I have argued, Hume thinks that the empirical science that undergirds ethics is as much social as it is individualistically psychological, and so an ethical theorist must be able to point to the customs that a society adopts and show how they are advantageous to that society and are inculcated in the members of that society in order to explain them. This Hume himself does when he considers the customs of his own society. Hardin (2007, 20) also defends a psychological and naturalistic reading of Hume’s ethics. 32.  Hume makes this point beautifully in the Enquiry Concerning the Principles of Morals: It appears, also, that in our general approbation of characters and manners, the useful tendency of the social virtues moves us not by any regards to self-​interest, but has an influence much more universal and extensive. It appears that a tendency to public good, and to the promoting of peace, harmony, and order in society, does always, by affecting the benevolent principles of our frame, engage us on the side of the social virtues. And it appears, as an additional confirmation, that those principles of humanity and sympathy enter so deeply into all our sentiments, and have so powerful an influence, as may enable them to excite the strongest censure and applause. (EPM 5.44; SBN 231) This idea follows Mandeville in part, who refers in his account of the origins of justice to the “Satisfaction that ensues upon [Virtuous Action, which] consists in a certain Pleasure he procures to himself by Contemplating his own worth” (1999, 92). But we must also note that here Hume departs significantly from Mandeville, as he does not argue that the social virtues are entirely egoistic. We often take the hallmark of a moral sensibility to be the absence of egoism; Hume agrees. His account of the nature

( 250 )  Books I and III

to recognize, but we also become more attractive collaborators, and hence more successful individuals, than we could hope to become as free riders. It hence becomes prudent overall to adopt a policy that might appear, on occasion, to lead us to act against what we might mistakenly see as our own narrow self-​interest.33 When Hume turns to promises, and to his most direct critique of the idea that government is founded upon a prior contract, he also adverts to the crucial role of convention in obligation: That the rule of morality, which enjoins the performance of promises, is not natural, will sufficiently appear from these two propositions  .  .  .  , viz. that a promise wou’d not be intelligible, before human conventions had establish’d it; and that even if it were intelligible, it wou’d not be attended with any moral obligation. [3.2.5.1, 516]34

Hume’s account of property and possession, and of the many virtues he takes as examples, follows this pattern. Custom lies at the basis of all and genesis of justice does not paint this virtue as covertly selfish, but rather shows how the interplay of original egoistic intentions, genuine original concern for others, our fundamentally social psychological instincts, and the collective conventions we are capable of constituting in response to these impulses—​all manifestations of custom—​ conspire to enable us to become beings who genuinely care for others and who genuinely value care for others. The “progress of sentiments” Hume characterizes is in the end the process by which we humanize one another. 33. And as Hume points out [3.2.5.5, 518], our motive to discharge a duty can never be a sense of duty, on pain of regress. The artifice of justice, as well as the artifice involved in all artificial virtues, is to construct a set of customs that constitute a framework in which our original passions can provide the motivation for behavior that benefits the community and that can also be clearly seen to be in the interest of the actor. This Hume inherits from Mandeville. But, as Cohon (2006) also notes, Hume’s position is more sophisticated than Mandeville’s, as he understands justice as conventional and artificial, but not therefore the result of deliberate manipulative schemes. See also his remark in a letter to Hutcheson: You are a great Admirer of Cicero, as well as I am. Please to review the 4th Book, de finibus bonorum & malorum; where you find him prove against the Stoics, that if there be no other Goods but Virtue, tis impossible there can be any Virtue; because the Mind woud then want all Motives to begin its Actions upon: And tis on the Goodness or Badness of the Motives that the Virtue of the Action depends. This proves, that to every virtuous Action there must be a Motive or impelling Passion distinct from the Virtue, & that Virtue can never be the sole Motive to any Action. You do not assent to this; tho’ I think there is no Proposition more certain or important. I must own my Proofs were not distinct enough, & must be alterd. You see with what Reluctance I part with you; tho’ I believe it is time I shoud ask your Pardon for so much trouble. (Letter 13, to Francis Hutcheson, in Hume 1932, 35) 34. Here, Hume is a bit sloppy: he is clearly using natural to mean, in his sense, original. He has already argued that human artifices, including such things as promises even though they are artificial, are natural. The point that their obligatory force and

E t hi c s  

( 251 )

morality, but custom always does so in its twofold guise. The account of morality is consistent with Hume’s methodology: it is never prescriptive, but instead is descriptive of our natural propensity to moral judgments; skeptical in the Pyrrhonian sense, and never nihilistic, in short, entirely continuous with the account of Book I, and equally founded in Book II.35 Nonetheless, just as Hume’s descriptive account of our epistemic principles grounds the possibility of epistemic prescriptions and evaluations, this account of ethics, although descriptive, grounds the possibility of moral prescriptions and evaluations. Hume distinguishes good from bad epistemic practice by describing the customs that constitute epistemic norms; he distinguishes virtue from vice, and specifies our duties by describing the customs (in both senses of that term) that constitute moral norms. As Garrett puts it, At no point in the course of the investigation does Hume invoke any substantive a priori normative principles. While this is certainly unusual, it should not be surprising. On his view, there can be no substantive a priori normative moral principles, for all awareness of moral truth depends in one way or another on the deliverances of the moral sense in experience. Equally, on his view, there can be no substantive a priori normative epistemic principles. (2007, 6)

6.  VIRTUE, STEADINESS, AND IMPARTIALITY I noted above that, according to Hume, our moral judgments flow from calm passions, and that what makes these passions calm, as opposed to violent, is that they are relatively stable in their operation. That is, they are fairly fixed elements of our character and our taste with regard to moral matters. They are also more or less disinterested, or nonegocentric, being caused by actions, traits of character, institutions, or states of affairs that are not immediately, or at least primarily, related to or evaluated in terms intelligibility presuppose convention entails only that the force of promises is not original, that is, that our regarding them as binding is not innate. 35. Baier (2010b) points out that Hume extends his account of justice considerably in the Enquiry Concerning the Principles of Morals, but that is beyond the scope of the present study. She also emphasizes the ways in which Hume’s historical research in the History of England informs his ethics, political philosophy, and moral psychology. I agree, and I recommend the History to anyone committed to a full understanding of Hume’s moral philosophy (or to anyone who loves a good read). Also see Baier (2008a) for a fine discussion of the impact of Hume’s history on his moral psychology. See Whelan (1985, 20–​21) for a similar analysis of the relationship between Hume’s skepticism, his naturalism, his appeal to custom, and his moral theory.

( 252 )  Books I and III

of our particular self-​interest. The disinterested nature of morality is hence traced, in Hume’s system, to the very nature of calm passions. In part III of Book III, and in particular in the first section, Hume expands on this idea, and it is useful to conclude this discussion of Hume’s moral theory with some attention to this discussion. Hume opens this section with a remark addressing the issue of morality and standing character: 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. [3.3.1.4, 575]

This is an interesting paragraph, and it is easy to misread it. One might think that Hume is arguing that there is nothing wrong with occasional wrongdoing, or nothing praiseworthy about the occasional good deed; that the only actions with any moral significance are those that are regular and habitual. Were that the case, and were that to follow from Hume’s moral theory, this would constitute an immediate reductio on the system, for it is clear that we do not customarily assess actions in that manner. But that would be at least an uncharitable reading. Hume is talking about the specific assessment of an action as virtuous or vicious, not good or bad in general, let alone just or unjust, honest or dishonest. He has argued that our primary object of moral assessment is not the individual act, but character, or institutions. Our assessment of actions, he argues, is parasitic on those assessments, not the other way around. This is something Hume inherits from the Hellenistic approach to morality that he admires so much. So all he is saying in the opening sentences is that we evaluate an action based upon what it tells us about the character of the actor; our evaluation of the character is the fundamental moral judgment. The Principle of Uniformity of Method is important in reading this passage. Judgments about causality reflect judgments about patterns of events, not about a single pair of events, and that is true even of a judgment that a particular pair of events is an instance of such a general pattern. Just so, Hume emphasizes here:  judgments about virtue and vice are judgments about patterns of behavior or of attitudes, and that is true even of a judgment that a particular action is virtuous or vicious.36 36. See Russell (1995, 199) for an extended discussion of Hume’s views about the relationship between action, character, and judgments of moral responsibility as well as a thorough review of the relevant secondary literature.

E t hi c s  

( 253 )

Hume then emphasizes [3.3.1.7–​9, 577–​78] the strong analogy between ethical and aesthetic judgment. We judge a work of art beautiful or ugly, or a trait of character admirable or despicable according to whether it is pleasing to us or not. But, he also emphasizes, the causes of that pleasure or displeasure are not the same in the aesthetic and ethical domain (although they may be coincident in some cases, as Hume points out—​aesthetic judgments of useful things like tools, animals, or fields may track their utility for us). Moral qualities, he emphasizes, are pleasing to us because of their social utility. He writes: No virtue is more esteem’d than justice, and no vice more detested than injustice; nor are there any qualities, which go farther to the fixing the character, either as amiable or odious. Now justice is a moral virtue, merely because it has that tendency to the good of mankind; and, indeed, is nothing but an artificial invention to that purpose. The same may be said of allegiance, of the laws of nations, of modesty, and of good manners. All of these are mere human contrivances for the interest of society. . . . Now as the means to an end can only be agreeable, where the end is agreeable; and as the good of society, where our own interest is not concern’d, or that of our friends, pleases only by sympathy: It follows, that sympathy is the source of the esteem, which we pay to all the artificial virtues. [3.3.1.9, 577]

At this point, Hume makes the crucial transition from the individual to the general point of view. Note first that Hume states clearly that moral judgment, at least in the case of the artificial virtues, tracks judgment about social utility, not judgment about whether the trait or action is beneficial to an individual. This marks a difference between the natural and artificial virtues. I feel and approve of benevolence toward my children because it pleases me to see my children prosper, but that very attitude, although the prospect of my children’s prosperity may be its cause, is egocentric—​it is directed to my children because they are mine. The fact that benevolence falls off so quickly with social distance means that this natural—​or more properly, original—​benevolence does not obviously benefit society; in fact, left alone, it could lead to factionalism and social disorder.37 37. Cohon (2008, 20) points out that this is one of the important motivations for the introduction of the impartial point of view as essential to moral reflection that we will discuss subsequently. The distinction between egocentric concern for others’ welfare because of their relationship to me, or because of how their prosperity affects my happiness (even if only in contemplation) and a nonegocentric concern for the welfare of others is also the basis of the Buddhist valorization of upekṣa, or nonegocentricity, and karuṇā, or nonegocentric care for others, as distinguished from attached desire for others’ welfare, which is often regarded in that moral tradition as a vice masquerading as a virtue.

( 254 )  Books I and III

The artificial virtues are different. They are not original qualities, but are “mere human contrivances.” In Mackie’s language, we “invent right and wrong” by establishing the customs that constitute the artificial virtues. But inventions are not ipso facto unreal; they are real inventions. And just as we come to enjoy and to encourage participation in other social inventions, such as musical concerts and games, we can come to enjoy and to encourage participation in social moral conventions. And we do. And we inculcate taste that ensures enjoyment of virtue and detestation of vice from an early age. By doing so, we create the ideological infrastructure that makes society possible, and by constituting our societies, we create the social infrastructure that makes morality possible. By constituting morality, we in turn make our characteristically human life possible, which in turn makes it possible for us to institute the conventions that constitute the artificial virtues. This may appear to be circular, but it is really a spiral, and surely a virtuous spiral, in more senses than one. So, although they are artificial, moral facts are hence just as real as any other social facts, or as real as the social institutions they enable. Note, as we observed earlier, that Hume does not take the artificial virtues to be in tension with natural virtues, as Cohon (2008) would have it; and here we can see why. The clue to understanding the consistency of these sets of virtues lies in Hume’s insight that sympathy is the foundation of the social virtues. Let us take as an example the use of sympathy to extend benevolence into justice. Benevolence, understood as an original quality, is necessarily restricted. By using the moral imagination and our natural sympathy for those we imagine to be like us in relevant ways, we develop a concern for the welfare not only for those immediately related to us, but for all of those with whom we might have transactions.38 This broader concern is socially beneficial (and for that reason, beneficial to me and to my immediate fellows as well). So we all come to esteem it, to encourage it, and to prosper from it. We come to see it as good. This new, broader, concern does not undermine or conflict with the initial benevolence that was its seed, but rather constitutes a tuning, or extension of that benevolence, mediated by the natural endowments of sympathy and the 38. It is also interesting in this context to compare Hume’s account of the role of the moral imagination to Buddhist meditational techniques for increasing the scope of care, such as the visualization of all sentient beings as once one’s mother, or the practice of exchange of self and others, practices which also use the moral imagination to broaden the reach of sentiments which might originate as egocentric, but which can be extended to a kind of universality that undermines that initial egocentricity while retaining the original affective valence.

E t hi c s  

( 255 )

moral imagination. The artificial is hence not in any conflict with the original, and both are natural. Hume is concerned to argue that this artifice is neither arbitrary nor cynical, as, for instance, Mandeville, to whom Hume owes much, suggests. Mandeville (1989) argues that the respect for the artificial virtues is the consequence of politicians cynically deploying moral propaganda to create a suitably compliant population. Hume notes and refutes this view: That many of the natural virtues have this tendency to the good of society, no one can doubt of. Meekness, beneficence, charity, generosity, clemency, moderation, equity, bear the greatest figure among the moral qualities, and are commonly denominated the social virtues, to mark their tendency to the good of society. This goes so far, that some philosophers have represented all moral distinctions as the effect of artifice and education, when skillful politicians endeavour’d to restrain the turbulent passions of men, and make them operate to the public good, by the notions of honour and shame. This system, however, is not consistent with experience. For, first, there are other virtues and vices beside those which have this tendency to the public advantage and loss. Secondly, had not men a natural sentiment of approbation and blame, it cou’d never be excited by politicians; nor wou’d the words laudable and praise-​worthy, blameable and odious, be any more intelligible, than if they were a language perfectly unknown to us. [3.3.1.11, 578–​79]

Mandeville’s account would undermine the Humean view that both original and artificial virtues are properly natural. Mandeville must be wrong, Hume thinks, first, because some of the original virtues do not conduce directly to the social good and are not so encouraged. One might count such virtues as partiality to one’s friends or clannish hospitality under that head. Second, he argues, there is a bootstrapping problem:  Mandeville’s account cannot get off the ground, as it presupposes an antecedent battery of moral attitudes, which therefore presuppose an antecedent battery of original objects to which these attitudes are addressed. Without these, there are no attitudes to extend so as to constitute those virtues or vices the politicians want to encourage or discourage. Virtue and vice must have an original basis. Hume concludes: The only difference betwixt the natural virtues and justice lies in this, that the good which results from the former, arises from every single act, and is the object of some natural passion:  Whereas a single act of justice, consider’d in itself, may often be contrary to the public good; and ’tis only the concurrence

( 256 )  Books I and III of mankind, in a general scheme or system of action, which is advantageous. [3.3.1.12, 579]

This passage introduces a new and important dimension to the natural/​ artificial distinction Hume has been adumbrating. The objects of the passions that approve the natural virtues are individual acts (taken always as signs of character, and so always in the context of patterns of behavior), whereas the objects of the passions that approve the artificial virtues are social structures or strictures.39 Particular instances, such as Hume’s example of the judge who decides in favor of a rich man and against a poor man, may conduce to social ill, but, as I emphasized above, the stable social patterns of behavior and expectation that are induced conduce to overall social benefit. And we each benefit from those structures and patterns. This is the recursive pattern of custom. We inculcate a custom of approving of and conforming to customary behavior. The first-​order custom constitutes the artificial pattern of morality; the higher-​order custom reinforces its normative and action-​guiding force. And custom in the sense of original instinct includes the passions that make the entire structure psychologically possible. As Hume puts it, “The approbation of moral qualities most certainly is not deriv’d from reason . . . ; but proceeds entirely from a moral taste, and from certain sentiments of pleasure and disgust, which arise upon the contemplation and view of particular qualities or characters” [3.3.1.15, 581]. This is Hume’s naturalistic account of how and why we invent right and wrong. Hume adds one more important dimension to the account, to complete the account of the normative force and depersonalization of the calm passions that guide morality: Every particular man has a peculiar position with regard to others; and ’tis impossible we cou’d ever converse together on any reasonable terms, were each of us to consider characters and persons, only as they appear from his peculiar point of view. In order, therefore, to prevent those continual contradictions, and arrive at a more stable judgment of things, we fix on some steady and general points of view; and always, in our thoughts, place ourselves in them, whatever may be our present situation. In like manner, external beauty is determin’d

39. Note once again that Hume emphasizes that the moral sentiments are passions, contra Loeb (1977).

E t hi c s  

( 257 )

merely by pleasure; and ’tis evident, a beautiful countenance cannot give so much pleasure, when seen at the distance of twenty paces, as when it is brought nearer us. We say not, however, that it appears to us less beautiful: Because we know what effect it will have in such a position, and by that reflexion we correct its momentary appearance. [3.3.1.15, 581–​82]

Here is where Hume appears to endorse what has been called an “ideal observer” theory of morality, according to which what is right is what somebody occupying that “steady and general point of view” would judge to be most pleasing. I hope by now that it is clear that this cannot be a correct understanding of his program. Hume is not in the business of specifying what makes an action or a character right, and in particular of specifying what grounds a judgment that a particular character or institution is morally pleasing or not. He is not a metaethicist, but a moral psychologist. He is therefore in the business of explaining the psychology and the complex social mechanisms that underlie our moral judgments and that constitute the framework in which they make sense. At this point in the account, he is confronting the decision problem that must arise in a society in which different persons come to different moral judgments about the same situation, in virtue of the contamination of their judgments by their own self-​interest. This contamination arises from a frequent kind of self-​deception: when addressing a hard moral problem, we are often required to ask whether our intuitions and motivations reflect a properly moral point of view or our own sense of our own personal interest. It is all too easy to convince oneself that motives based on self-​interest are in fact moral. This is a problem any reflective person has encountered, and the resolution of moral disputes requires that in dialogue we be able to surmount this kind of self-​deception. Hume sees this problem as analogous to problems we might encounter in aesthetic judgment. If you and I disagree about the beauty or artistic merit of a painting, how do we adjudicate the disagreement? Do we just say, de gustibus non est disputandum? That is more a confession of failure of aesthetic discourse than a resolution, and in any case, is not really open to us in the moral case. Or do we try to abstract from our respective preferences and personal taste and ask about the aesthetic qualities of the piece of art from a more detached standpoint? This may be harder, but this is the task of good criticism. Mutatis mutandis in the ethical case, argues Hume. Our own private perspectives might well color our ethical judgment. So if we care about getting it right—​about judging an act, an institution, a character according to the customs on which we all agree—​we must attempt to abandon our

( 258 )  Books I and III

particular perspectives. When we do so, we ask not what we prefer, but what a disinterested party who nonetheless participates in our moral customs would say. That may be difficult, but it is the task of good ethical discourse. This does not require that we adopt some idealized position, just that we abstract, in good faith, from our own particularity, to ask what our own customs actually dictate. More importantly, it is not to search for an arché beyond custom itself to find an ideal observer beholden to no moral conventions whatsoever. That would be to forget the role of custom in the constitution of ethical norms in the first place.40 That would be analogous to trying to solve the problem of the criterion in epistemology. And Hume, as a good skeptic, could never endorse that, any more than he, as a good naturalist, could ever endorse moral realism. Instead, reflecting his sense of the recursivity of custom, and in complete methodological harmony with his treatment of skepticism with regard to reason, he insists that one of our moral customs is the custom of trying to achieve as much disinterest as possible in the quest to bring our own judgments and behavior in harmony with the standards that we ourselves reflectively endorse.41 Who could ask more of moral reflection? In reading Hume’s treatment of ethics, we see that all of the principles of interpretation I introduced in ­chapter 1 pay off: we see Hume as a skeptic, as a naturalist, as a nominalist, using exactly the same pattern of argument he does elsewhere in the Treatise, and, most importantly, grounding his account in custom—​understood both in terms of habit grounded in the passions, and in terms of social convention. It is this ground that explains 40. And as Cohon (2008, 20) points, out, assuming this impartial point of view does not involve an escape from the passions, but rather ensures that the calm—​hence the moral—​and not the violent passions guide our action. See Rorty (1993) for a similar view. 41.  Again, see Cohon (2008) for a similar account. Hardin points out that Rawls’s complaint (2000) that Hume never provides an answer to the question “Why be moral?” and so never provides a genuine moral theory, is misguided. I agree. Hume demonstrates that we have moral norms that provide motivations for morality and grounds for moral approbation and disapprobation, and that the complex web of sentiments and customs that establish these norms also constitutes and enforces obligations. He also shows that these norms and institutions are socially advantageous, and that it is advantageous to each of us to participate in the social networks these norms enable and which in turn enforce these norms. This provides a compelling answer to the question “Why be moral?” on two levels: it shows why we should collectively institute moral practices and criticism, and why, on the individual level, we are each bound by the rules so instituted, and why it is prudent for us to recognize those bonds. To criticize Hume for failing to provide a transcendental ground, as opposed to this naturalistic ground, is to beg the question not locally, but against the entire philosophical enterprise of the Treatise.

E t hi c s  

( 259 )

the normative force of ethical discourse, just as it explains the normative force of the law. The account of the passions in Book II and the background of custom are the keys to understanding Hume’s treatment of ethics, just as they are the keys to understanding his treatment of human nature in the epistemic domain.

PA R T  I V

vwv Living Carelessly

C H A P T E R  1 2

w

The Appendix Second Thoughts about Second Thoughts

R

eaders of the Treatise have long been perplexed by the section in the  appendix in which Hume seems to recant his position on personal identity. There are several reasons to wonder about what to make of this brief discussion. First, Hume himself, while clearly expressing some reservations about the arguments he advances in Book I  of the Treatise, neither rebuts nor withdraws any of these arguments; he only doubts their conclusion. Second, he never replaces the account with any other. These facts at least suggest his ambivalence. Moreover, were he to recant these arguments, given the complete consistency in method we have observed in the Treatise as a whole, it is not at all clear that he would be entitled to any of the principal arguments or conclusions of the Treatise, inasmuch as all proceed from the same foundations via analogous reasoning (although Hume himself does not appear to see the consequences of these doubts as this far-​reaching). It is therefore worth closing with an assessment of Hume’s own second thoughts. For if we can lay them to rest, we rescue not only Hume’s account of personal identity from his own worries, but also the project of the Treatise as a whole; if we cannot reply adequately to these doubts, not only is the account of personal identity in peril, but so is his entire philosophical program. I hope that one merit of my reading is that it allows Hume a reply to his second thoughts, even if it is a reply he himself did not consider. The literature on the appendix is vast, and it is not possible to address all of it here. But it is worth noting a few alternative readings that

( 263 )

( 264 )  Living Carelessly

contrast with my own. Kemp Smith (1941) suggests that, in the appendix, Hume confronts a fundamental tension between his Newtonian and Hutchesonian impulses: as a Newtonian, he is committed to the claim that the mind is a complex of interacting perceptions and that it is apprehended as such; but as a Hutchesonian, he is committed to an observer—​a singular mind—​that apprehends that complexity and that confuses it with a unity. I see little textual evidence to support this reading. Fogelin (1985) puts this problem in a slightly different, and illuminating way:  on Hume’s account, perceptions are metaphysically independent mental events, and while minds are bundles of perceptions, those minds are ontologically posterior to, not anterior to, those perceptions. Indeed, in the context of the discussion of skepticism with regard to the senses, Hume countenances the possibility that any perception could exist independently of the bundle of which it is in fact a part. But that would seem to amount to the possibility that a perception could exist unperceived. That, as Fogelin perceptively puts it, means that Hume has no reply to Berkeley’s argument for the reality of mental substance. So one way to put the problem is in terms of the need for a synthesis that makes my perceptions mine; another is in terms of the need for a perceiver of any perception. I believe that my reading will preserve these insights, but with a bit more economy. Swain (1991) doesn’t think that there is much of a problem to solve in the appendix. She argues instead that all that Hume is doing in the appendix is reminding us of the conclusions he has drawn earlier in the Treatise regarding the impossibility of “finding the ultimate principles of the soul” (118). Although I agree with her that Hume is not identifying an insoluble problem for his program, I find it hard to square this reading with Hume’s own confession of despair in the appendix, and the need he felt to write it. Garrett (1997) offers a very different reading, arguing that the problem arises from Hume’s denial of spatial location to perceptions, with the consequence that there would be no way to establish the difference between qualitatively identical perceptions arising simultaneously in different minds. Without a difference in location, there is no way to distinguish such perceptions. There are, I  think, at least two things wrong with this reading. First, as Garrett acknowledges, there is no textual evidence that this is what bothered Hume; he never mentions or argues for this difficulty. But second, Hume does have the resources to distinguish qualitatively identical but non-​spatially located perceptions. He can note that they have distinct causes and effects, and hence belong in distinct causal continua.

The Appe n di x  

( 265 )

Garrett concludes his discussion of this problem with an astute observation: “It seems that the representational resources of [Hume’s] cognitive psychology are ultimately inadequate to the self-​referential task of explaining the representation of one’s own mind as an entity potentially distinct from others” (1997, 186). Once again, I  don’t think that the problem has to do with the distinctness of one mind from others, but rather with the impossibility of representing one’s own mind. That problem is soluble, I will argue, but Hume’s psychology does not make it easy for him to see the way to that solution. Garrett (2006), in a perceptive discussion of the implications of the appendix for Hume’s views on reason, points out that Hume notices that reason itself undermines the epistemic warrant of all of the faculties of the mind, including itself. This is right, but it is not the problem that animates these second thoughts. Garrett himself seems to grant this point when he notes (2006, 20)  that Hume concludes that his skeptical principles allow him to proceed with the investigation of human nature despite this subversion. Pears presents yet another diagnosis of Hume’s problem. He also takes Hume to be worried about the difficulty of individuating minds. He argues (1990, 1)  that without associating minds with the bodies that support them, there is no principle to which Hume can appeal in order to distinguish minds one from another. So, he argues, when we join Hume’s argument against the numerical identity of the person (individuated purely mentally) over time with the conclusion that without any such numerical identity we cannot distinguish persons one from another, we have a reductio on the idea that persons are distinct from their bodies. Beyond the fact that there is no real textual support for this reading, either, I will argue subsequently that Hume does in fact have the resources to draw the distinction he needs to draw. Nonetheless, there is something important in Pears’s idea: Hume’s accounts of our epistemological and ethical practice require that we recognize one another as distinct persons, and that we interact with one another as individuals. And the only basis for that recognition and interaction can be our respective bodies. The fact that Hume does not discuss physical identity in the context of these worries about personal identities is hence a problem, even if it is not the problem that Hume confronts in the appendix. McIntyre (1979, 1993) offers an account of Hume’s problem very similar to that which I present below, but with this difference: she argues that the problem concerns only the fact that the imagination is taken by Hume to yield contradictory conclusions: both a belief in a continuing self and a belief that there can be no such thing. She then argues for a very different

( 266 )  Living Carelessly

solution, suggesting that Hume can resolve this problem by appeal to the role of the passions and to the intimate relationship between the mind and body. I doubt that this is the right way to see the problem. For one thing, Hume is pretty clear about the role of the passions in generating the idea of the person; and for another, as McIntyre herself notes, there is no greater unity to be found in bodies than in minds, on Hume’s view. See Beauchamp (1979) for a more extensive reply to McIntyre. I now turn to my own understanding of Hume’s worries, and a prescription for alleviating them in a very Humean way. Hume begins the discussion in the appendix by noting that a philosophical system should not force upon us “contradictions and absurdities,” and so he implicates that he finds such contradictions and absurdities in his discussion of personal identity: I had entertain’d some hopes, that however deficient our theory of the intellectual world might be, it wou’d be free from those contradictions, and absurdities, which seem to attend every explication, that human reason can give of the material world. But upon a more strict review of the section concerning personal identity, I find myself involv’d in such a labyrinth, that, I must confess, I neither know how to correct my former opinions, nor how to render them consistent. If this be not a good general reason for scepticism, ’tis a least a sufficient one (if I were not already abundantly supplied) for me to entertain a diffidence and modesty in all my decisions. I shall propose the arguments on both sides, beginning with those that induc’d me to deny the strict and proper identity and simplicity of a self or thinking being. [Appendix 18, 633]

He immediately follows this introductory paragraph with a clear and fair, if brief, rehearsal of the central arguments for his account of the origin of the conventions regarding selves and his account of the nature of the person [Appendix 19–​27, 633–​35]. He expresses satisfaction with those arguments, affirming that he is “attended with sufficient evidence” for the position he defends. But then he immediately raises serious doubts about this very position: So far I seem to be attended with sufficient evidence. But having thus loosen’d all our particular perceptions, when I proceed to explain the principle of connexion, which binds them together, and makes us attribute to them a real simplicity and identity; I am sensible, that my account is very defective, and that nothing but the seeming evidence of the precedent reasonings cou’d have induc’d me to receive it. If perceptions are distinct existences, they form a whole only by being connected together. But no connexions among distinct existences are ever discoverable by human understanding. We only feel a connexion or a determination

The Appe n di x  

( 267 )

of the thought, to pass from one object to another. It follows, therefore, that the thought alone finds personal identity, when reflecting on the train of past perceptions, that compose a mind, the ideas of them are felt to be connected together, and naturally introduce each other. However extraordinary this conclusion may seem, it need not surprise us. Most philosophers seem inclin’d to think, that personal identity arises from consciousness; and consciousness is nothing but a reflected thought or perception. The present philosophy, therefore, has so far a promising aspect. But all my hopes vanish, when I  come to explain the principles, that unite our successive perceptions in our thought or consciousness. I cannot discover any theory, which gives me satisfaction on this head. [Appendix 28, 635–​36]

Let us examine this somewhat extraordinary reflection with care to see where and why the doubts intrude, and to what degree they are justified. We can represent the principal argument as follows: 1. If perceptions are distinct existences, they form a whole only by being connected together. 2. No connections among distinct existences are ever discoverable by the human understanding. 3. The only impression of unity comes from a feeling of a determination of thought to pass from one object to another. 4. Therefore, thought constitutes personal identity by passing from idea to idea and feeling them to be connected to one another. So far, so good. Hume has given us no reason to worry about the account, and he himself seems to be satisfied with it to this point. The problem appears to arise when he tries to “explain the principles, that unite our successive perceptions in our thought or consciousness.” So the difficulty seems to be that Hume has no theory or explanation of just how and why the mind connects particular perceptions together to constitute a self: why these perceptions and not those? One can feel the pull of this problem. If there is no principle at all to determine which perceptions are united to form a person, our identity would seem to be random. But if there were a principle, then the unity we experience would be independent of our construction, and would constitute the self of which Hume is sure we can have no impression.1 Moreover, if we are to take seriously much of Hume’s psychology—​committed as it is 1. Baxter puts it nicely: “The principle of connexion that he is worried he needs is identity itself” (2008, 200).

( 268 )  Living Carelessly

to mechanisms of learning and causation among psychological states—​we might think that we need a continuing self to serve as the substratum for the causal links between earlier events and the later events they cause, between impressions and the ideas that arise from them, as well as to explain why my impressions cause my ideas, and yours cause yours, and to explain why we never have cross-​individual causation of this type.2 This is the problem that led Kant to posit the transcendental unity of apperception. He was to reason that if we don’t have a unity of consciousness to start with, there is no subject position from which to unify a disparate set of representations, and so we could never achieve even the appearance of such a unity. So, Kant argues, we either start with a primordial unity of consciousness—​a transcendental subject—​or we never achieve unity of any kind. Now, it is worth noting that Kant must help himself to the distinction between a transcendental and an empirical subject here, and indeed he credits Hume with calling his attention to the need for a transcendental unity of apperception in order to solve the problem of the unity of the ego. That is, Kant agrees with Hume that the empirical self has only a constructed or represented unity, but thinks that the problem that Hume confronts here forces us to posit a higher unity as condition of constructing that very empirical integrity. But Hume, unlike Kant, cannot avail himself of a transcendent self, or even a transcendental unity of apperception as a condition for self-​consciousness; for, again in agreement with Kant, he does not believe that we could ever have an impression of any such thing; so, Hume concludes, we could have no idea of it. That is, Hume agrees with Kant that awareness of a person demands a synthesis; but he has no room for an agent of that synthesis, and the synthesis is hence appears to be inexplicable. That is Hume’s problem. Hume, without the benefit of Kant’s edifice as a framework for articulating this problem, sums up the predicament this way: In short, there are two principles, which I cannot render consistent; nor is it in my power to renounce either of them, viz. that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences. Did our perceptions either inhere in something simple and individual, or did the mind perceive some real connexion among them, there wou’d be no difficulty in the case. For my part, I  must plead the privilege of a sceptic, and confess, that this difficulty is too hard for my understanding. [Appendix 29, 636]

2. Passmore (1952) sees this as the central issue.

The Appe n di x  

( 269 )

Here are the two principles that together cause the problem: 1. All of our distinct perceptions are distinct existences; that is, they are independent of one another, and nothing intrinsic to any of them determines to what others they may be related or how. 2. The mind never perceives any real connection among distinct existences. That is, the only connections between them are those the mind constructs in virtue of its tendency to pass from one to another. But we need a third fact to get the inconsistent triad that bothers Hume: 3. But the mind does unite some perceptions in a regular way and not others, and there is no mind to do the uniting or to employ any principle of selection until the uniting is accomplished. If the question is seen this way, Hume has two problems, but only one is serious. The first is the one that seems to bother him at the outset: that is, what are the principles that determine the unity of the self? It seems that if we are appealing to some mechanisms that create the unity we experience, we should be able to say what they are. The second, which we have been bringing into focus in the previous discussion, is what we might call a “bootstrapping” problem: how do we get from a set of perceptions that has no intrinsic unity to the awareness of that set as a unity? The first problem either shouldn’t bother Hume at all or it should bother him a great deal more than it does. If it is a real problem, it is a problem for the entire enterprise of the Treatise. For not only in the case of personal identity does Hume end explanation at a bedrock of innate principles of the mind whose origin and fundamental nature is unknowable, but this is also the case in the discussion of causality, of skepticism with regard to the senses, and of morality. In general, Hume appeals to basic laws of psychology that govern the operations of the passions, and leaves those laws themselves unexplained. If this is a problem in the account of personal identity, it is therefore not specific to that topic, but is a fundamental problem with the methodology of the Treatise. On the other hand, I don’t think it is a problem, either in this case or in general; and this falls out from our reading of the text. Hume’s program, I have argued, is, in a broad sense, fundamentally Newtonian; it is a natural science of individual and social psychology. And just as Newton is free to discover laws of nature, but to leave those fundamental laws unexplained, Hume must be allowed the same liberty. Not forever: physics progresses, and so does cognitive science. But Hume, like

( 270 )  Living Carelessly

Newton, is inventing a science, in this case cognitive science, and just as Newton leaves problems for Einstein to solve, Hume, like any consistent naturalistic skeptic, can leave problems for a later generation of cognitive scientists to solve. So, even though he puts it that way, I  don’t think that the source of Hume’s problem is the inexplicability of the fundamental principles by means of which the mind synthesizes itself; as I just noted, Hume can postpone that one. Let us then examine the second problem in detail, for the hard problem is the bootstrapping problem: how does a mind with no identity of its own come to recognize past ideas as its own, to bundle them in order to synthesize a mind to enable a synthesis of a mind? This one is hard, and this one is specific to the case of personal identity, for all of the other cases invoking fundamental principles presuppose the integrity of a mind operating according to original principles. This would appear to drive one to posit a transcendental subject. Is Kant’s solution to this problem Hume’s only way out? Need we go transcendental to solve the bootstrapping problem? It is easy to see why Hume might have despaired, and why Kant may have taken the route he took. But in hindsight, I think we can see a solution for Hume more congenial to his naturalistic temperament and project. That solution is not to rise to the transcendental level but to sink to the physical level, and to locate the original principles that generate the effective illusion of the unity of mind from a plurality of disunified cognitive processes in subdoxastic cognitive and brain processes. And the fact that Hume could have pursued this strategy provides a kind of vindication for his skeptical naturalism.3 Here is what Hume could have said (though this might require a considerable rethinking of the relation between mind and body in 1.4.5): We are not born with minds; we are born with neural systems that operate in lawlike ways to generate minds. They associate and make available for association complex states that supervene at least in part in our brains. The processes by means of which they do that are cognitively opaque, just as are so many of the original processes that govern our passions as adults. But the pathways they create and the networks of association they construct, as they become more complex, become causally efficacious. These initial pathways and networks create more complex pathways and networks (just

3. See Waxman (1992) for a more Kantian, transcendental recommendation for the solution of Hume’s problem.

The Appe n di x  

( 271 )

as Malebranche suggested—​a suggestion that we have seen Hume take up, and just as initial customs can include customs for crafting better customs). The autopoietic processes that constitute our minds thus begin as simple, primitive reflexes, but their reflexivity and plasticity, in interaction with the physical and social environment and with our trainers, gradually enable them to become minds, systems that can experience impressions, form ideas, and come to use language and to discriminate, to associate, and to unify sets of impressions as their own, and eventually to recognize themselves as the subjects of the narratives they thereby write of their own lives. This is the route to an empirical, naturalistic account—​as opposed to a Kantian transcendental account—​of synthesis, and one that does not presuppose the unity it seeks to explain.4 The science was not yet there for Hume, and so claiming “the privilege of the sceptic” was the intellectually honest course of action. But everything in the Treatise points the way to the science that was to emerge. That science is what we now recognize as embodied, embedded, and extended cognitive science, a psychology that is at the same time social and biological, and one that reveals our human subjectivity as an achievement, not a given; but an achievement that good science can explain, with our original

4. Here again, I diverge from Baxter. While we agree that Hume’s desire for a naturalistic account of mind that would solve this problem is not satisfied by the end of the Treatise and that this causes his despair, Baxter takes the solution to be blocked by Hume’s realization that consciousness itself is, while the only instrument available for the study of mind, itself an inaccurate instrument, subject to contradictory findings (2008). On my view, it is simply Hume’s inability to think through the process of the ontogenesis of mind. Baxter takes the problem Hume encounters in the appendix to be principled and epistemological; I take it to be psychological and merely the consequence of the youth of his science. Baxter, at the end of his book, defends a more nuanced view of Hume’s problem according to which it reflects a larger problem, viz., the intensionality of identity. Baxter describes this as the problem of even asking whether two things are identical or not. To pose the problem of identity, we have to take them as two; to take seriously the possibility that they are identical is to represent them as one. So, to ask whether I am identical now to the self who began writing this sentence is to represent myself now and myself then, both as the same and as different. Baxter points out that Hume does not have the resources to solve this puzzle regarding the nature of thought about identity. (Indeed, he claims that it is as yet unsolved.) Whether or not one agrees regarding how serious this puzzle is, however, I do not think that this is Hume’s puzzle, the one that worries him in the appendix. For one thing, I think that we have just provided an account that accords more closely with the words Hume uses in posing the problem. But for another, while I agree with Baxter that Hume does not have the resources to solve this problem, I do not think that Hume even has the resources to pose this problem. Hume nowhere in the Treatise seems to be bothered with problems about the intensionality of language and thought, and I see no reason to think that he suddenly discovers these problems in the appendix. This reading therefore just seems more tendentious than is necessary.

( 272 )  Living Carelessly

psychological constitution and the social matrix into which we are born as the starting points. If this is right, then while, to be sure, Kant’s critical philosophy is in many respects an advance over Hume’s empiricism—​particularly insofar as it recognizes the need for a theory of mental activity and of the interaction of top-​down and bottom-​up processes—​in some other respects Hume has the better story. For Kant—​his concern for the status of scientific knowledge not withstanding—​abandons Hume’s commitment to a thoroughgoing naturalism. In positing a transcendental ego, he offers an explanation of the unity of consciousness that requires us to leave the spatiotemporal and causal order. This is the move that Hume properly resists.5 If we are to explain human cognition, he sees, that explanation must itself be scientific and causal, adverting to natural processes. If this line of thought is correct, all that Hume missed in the appendix is the possibility that a set of relatively automatic, blind, subdoxastic processes, set in an appropriate biological, cultural, and linguistic milieu, could, in interaction with that social environment, organize so as to become a person capable of thinking of itself as an individual, capable of recognizing itself as an extended psychological continuum among others. It is not surprising that he missed that—​the science of cognitive development was not available to him. But since the time of Vyogotsky, this is the picture of human ontogenesis we have come to accept, and it is not only a solution to the bootstrapping problem, but it is a nontranscendental, empirical, and plausible solution to it. Hume needn’t have despaired: his own naturalism and his own commitment to a socially embedded account of human nature provide the key to the escape from the predicament that, in the appendix, he took his position in Book I to create.

5. And we could draw a similar contrast in the case of ethics: where Kant sees the need to go transcendental and to posit at least the idea of a free agent standing outside of the empirical order in order to account of obligation, Hume remains resolutely naturalistic, arguing that the moral obligations we recognize can only emerge from patterns of human conduct explicable by natural and social science. Once again, Hume seems to hold the high ground, and for the same reasons. But see Korsgaard (1986, 2009) for a very different assessment.

C H A P T E R  1 3

w

Persons as Customary Creatures

I

hope to have vindicated the interpretative principles that animate this study. In particular, I hope to have shown that Hume offers a naturalistic account of human nature, using a Pyrrhonian methodology, and that that methodology is completely consistent with his naturalism. That naturalism, I have argued, involves an account of human behavior as driven by the passions. For this reason, Book II must be taken as foundational to the project of the Treatise. And I hope to have shown that the Treatise is a methodologically unified study, one which joins that account of the passions as the foundation of our psychology to an account of custom inspired by reflection on British customary law and its account of the sources of normativity as the core of his vision of human life. But, we might now ask, what kinds of beings does Hume take us to be? And how plausible is that vision? One leading idea is that we are, first and foremost, organisms in a natural world, as opposed to rational souls or transcendental egos that stand outside the world as disembodied subjects to the world as our empirical object. We are instead ourselves natural objects firmly embedded in the natural world, and our behavior, including our thought, is as subject to natural law, and as determined, as that of any stone or insect. Our behavior may be more complex than that of an insect, just as that of an insect is more complex than that of a stone, requiring more explanatory principles. But that complexity reflects differences in degree in the number of variables and range of interactions necessary to consider, not a difference between inanimate objects, irrational, natural brutes, and rational persons partaking of the divine or the transcendent.

( 273 )

( 274 )  Living Carelessly

Hume certainly underestimated that complexity. He (like quite a few of his behaviorist descendants a couple of centuries down the line) was convinced that a relatively few fairly simple associationist principles would suffice for the explanation of human thought and behavior. And he was pretty sure that not much needed to be said about human cognitive architecture and the dynamics of cognition in the explanation of our behavior. Subsequent cognitive science has revealed those views about the simplicity of psychology to be significant errors. But as regulative principles, they aren’t bad. After all, it is a good idea in science to see how few and how simple one’s basic principles can be in order to explain the dynamics of a complex system. Even bees (and one has to suspect that Hume—​ following Mandeville—​ often takes those social insects as inspiration when he imagines the dynamics of human behavior) exhibit remarkably complex behavior that can be explained by reference to a few fairly simple empirical generalizations, with much of the apparent complexity due to the complexity of their environment, and, especially, their complex social structure—​a social structure itself explicable based on a few simple principles. It was not irrational for Hume to apply the same kind of ethological approach to Homo sapiens, at least provisionally. And we are, according to Hume, essentially social creatures. We are governed by custom; and custom, as I  have been at pains to emphasize, comprises both our original habits and instincts and those established by convention. Most importantly, we are originally tuned to the social, and this enables custom to have the normative force it does, governing our ethical as well as our cognitive lives. As we have discovered in the twentieth century, we are wired to learn language, even though we are not wired to speak any particular language; we are wired to imitate facial expressions; we are wired to respond to the cognitive states of our conspecifics; we are wired to learn and to participate in the customs of our social groups; we are wired both for limited natural sympathy and for the ability to extend that sympathy; we are wired to think and to reason; we are even wired for pervasive cognitive illusions, just as we are wired for sensory illusion. Moreover, those cognitive illusions, despite being illusions, underlie our ability to behave socially: they allow us to take ourselves to be free; to be guided by reason; to be discrete individuals; to treat our customs as binding, and so to take ourselves to be bound to one another through them. It is therefore impossible to understand persons correctly as individuals abstracted from our social environments and the customs they institute, just as it is impossible to understand us as thinkers abstracted from our physical embodiment. That is so whether one is trying to explain behavior or cognition; the epistemic realm or the moral. And this despite the pervasive

P e r s o n s a s C u s t o m a r y C r e at u r e s  

( 275 )

illusion that we are independent individuals responding to a world that exists just as it presents itself to us. One of Hume’s most important insights in the Treatise is that we enact even our individuality, not as solo performers at an open mic, or even as karaoke performers with a preselected backing tape, but rather as members of an improv ensemble in a club that constantly blurs the distinction between audience and performers. We constitute ourselves as persons through a whirl of passions that have social causes and that constantly involve representing others as individuals among whom we live; we come to speak languages, and so to be capable of abstract thought by participating in the social conventions that constitute linguistic meaning, a participation in turn enabled by our original psychobiological endowment, an endowment whose evolutionary explanation in turn requires that social context. We come to function as agents in a moral order by participating in the artificial virtues and the customary practices that make social intercourse possible, social intercourse that in turn makes our social environments, including both material and intellectual culture, possible. The norms to which we are responsive, and which constitute us as persons—​as those who can occupy roles governed by rules—​themselves emerge from this web of custom in ways that we can understand causally, an understanding that we have seen that Hume derives from his understanding of the force of customary law. While it may be tempting to think that we can understand the social world reductively, by first understanding the dynamics of individual behavior, and then constructing our theory of social dynamics on that foundation (a temptation to which Hume in one moment succumbs in his associationist ambitions), that approach is doomed:  we can only understand individuals who, like us, are governed by social as well as by biological custom—​who are naturally social—​by understanding the societies that shape us first, and then by coming to understand how we are shaped by them. This is something that Hume sees more clearly in Books II and III than he does in the early pages of Book I that draw so much scholarly attention.1 1. I have noted several times the ways in which Hume anticipates ideas (or perhaps actually influenced the ideas) of Edward Chace Tolman. Here one might mention Tolman’s argument in (1951). Tolman argues that although one might suppose that physiology is a more basic science than psychology and psychology a more basic science than sociology, in fact the hierarchy should be reversed: we can only understand human psychology by understanding the social matrices in which our minds and behavior unfold; we can only understand our biology by understanding the complex behavior and cognition it evolved to subserve and that constantly shapes it. And Tolman emphasizes throughout his work that behavioral regularities are mediated by imagelike

( 276 )  Living Carelessly

All of this in turn suggests that we should think of Hume not merely as the progenitor of behaviorism in twentieth-​century psychology and of connectionism in more recent psychology, but also of the social cognitive psychology initiated by such figures as Vygotsky (1978) that encourages us to understand human behavior from the social level down, and of the embodied, embedded, and extended approach to cognition that is now dominating cognitive science (Dennett 2010, 2017; Thompson 2010) (although once again, Hume’s atomistic tendencies when he thinks about perception sometimes obscure or even undermine these aspects of his thought). All of this connects immediately to Hume’s insightful deconstruction of the distinction between nature and artifice. Hume reveals us to be natural artificers, and the artifacts for which we are responsible include not only buildings, bridges, and boats, but also currencies, communities, courts, and customs. We create, as complex artifacts, both the physical and the social milieus that make human life possible. And our artifacts go beyond currencies, communities, and courts and include languages, laws, and legacies, and so the ways we understand and regulate ourselves, both synchronically and diachronically. Inasmuch as the narratives that constitute us as persons intelligible to ourselves and to each other are so constructed, we can say without exaggeration that we ourselves are the most important artifacts for which we are responsible as craftspersons. Everything we are, and every frame through which we can understand ourselves and our fellows, is itself a human construction. While humanity may not be the crown of any divine creation, we bid fair to be the crowns of our own creative activity. But we and so much of the world we inhabit (whether the notion of inhabiting a world is understood physically, socially, or phenomenologically) are, Hume argues, fictional. The self, external objects as we understand them, the causal nexus that underlies our explanatory behavior, and even our self-​individuation, moral facts, are all, as we have seen, characterized by Hume as fictions. To be fictional, however, from a Humean perspective, is not to be unreal. Instead, it is the way that the human lifeworld is real.2 Ishmael and Ahab inhabit a fictional world, and their reality is a world representations—​what he called cognitive maps—​that guide behavior in lawlike ways, but that must be understood as representational, with their representational character deriving from their causal linkages to perception and to action. 2. And this is yet another site of affinity between the Humean and the Madhyamaka Buddhist ways of seeing things, and in particular between the thought of Hume and Candrakīrti. For discussion of conventional reality not as a second-​rate reality, but as the way that real things exist, see Nāgārjuna (1995) and Garfield (2015), and for

P e r s o n s a s C u s t o m a r y C r e at u r e s  

( 277 )

within that fiction. We do not inhabit that fictional world. But we inhabit one that is no less fictional than theirs, so long as we keep the right sense of fiction in mind; it is just one with respect to which theirs is an embedded fiction. The objects we see and hear are created by our sensory and cognitive faculties; there is no sense whatsoever—​literally—​in pretending that we experience objects as they are independent of those faculties.3 Hume emphasizes that our identity is also fictional. That is to say that the world we inhabit is a world that we create; but it is also a world that we reflexively take to be something not that we have made, but that we have discovered. This is a deep form of the Myth of the Given, as Sellars (1963, 196) was to call it about two centuries down empiricist history from the Treatise. That world—​the one we inhabit and take to be given—​Hume emphasizes, can only be fictional. This has further implications for how we think of truth from a Humean perspective. For we are often inclined to contrast truth and fiction, and to see fiction as necessarily a kind of falsehood. But if we take the moral of the story told in the Treatise seriously, we should abandon this contrast, just as we abandon that between the natural and the artificial. Fiction is not opposed to truth, but often constitutes it. As we noted earlier, Moby Dick the novel makes it true that Ahab captained the Pequod; the narrative does not reflect that truth; it is not responsible to it. Instead, it constitutes it. Just so, the fictions we create constitute truths, including truths about persons, causal relations, morality, epistemology, meaning, and the rest—​all of the truths that make science, society, and human life itself possible. These truths as much determine the character of our world as the truths constituted by Moby Dick determine the world of Ishmael and Ahab. And like those truths in that world, the fictional character of these truths in our world does not entail that they are arbitrary: we—​thrown as we are into this world, in Heideggerian terms—​can never take them as optional. They are truths, for to be a fictional truth is not to fail to be a truth, but rather is a way of being a truth, and indeed may be the only way that truth can be constituted.4 a discussion of how to see Madhyamaka as a kind of panfictionalism, see Cowherds (2010), Matilal (1970), and Tsongkhapa (2006). 3.  Kant is often credited with this insight, but it was a commonplace for Indian Buddhists and was a standard doctrine of the Pyrrhonian skeptics. Hume is responsible for bringing it to the early modern European world and to Kant’s attention. 4. See Cowherds (2010) for a parallel discussion of conventional truth as a kind of truth, this time in the context of Madhyamaka Buddhist philosophy. Candrakīrti, for instance, emphasizes that it is necessary within the domain of conventional truth to distinguish between conventional truth and conventional falsehood, and that this distinction really is a distinction between the true and the false. He argues that it can

( 278 )  Living Carelessly

Finally, this fictionalism, like Hume’s Pyrrhonism, is not only consistent with Hume’s naturalism, but is entailed by it. It is Hume’s resolutely naturalist investigation into the mechanisms of thought and human life that leads him to see that and how we create the world we inhabit. Like any good Pyrrhonian, he must accept the reciprocity of that relationship: the very evidence and mode of reasoning that leads us to that fictionalism is itself part of the fiction, and the natural world in which we embed ourselves in order to discover the inevitability of skepticism is as fictional as anything else, and known subject to the same skeptical strictures.5 Hume hence also asks us to abandon any sense that philosophy can presuppose a starting point for reflection, begin from or find any Archimedean point outside of human life. Philosophy, like everything else we do, begins and ends in the human world we create in part by doing philosophy. I hope that these reflections also help us better to situate Hume within the history of philosophy. Hume is heir not only to the naturalistic and sentimentalist ideas of Mandeville, Shaftesbury, and Hutcheson and to the empiricism of Berkeley and Locke. He is also heir to a subtle skeptical tradition he inherits, through Bayle, from Sextus Empiricus, Pyrrho, and eventually from Indian Buddhist philosophers of whose existence he was undoubtedly ignorant, but whose thought probably influenced Pyrrho himself and his Greek followers, and so, through a long line of transmission, eventually influenced Hume. He is a bridge linking that Hellenistic world to such twentieth-​century figures as Wittgenstein (as a skeptic and

only be drawn on epistemological grounds, and specifically with reference to human epistemic conventions, as there is no convention-​independent reality against which, say, correspondence could be measured. 5. Hume puts the point nicely in the Enquiry: And though a Pyrrhonian may throw himself or others into a momentary amazement and confusion by his profound reasonings; the first and most trivial event in life will put to flight all his doubts and scruples, and leave him the same, in every point of action and speculation, with the philosophers of every other sect, or with those who never concerned themselves with any philosophical researches. When he awakes from his dream, he will be the first to join in the laugh against himself, and to confess, that all his objections are mere amusement, and can have no other tendency than to show the whimsical condition of mankind, who must act and reason and believe; though they are not able, by their most diligent enquiry, to satisfy themselves regarding the foundation of these operations, or to remove the objections, which may be raised against them. (EHU 12.23; SBN 160) It is important to note that while Hume may deprecate Pyrrhonism, however gently, in the Enquiry, it is in fact the philosophical strategy he adopts in the Treatise, and that while he defends (EHU 12.24; SBN 161)  an alternative, “mitigated skepticism,” in the Enquiry, this same remark demonstrates that he never abandons his critique of foundationalism in epistemology.

P e r s o n s a s C u s t o m a r y C r e at u r e s  

( 279 )

a communitarian), Quine, and Sellars (as a naturalist). Hume’s genius consists in part—​as Kemp Smith realized—​in the synthesis of this naturalism and skepticism. But that genius is also evident in his insight that this naturalism and skepticism lead inevitably to a communitarian view of human life, of meaning, and even of the constitution of the world we inhabit. It is also evident in his understanding that custom—​including both social convention and our innate and invisible cognitive structures—​ determine our cognition, behavior, experience, and values to a far greater extent than does anything of which we are typically aware. Hume’s understanding of custom, we have seen, is grounded in an appreciation of its role in law. Custom is both biological and social; it is explained in part by our original, genetically determined constitution; but inasmuch as we have evolved as social creatures, our original nature can only unfold in the context of a collectively constructed social matrix into which we are born. That matrix, itself causally explained by our biological capacities, institutes a web of practices in which we are raised, and which we come to take for granted. These practices may be local, but they perforce interact with those of other local communities; they may be contingent, but they are sufficient to induce normativity; they may be ancient, but they contain within them the resources for their own adaptive evolution. Despite its pervasive influence, as Hume notes so perceptively in the Enquiry, custom conceals itself,6 just as the very air we breathe disappears from our awareness when we are not attending to it. As a consequence, we often take ourselves to be independent subjects inhabiting a natural world that just happens to be as we found it, subject to epistemic and moral norms whose sources lie outside of us. We are, as Hume saw so clearly, strangers to ourselves, capable of being revealed only by careful natural science. That Hume was unable to complete that science is not his fault; that he saw the need for it, as opposed to a priori speculation about the nature of reason, is part of his untimely genius. Hume is an important precursor to Kant, but not simply as an alarm clock to awaken him from his dogmatic slumbers, or as a whipping boy whose empiricist errors are to be corrected by transcendental idealism. He anticipates much of what is great in the first Critique, including the centrality of the imagination to any theory of human synthetic activity, and the constructed character of the empirical self. And, as I  argued in the

6. A final delicious irony in Hume’s relation to Buddhism is that the Sanskrit term often translated as custom—​samvṛti—​means both convention, as in social convention, or ordinary everyday practice, and concealment or obscuration. See Garfield (2015, chap. 3).

( 280 )  Living Carelessly

previous chapter, Kant’s retreat from Locke’s and Hume’s naturalism may not be such a good thing. While it may offer a more satisfying theoretical epistemology, it does so at the cost of removing the subject from the empirical world, a move that led inevitably to the mysticism of Wittgenstein’s Tractatus. Hume is more the precursor to the Wittgenstein of Philosophical Investigations, to what we are now accustomed to recognizing as the postmodern collectivist understanding of meaning, thought, and mind, and of persons as members of a community that simultaneously constitutes and inhabits a world of meaning, as beings opaque to themselves, who can understand themselves only in the mirror of the lives of others. We read Hume best when we read him in the context of the naturalistic, skeptical, and, above all, communitarian tradition that reveals our human nature to consist in our nature as fallible social animals, animals only dimly aware of their own cognitive powers and even of their own cognitive and affective states, and who, while they might seem to themselves to be independent individuals, are ineliminably constituted by their social, as well as their biological, histories and contexts.

R EF E R E NC E S

Abramson, Kate. 2015. “What’s So ‘Natural’ about Hume’s Natural Virtues?” In The Cambridge Companion to Hume’s Treatise, edited by Donald Ainsley and Annemarie Butler, 333–​68. Cambridge: Cambridge University Press. Ainslie, Donald. 1999. “Scepticism about Persons in Book II of Hume’s Treatise.” Journal of the History of Philosophy 37 (3): 469–​92. Ainslie, Donald. 2005. “Sympathy and the Unity of Hume’s Idea of Self.” In Persons and Passions: Essays in Honor of Annette Baier, edited by Joyce Jenkins, Jennifer Whiting, and Christopher Williams, 143–​73. Notre Dame, IN:  University of Notre Dame Press. Ainslie, Donald. 2015. Hume’s True Scepticism. Oxford: Oxford University Press. Alanen, Lilli. 2005. “Reflection and Ideas in Hume’s Account of the Passions.” In Persons and Passions: Essays in Honor of Annette Baier, edited by Joyce Jenkins, Jennifer Whiting, and Christopher Williams, 117–​42. Notre Dame, IN:  University of Notre Dame Press. Alanen, Lilli. 2006. “The Powers and Mechanisms of the Passions.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 179–​98. Oxford: Blackwell. Allison, Henry E. 2008. Custom and Reason in Hume: A Kantian Reading of the First Book of the Treatise. Oxford: Clarendon Press; New York: Oxford University Press. Ardal, Páll S. 1966. Passion and Value in Hume’s Treatise. Edinburgh:  Edinburgh University Press. Bacon, Francis. 1909. “Of Custom and Education.” In Essays, Civil and Moral: And the New Atlantis, vol. 3, edited by Charles W. Eliot, L.L.D., 103-​104. New  York, NY: P.F. Collier & Son. Baier, Annette. 1991. A Progress of Sentiments: Reflections on Hume’s Treatise. Cambridge, MA: Harvard University Press. Baier, Annette. 1994. “Hume:  The Reflective Women’s Epistemologist?” In Moral Prejudices: Essays on Ethics, 76–​94. Cambridge, MA: Harvard University Press. Baier, Annette. 1997. The Commons of the Mind. Edited by American Philosophical Association. Chicago: Open Court. Baier, Annette. 2010a. Reflections on How We Live. Oxford: Oxford University Press. Baier, Annette. 2010b. The Cautious Jealous Virtue:  Hume on Justice. Cambridge, MA: Harvard University Press. Baier, Annette. 2011. The Pursuits of Philosophy: An Introduction to the Life and Thought of David Hume. Cambridge, MA: Harvard University Press. Balfour, James. [1768] 2000. “Philosophical Essays.” In Early Responses to Hume’s Metaphysical and Epistemological Writings, edited by James Fieser, 3:185–​99. Bristol: Thoemmes Press.

( 281 )

( 282 )  References Baxter, Donald L.  M. 1993. “Hume’s Theory of Space and Time in Its Skeptical Context.” In The Cambridge Companion to Hume, edited by David Fate Norton and Jacqueline Taylor, 105–​146. Cambridge: Cambridge University Press. Baxter, Donald L. M. 2006. “Identity, Continued Existence, and the External World.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 114–​ 32. Oxford: Blackwell. Baxter, Donald L.  M. 2008. Hume’s Difficulty:  Time and Identity in the Treatise. New York: Routledge. Bayle, Pierre. [1696] 1965. Historical and Critical Dictionary (Selections). Edited by Richard H. Popkin. Indianapolis, IN: Bobbs-​Merrill. Beauchamp, Thomas. 1979. “Self Inconsistency or Mere Self Perplexity?” Hume Studies 5 (1): 36–​44. Beck, Lewis White. 1978. Essays on Kant and Hume. New Haven: Yale University Press. Beckwith, Christopher I. 2015. Greek Buddha: Pyrrho’s Encounter with Early Buddhism in Central Asia. Princeton, NJ: Princeton University Press. Beebee, Helen. 2006. Hume on Causation. New York: Routledge. Bell, Martin. 1993. “Hume on Causation.” In The Cambridge Companion to Hume, edited by David Fate Norton and Jacqueline Taylor, 147–​76. New York: Cambridge University Press. Bennett, Jonathan Francis. 2003. Learning from Six Philosophers:  Descartes, Spinoza, Leibniz, Locke, Berkeley, Hume. Oxford: Clarendon Press. Bentham, Jeremy. 1825. A Treatise on Judicial Evidence. Edited by M. Dumont. London: Paget. Biro, John I. 1993. “Hume’s New Science of the Mind.” In The Cambridge Companion to Hume, edited by David Fate Norton, 33–​63. Cambridge: Cambridge University Press. Blackburn, Simon. 2000. “Why We Pull Together: Hume’s Understanding of Habit and Sentiment in Human Life.” Times Literary Supplement 5056: 14–​15. Bricke, John. 2011. “Hume on Liberty and Necessity.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 201–​66. Malden, MA: Wiley-​Blackwell. Broughton, Janet. 2003. “Hume’s Naturalism about Cognitive Norms.” Philosophical Topics 31 (1): 1–​19. Broughton, Janet. 2004. “The Inquiry in Hume’s Treatise.” Philosophical Review 113 (4): 537–​56. Brown, Charlotte. 2011. “Hume on Moral Rationalism, Sentimentalism, and Sympathy.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 219–​39. Malden, MA: Wiley-​Blackwell. Burnyeat, Myles. 1983. “Can the Skeptic Live His Skepticism?” In The Skeptical Tradition, 117–​48. Berkeley: University of California Press. Capaldi, Nicholas. 1966. “Hume’s Rejection of ‘Ought’ as a Moral Category.” Journal of Philosophy 63 (5): 126–​37. Cohen, L. 1977. The Probable and the Provable. Oxford: Oxford University Press. Cohon, Rachel. 2006. “Hume’s Artificial and Natural Virtues.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 256–​75. Oxford: Blackwell. Cohon, Rachel. 2008. Hume’s Morality:  Feeling and Fabrication. New  York:  Oxford University Press. Cohon, Rachel. 2011. “Hume’s Indirect Passions.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 159–​84. Malden, MA: Wiley-​Blackwell. Connolly, John. 1987. “David Hume and the Concept of Volition:  The Will as Impression.” Hume Studies 13 (2): 276–​307.

References  ( 283 ) Coventry, Angela Michelle. 2006. Hume’s Theory of Causation:  A Quasi-​ Realist Interpretation. New York: Continuum. Cowherds. 2010. Moonshadows: Conventional Truth in Buddhist Philosophy. New York:  Oxford University Press. Cowherds. 2015. Moonpaths: Ethics and Emptiness. New York: Oxford University Press. Darwall, Stephen. 2002. “Ethical Intuitionism and the Motivation Problem.” In Ethical Intuitionism:  Re-​evaluations, edited by Phillip Stratton-​Lake, 248–​270. Oxford: Oxford University Press. Dauer, Francis Watanabe. 2011. “Hume on the Relation of Cause and Effect.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 89–​ 105. Malden, MA: Wiley-​Blackwell. Davies, John. 1615. Le primer report des cases and matters en ley resolues & adiudges en les courts del roy en Ireland. Dublin. Dennett, Daniel. 2010. Consciousness Explained. Boston: Back Bay Books. Dennett, Daniel. 2017. From Bacteria to Bach and Back:  The Evolution of Minds. New York: Norton. Desideri, Father Ippolito, S.J. 2010. Mission to Tibet:  The Extraordinary Eighteenth-​ Century Account of Father Ippolito Desideri S.J. Edited by Leonard Zwilling. Translated by Michael J. Sweet. Boston: Wisdom Publications. Duckworth, D., et al. 2016. Dignāga’s Investigation of the Percept: A Philosophical Legacy in India and Tibet. New York: Oxford University Press. Ekman, Paul, and Richard J. Davidson. 1994. The Nature of Emotion:  Fundamental Questions. New York: Oxford University Press. Elsky, Stephanie. 2013a. “Common Law and the Commonplace in Thomas More’s Utopia.” English Literary Renaissance 181–​210. Elsky, Stephanie. 2013b. “‘Wonne with Custome’:  Conquest and Etymology in the Spenser-​Harvey Correspondence and A View of the Present State of Ireland.” Spenser Studies 28:  168–​92. https://​www.questia.com/​library/​journal/​1P3-​ 3717200871/​wonne-​with-​custome-​conquest-​and-​etymology-​in-​the. Elsky, Stephanie. 2014. “Lady Anne Clifford’s Common-​Law Mind.” Studies in Philology 111 (3): 521–​46. Falkenstein, Lorne. 2006. “Space and Time.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 59–​76. Oxford: Blackwell. Fogelin, Robert J. 1985. Hume’s Skepticism in the “Treatise of Human Nature.” London: Routledge & Kegan Paul. Fogelin, Robert J. 2009. Hume’s Skeptical Crisis:  A Textual Study. New  York:  Oxford University Press. Frede, Michael. 1997. “The Sceptic’s Beliefs.” In The Original Sceptics:  A Controversy, 1–​24. Indianapolis, IN: Hackett. Gamboa, Steven. 2007. “Hume on Resemblance, Relevance, and Representation.” Hume Studies 33 (1): 21–​40. Garfield, Jay. 1988. Belief in Psychology: A Study in the Ontology of Mind. Cambridge, MA: MIT Press. Garfield, Jay. 1990. “Epoche and Śūnyatā: Skepticism East and West.” Philosophy East and West 40 (3): 285–​307. Garfield, Jay. 1994. “Dependent Arising and the Emptiness of Emptiness:  Why Did Nāgārjuna Start with Causation?” Philosophy East and West 44 (2): 219–​50. Garfield, Jay. 2002. Empty Words: Buddhist Philosophy and Cross-​Cultural Interpretation. New York: Oxford University Press.

( 284 )  References Garfield, Jay. 2010. “What Is It Like to Be a Bodhisattvaa? Moral Phenomenology in Śāntideva’s Bodhicaryāvatāra.” Journal of the International Association of Buddhist Studies 33 (1–​2): 327–​51. Garfield, Jay. 2011. Western Idealism and Its Critics. Sarnarth, India: Central University of Tibetan Studies Press. Garfield, Jay. 2012. “Mindfulness and Ethics: Attention, Virtue and Perfection.” Thai International Journal of Buddhist Studies 3: 1–​24. Garfield, Jay. 2014. “Just Another Word for Nothing Left to Lose: Freedom, Agency and Ethics for Mādhyamikas.” In Freedom of the Will in a Cross-​Cultural Perspective, edited by M. Dasti and E. Bryant, 164–​85. New York: Oxford University Press. Garfield, Jay. 2015. Engaging Buddhism: Why It Matters to Philosophy. New York: Oxford University Press. Garfield, Jay, and Graham Priest. 2002. “Nāgārjuna and the Limits of Thought.” Philosophy East and West 53 (1): 1–​21. Garrett, Don. 1997. Cognition and Commitment in Hume’s Philosophy. New York: Oxford University Press. Garrett, Don. 1999. “Owen on Humean Reasoning.” Hume Studies 26 (2): 291–​304. Garrett, Don. 2006. “Hume’s Conclusions in ‘Conclusion of This Book.’” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 151–​ 75. Oxford:  Oxford University Press. Garrett, Don. 2007. “Reasons to Act and Reasons to Believe: Naturalism and Rational Justification in Hume’s Philosophical Project.” Philosophical Studies 132 (1): 1–​16. Garrett, Don. 2015. Hume. New York: Routledge. Gibson, James J. 1979. “The Ecological Approach to Visual Perception of Pictures.” Leonardo 11: 227–​35. Gopnik, Alison. 2009. “Could David Hume Have Known about Buddhism? Charles Francis Dolu, the Royal College of La Flèche, and the Global Jesuit Intellectual Network.” Hume Studies 35: 5–​28. Govind, Rahul. 2015. The Infinite Double: Persons: Things /​Empire: Economy. Simla: Indian Institute for Advanced Study. Govind, Rahul. 2017. “The King’s Plunder, the King’s Justice: Sovereignty in British India, 1756–​76.” Studies in History 33 (2): 1–​36. Green, Thomas Hill, and Ramon M. Lemos. 1968. Hume and Locke. New York: Crowell. Hale, Matthew. 1713. History of Common Law. Chicago: University of Chicago Press. Hardin, Russell. 2007. David Hume:  Moral and Political Theorist. New  York:  Oxford University Press. Home, Henry, Lord Kames. 2000. “Essays on the Principles of Morality and Natural Religion.” In Early Responses to Hume’s Metaphysical and Epistemological Writings, edited by James Fieser, 3:117–​34. Bristol: Thoemmes Press. Hull, Clark. 1943. Principles of Behavior. New York: New Century. Hume, David. 1932. The Letters of David Hume. Edited by J. Y.  T. Greig, 2 vols. Oxford: Clarendon Press. Hume, David. 1975. Enquiries Concerning Human Understanding and Concerning the Principles of Morals. Edited by L. A. Selby-​Bigge and P. H. Nidditch. 3rd ed. Oxford: Oxford University Press. Hume, David. 1978. A Treatise of Human Nature. Edited by L. A Selby-​Bigge and P. H. Nidditch. 2nd ed. Oxford: Oxford University Press. Hume, David. 1983–​85. The History of England: From the Invasion of Julius Caesar to the Revolution in 1688. 6 vols. Indianapolis, IN: Liberty Classics.

References  ( 285 ) Hume, David. 2007. A Dissertation on the Passions and The Natural History of Religion. Edited by Thomas Beauchamp. Oxford: Oxford University Press. Huntington, C. W., and Geshe N. Wangchen. 1995. The Emptiness of Emptiness:  An Introduction to Early Indian Maedhyamika. Honolulu: University of Hawaii Press. Hutcheson, Francis. 1969. An Essay on the Nature and Conduct of the Passions and Affections; with Illustrations on the Moral Sense. Gainesville, FL: Scholars’ Facsimiles & Reprints. James, William. 1918. Principles of Psychology. Vol. 1. New York: Dover. Kail, P. J.  E. 2001. “Hutcheson’s Moral Sense:  Skepticism, Realism, and Secondary Qualities.” History of Philosophy Quarterly 18 (1): 57–​77. Kail, P. J.  E. 2007. Projection and Realism in Hume’s Philosophy. New  York:  Oxford University Press. Kemp Smith, Norman. 1905. “The Naturalism of Hume (I.).” Mind 14 (54): 149–​73. Kemp Smith, Norman. 1941. The Philosophy of David Hume: A Critical Study of Its Origins and Central Doctrines. London: Macmillan. Klibansky, Raymond, and Ernest Campbell Mossner, eds. 2017. New Letters of David Hume. Oxford: Oxford University Press. Korsgaard, Christine. 1986. “Skepticism about Practical Reason.” Journal of Philosophy 83 (1): 5–​25. Korsgaard, Christine. 1996. The Sources of Normativity. Cambridge:  Cambridge University Press. Korsgaard, Christine. 2009. Self-​Constitution:  Agency, Identity and Integrity. Oxford:  Oxford University Press. Kripke, Saul A. 1982. Wittgenstein on Rules and Private Language:  An Elementary Exposition. Cambridge, MA: Harvard University Press. Kuzminski, Adrian. 2008. Pyrrhonism:  How the Ancient Greeks Reinvented Buddhism. Lanham, MD: Lexington Books. Laird, John. 1932. Hume’s Philosophy of Human Nature. London: Methuen. Lewis, David. 1969. Convention:  A Philosophical Study. Cambridge, MA:  Harvard University Press. Loeb, Louis E. 1977. “Hume’s Moral Sentiments and the Structure of the Treatise.” Journal of the History of Philosophy 15: 395–​403. Loeb, Louis E. 2005. Stability and Justification in Hume’s Treatise. New  York:  Oxford University Press. Loeb, Louis E. 2006. “Psychology, Epistemology, and Skepticism in Hume’s Argument about Induction.” Synthese 152 (3): 321–​38. Loeb, Louis E. 2011a. “Epistemological Commitment in Hume’s Treatise.” In Oxford Studies in Early Modern Philosophy, vol. 6, edited by Daniel Garber and Donald Rutherford, 309–​47. Oxford: Oxford University Press. Loeb, Louis E. 2011b. “Inductive Inference in Hume’s Philosophy.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 106–​25. Malden, MA: Wiley-​Blackwell. Malebranche, Nicolas. 1879. De la recherche de la vérité. Vol. 1. Paris: Garnier Freres. Malherbe, Michel. 2010. “Hume on Belief in the External World.” In A Companion to Hume, edited by Elizabeth S. Radcliffe, 126–​39. Malden, MA: Wiley-​Blackwell. Mandeville, Bernard. 1989. The Fable of the Bees. First Printing edition. London: Penguin. Marušić, Jennifer Smalligan. 2016. “Locke’s Simple Account of Sensitive Knowledge.” Philosophical Review 125 (2): 205–​39. Mates, Benson, ed. 1996. The Skeptic Way:  Sextus Empiricus’s Outlines of Pyrrhonism. Oxford: Oxford University Press. Matilal, Bimal K. 1970. “Reference and Existence in Nyāya and Buddhist Logic.” Journal of Indian Philosophy 1 (1): 83–​110.

( 286 )  References McArthur, Neil. 2007. David Hume’s Political Theory: Law, Commerce and the Constitution of Government. Toronto: University of Toronto Press. McEvilley, Thomas. 1981. “Early Greek Philosophy and Mādhyamika.” Philosophy East and West 31 (2): 141–​64. McEvilley, Thomas. 1993. “Hume and the Problem of Personal Identity.” In The Cambridge Companion to Hume, edited by David Fate Norton and Jacqueline Taylor, 177–​208. New York: Cambridge University Press. McEvilley, Thomas. 2002. The Shape of Ancient Thought. New York: Allworth Press. McEvilley, Thomas. 2012. The Shape of Ancient Thought:  Comparative Studies in Greek and Indian Philosophies. E-​book. New York: Allworth Press. McIntyre, Jane L. 1979. “Further Remarks on the Consistency of Hume’s Account of the Self.” Hume Studies 5 (1): 55–​61. Miller, David. 1980. “Hume and Possessive Individualism.” History of Political Thought 1: 261–​78. Millican, Peter. 1998. “Hume on Reason and Induction.” Hume Studies 24 (1): 141–​59. Morris, William E. 2006. “Belief, Probability, Normativity.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 77–​94. Oxford: Blackwell. Nāgārjuna. 1995. The Fundamental Wisdom of the Middle Way:  Nāgārjuna’s Mulamadyamakakārikā. Translated by Jay Garfield. Oxford:  Oxford University Press. Norton, David Fate. 1982. David Hume, Common-​Sense Moralist, Sceptical Metaphysician. Princeton, NJ: Princeton University Press. Owen, David. 1999. Hume’s Reason. New York: Oxford University Press. Passmore, John A. 1952. Hume’s Intentions. Cambridge: Cambridge University Press. Pears, David. 1990. Hume’s System:  An Examination of the First Book of His Treatise. Oxford: Oxford University Press. Penelhum, Terence. 2000a. “Hume’s Moral Psychology.” In Themes in Hume:  The Self, the Will, Religion, 127–​55. Oxford:  Clarendon Press; New  York:  Oxford University Press. Penelhum, Terence. 2000b. “Self-​Identity and Self-​Regard.” In Themes in Hume:  The Self, the Will, Religion, 61–​87. Oxford:  Clarendon Press; New  York:  Oxford University Press. Penelhum, Terence. 2000c. Themes in Hume: The Self, the Will, Religion. Oxford: Clarendon Press; New York: Oxford University Press. Pocock, J. G.  A. 2016. The Machiavellian Moment:  Florentine Political Thought and the Atlantic Republican Tradition. 2nd ed. Princeton, NJ: Princeton University Press. Popkin, Richard H. 1980. “David Hume:  His Pyrrhonism and His Critique of Pyrrhonism.” In The High Road to Pyrrhonism, 103–​32. San Diego, CA:  Austin Hill Press. Pringle-​Pattison, Andrew S. 1890. Scottish Philosophy: A Comparison of the Scottish and German Answers to Hume. Edinburgh: Edinburgh Blackwood. Qu, Hsueh. 2014. “Hume’s Practically Epistemic Conclusions?” Philosophical Studies 170 (3): 501–​24. Qu, Hsueh. 2015a. “Hume on Mental Transparency.” Pacific Philosophical Quarterly 97: 576–​601. Qu, Hsueh. 2015b. “Justifying Hume’s Epistemology.” 1. New York University. Qu, Hsueh. 2017. “Hume’s Dispositional Account of the Self.” Australasian Journal of Philosophy 95 (4): 644–​57. Rawls, John. 2000. Lectures on the History of Moral Philosophy. Edited by Barbara Herman. Cambridge, MA: Harvard University Press.

References  ( 287 ) Richter, Henry J. [1797] 2000. “On Mr. Hume’s Account of the Origin of the Idea of Necessary Connection.” In Early Responses to Hume’s Metaphysical and Epistemological Writings, edited by James Fieser, 4:20–​ 26. Bristol: Thoemmes Press. Rorty, Amelie Oksenberg. 1982. “From Passions to Emotions and Sentiments.” Philosophy 57 (220): 159–​72. Rorty, Amelie Oksenberg. 1993. “From Passions to Sentiments:  The Structure of Hume’s ‘Treatise’.” History of Philosophy Quarterly 10 (2): 165–​79. Rorty, Amelie Oksenberg. 2006. “‘Pride Produces the Idea of Self’:  Hume on Moral Agency.” Australasian Journal of Philosophy 68 (3): 255–​69. Roth, Abraham S. 2006. “Causation.” In The Blackwell Guide to Hume’s Treatise, edited by Saul Traiger, 95–​113. Oxford: Blackwell. Russell, Paul. 1995. Freedom and Moral Sentiment:  Hume’s Way of Naturalizing Responsibility. New York: Oxford University Press. Russell, Paul. 2008. The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion. New York: Oxford University Press. Schmitt, Frederick F. 2014. Hume’s Epistemology in the Treatise: A Veritistic Interpretation. New York: Oxford University Press. Schneewind, J. B. 1983. “Moral Crisis and the History of Ethics.” Midwest Studies in Philosophy 8 (1): 525–​39. Sellars, Wilfrid. 1963. Science, Perception, and Reality. London:  Routledge and Kegan Paul. Sellars, Wilfrid. 1977. “Hochberg on Mapping, Meaning, and Metaphysics.” Edited by P. French, T. Uehling, and H. Wettstein. Midwest Studies in Philosophy 2: 214–​24. Sellars, Wilfrid. 1978. “The Role of the Imagination in Kant’s Theory of Experience.” In Categories: A Colloquium, edited by Henry W. Johnstone. State College: Penn State University Press. Sellars, Wilfrid. 1980. “Behaviorism, Language, and Meaning.” Pacific Philosophical Quarterly 61: 3–​30. Sextus Empiricus. 1985. Sextus Empiricus:  Selections from the Major Writings on Scepticism, Man and God. Edited by Philip P. Hallie. Indianapolis, IN: Hackett. Sextus Empiricus. 1996. The Skeptic Way:  Sextus Empiricus’s Outlines of Pyrrhonism. Edited by Benson Mates. New York: Oxford University Press. Shaftesbury, Anthony A.  C. 1964. Characteristics of Men, Manners, Opinions, Times. Edited by John M. Robertson. 2 vols. in 1. Indianapolis, IN: Bobbs-​Merrill. Shaver, Robert. 1995. “Hume’s Moral Theory?” History of Philosophy Quarterly 12 (3): 317–​31. Sidelle, Alan. 1989. Necessity, Essence, and Individuation: A Defense of Conventionalism. Ithaca, NY: Cornell University Press. Smith, Adam. 1759. The Theory of the Moral Sentiments. London: Printed for A. Millar, and A. Kincaid and J. Bell. Stove, David C. 1966. “Hume, Probability, and Induction.” Philosophical Review 74 (2): 160–​77. Stove, David C. 1973. Probability and Hume’s Inductive Scepticism. Oxford:  Oxford University Press. Strawson, Galen. 1989. The Secret Connexion:  Causation, Realism, and David Hume. Oxford: Oxford University Press. Strawson, Galen. 2002. “David Hume: Objects and Power.” In Reading Hume on Human Understanding: Essays on the First Enquiry, edited by P. J.  R. Millican, 231–​58. Oxford: Clarendon Press; New York: Oxford University Press.

( 288 )  References Strawson, Peter F. 1985. Skepticism and Naturalism: Some Varieties. New York: Columbia University Press. Stroud, Barry. 1984. The Significance of Philosophical Scepticism. Oxford:  Clarendon Press; New York: Oxford University Press. Swain, Corliss. 1991. “Being Sure of Oneself: Hume on Personal Identity.” Hume Studies 17 (2): 107–​24. Tanaka, R. 2016. “Hume on Nonhuman Animals, Causal Reasoning, and General Thoughts.” Unpublished manuscript. Taylor, Jacqueline. 2002. “Hume on the Standard of Virtue.” Journal of Ethics 6 (1): 43–​62. Thompson, E. P. 1991. Customs in Common:  Studies in Traditional Popular Culture. New York: New Press. Thompson, Evan. 2010. Mind in Life: Biology, Phenomenology and the Sciences of Mind. Cambridge, MA: Harvard University Press. Tolman, Edward C. 1951. “Physiology, Psychology and Sociology.” In Behavior and Psychological Man, 179–​89. Berkeley: University of California Press. Townley, Cynthia. 2003. “Trust and the Curse of Cassandra.” Philosophy and the Contemporary World 10 (2): 105–​11. Traiger, Saul. 1987. “Impressions, Ideas and Fictions.” Hume Studies 13 (2): 381–​99. Tsongkhapa. 2006. An Ocean of Reasoning:  Tsong Kha Pa’s Great Commentary on Nāgārjuna’s Mūlamadhyamakakārika. Translated by Jay Garfield and Geshe N. Samten. New York: Oxford University Press. Vygotsky, Lev S. 1978. Mind in Society: Development of Higher Psychological Processes. Edited by Michael Cole, Vera John-​Steiner, Sylvia Scribner, and Ellen Souberman. Cambridge, MA: Harvard University Press. Waxman, Wayne. 1992. “Hume’s Quandary Concerning Personal Identity.” Hume Studies 18 (2): 233–​54. Westerhoff, Jan. 2010. Dispeller of Disputes:  Nāgārjuna’s Vigrahavyāvartanī. New York: Oxford University Press. Whelan, Frederick G. 1985. Order and Artifice in Hume’s Political Philosophy. Princeton, NJ: Princeton University Press. Wigmore, J. 1983. Evidence in Trial at Common Law. Edited by P. Tiller. Boston: Little, Brown. Wilson, Fred. 1997. Hume’s Defence of Causal Inference. Toronto: University of Toronto Press. Winch, Donald. 1996. Riches and Poverty: An Intellectual History of Political Economy in Britain, 1750–​1834. Cambridge: Cambridge University Press. Winkler, K. P. 2000. “All Is Revolution in Us:  Personal Identity in Shaftesbury and Hume.” Hume Studies 26 (1): 3–​40. Wittgenstein, Ludwig. 1953. Philosophical Investigations. Translated by G. E. M. Anscombe. Edited by P. M. S. Hacker and Joachim Schulte. 4th ed. Chicester: Wiley-​Blackwell. Wolff, Robert Paul. 1960. “Hume’s Theory of Mental Activity.” Philosophical Review 69 (3): 289–​310. Wright, John P. 1983. The Sceptical Realism of David Hume. Minneapolis: University of Minnesota Press. Wright, John P. 2000. “Hume’s Causal Realism: Recovering a Traditional Interpretation.” In The New Hume Debate, edited by R. J. Read and K. A. Richman, 88–​99. Abingdon: Routledge.

I N DE X

Abramson, K., 248n30 Abstraction, Aristotle on, 114 Berkeley on, 114 Accident and substance, 111 Action, and passions, 30–​31 Aesthetics, 33 and ethics, 253 and judgement, 257 Agency,  86–​89 and causation, 85 and freedom, 23, 32, 82n8, 85–​89 Agogē, 11, 12n15, 16–​17 Ainslie, D., 73n1, 77n3, 155, 158n8, 222n14 on language, 115n6 on passions, 68n26 on persons, 208n4 on property rights, 115n6 on psychology, 58n13, 68n26 on sympathy, 58n13 Alanen, L. 8n8, 54, 58n12 Allison, H., 122n10, 144n13, 159n10, 162n12 on external world, 199n15 on ideas, 121n10 on justification, 221n13 on mind, 147n16 on necessary connection, 146 on Pyrrhonian skepticism, 221n13 on uniformity of method, 146 Anasademus, tropes of, 152n2 Arbuthnot, J., 6n6 Archimedean fulcrum, 155n5, 201n19, 243n22, 278 Archimedean transcendentalism, 219 Ardal, P., 55 Aristotle, 117n8

on abstraction, 114 on custom, 248 on habit, 120 on human behavior, 80 on morality, 9n10 Artifice,  246–​47 Enquiry on, 242n21 and justice, 16n21, 24n33, 26n34, 241–​42,  243–​44 Mandeville on, 250n33, 255 and nature, 58, 238–​44, 249–​50, 250–​51n34, 253–​54,  277 Āryadeva, 246n26 Ataraxia, 108, 222 Atheism and Pyrrhonism, 6 Atomism, 276 Augustine, 87–​88, 87n11 Authority,  41–​42 and custom, 45 of reason, 170–​71   Bacon, F., 35, 79 Baier, A., 29–​37, 128, 142–​43, 164n14, 244 on communitariansim, 3 on custom, 19, 26n35, 129, 223n15 on determination of the mind, 140 on epistemology, 170n18 on ethics, 170n18, 226, 227n2 on freedom, 23n32 on impressions, 26n35 on justice, 32–​33, 239–​40n18, 245n24, 249–​50, 251n35 on natural sympathy, 33 on persons, 208n4 on reason, 170–​71, 220, 223n15 on skepticism, 153n3, 168, 223n15 on society, 243n23 ( 289 )

( 290 )  Index Balfour, 152n2 Baxter, D., on epistemology, 271n4 on external world, 201n18 on ideas, 67n25, 115n6 on identity, 186–​87n10, 267n1 on metaphysics, 5n2, 6n3 on naturalism, 271n4 on nominalism, 115n6 on Pyrrhonism, 10n12, 12n15, 13n16, 201n18 on skepticism, 16n21 Bayesin inference, 157n7 Bayle, P., 13, 14–​15, 20, 20n27, 152–​53, 160, 163n13, 278–​79 Beauchamp, T., 265–​66 Beck, L., 124, 145 Beckwith, C., 4n1, 20n27 Beebee, 121n10, 129n3, 140, 142n9, 147n17 Behavior, 80, 85–​86, 89, 273–​75, 275–​76n1 and Buddhism, 18n25 and convention, 32–​33 and naturalism, 85 and passions, 85 as passive, 33–​34 and society, 107–​8, 109–​10 Behaviorism, 33–​34, 51–​52, 82–​83, 192–​93,  279 Belief, 23, 25–​26, 121, 121n10, 155–​56, 165, 167, 169–​70 and causation, 77 as custom, 152n1 in external world, 64, 121, 129, 129n3, 172–​76, 177n6, 179–​81, 218, 247 Fogelin on, 174n3 and imagination, 172, 174–​75, 175n5 justification of, 185–​86, 195, 227 moral, 97n8 psychology of, 45n15 and Pyrrhonism, 152n1, 162–​63 and reason, 162–​67, 173–​75 suspension of, 45n15 Bell, M., 129n2 Bennett, J., 155–​56 Bentham, J., 10 Berkeley, G., 177, 210, 264 on abstraction, 114

and conceivabilty-​possibility link, 64n20 Dialogues, 177, 178 on empiricism, 152n1, 278–​79 on epistemology, 219 on external world, 177n6 on human nature, 106 on imagination, 98 and idealism, 210n5 on ideas, 62–​63 on knowledge, 97 on language, 116–​17 on metaphysics, 6 and nihilism, 10 on perception, 117n8, 180n7 on reasoning, 158 on senses, 97 and skepticism, 186 Biology, 15 and custom, 275, 279 and morality, 229–​30 and psychology, 271–​72 Biro, J., 6n4 Blackburn, S., 121–22n10, 143n10 Bricke, J., 89n13 Broughton, J., 6n4, 153n4, 168n17, 169 Brown, C., 92–​93 Buddhaghosa, 233 Buddhism, 276–​77n2,  278–​79 and compassion, 246n26, 254n38 and convention, 18n25, 231n6, 279n6 and custom, 18n25, 20n27, 279n6 and ethics, 56n9, 94n5, 95n6, 246n26, 247n27 and external world, 189n11 and identity, 212n8 middle path, 11n13 and psychology, 79n6, 82–​83 and Pyrrhonism, 20n27 and reality, 277n3 and self, 69n27, 74n2 and senses, 52n5 and skepticism, 3–​4 and truth, 277–​78n4 Burke, E., 36 Butler, J., 93n4   Campbell, G., 20n27 Canalization, 82n8 Canberra plan, 64n20

Index  ( 291 ) Candrakīrti, 3–​4, 276–​77n2 on convention, 6n4, 22n30, 74n2 on custom, 20n27 Introduction to the Middle Way, 95n6, 204, 211n6 Lucid Words, 161n11 on person, 66, 66n23, 66n24, 74n2, 208n4 on self, 66, 66n23, 66n24, 74n2, 208n4 Capaldi, N., 249n31 Carter, S., 38 Cartwright, N., 110–​11 Causation, 13–​14, 26, 125n13, 125–​27,  128–​51 and agency, 85–​87 and belief, 76–​77 and custom, 128–​51 Dignāga on, 61–​62n19 Enquiry on, 121n10, 132n5 and externality, 122–​24 and external world, 118–​24 and freedom, 31–​32, 85–​89 and identity, 203, 214–​15 and induction, 29n1 and judgement, 252 and mind, 31–​32 and nihilism, 145–​46 and realism, 121n10, 142, 145–​46 reality of, 125n13 and reasoning, 31–​32n3, 117n8, 123–​ 24, 127, 129, 130n4 and skepticism, 118, 131–​33, 168 Change,  211–​12 Cheyne, G., 6n6, 225n1 Christianity, 20n27, 23n31, 59n15, 68n26, 87n11, 200n16 Clifford, A., 37 Cognitive science, 49, 57, 82n8, 269–​70, 271–​72,  276 Cohen, L., 157n7 Cohon, R., 227n4, 234n11, 258n40, 258n41 on empiricism, 227 on justice, 250n33 on morality, 58n12, 227n2, 235n14, 247n28, 253n37 on naturalism, 227 on passions, 11n13, 58n12, 258n41 on sympathy, 58n12

on virtue, 241, 254–​55 on will, 82n8 Coke, E., 38n10 Communitarianism, 3, 105, 278–​79, 280 Compassion and Buddhism, 246n26, 254n38 Conceivability-​possibility link,  63–​64 Berkeley on, 64n20 Descartes on, 64n20 Connectionism, 82, 276 Connolly, J., 60n16, 82n8, 89n13 Convention, 12–​13, 211, 249–​51, 254–​55, 276–​77n2, 277–​78n4,  278–​79 and behavior, 32–​33 and Buddhism, 231n6, 279n6 Candrakīrti on, 22n30, 74n2, 277–​78n4 custom as, 227, 244–​51 and epistemology, 11–​12 and identity, 204–​6, 214–​15 and justice, 241n20, 250–​51n34 and morality, 90–​93, 227–​31, 247–​51 and ontology, 11–​12 and psychology, 275–​76n1 and reality, 211n6, 276–​77n2 and truth, 211n6, 276–​77n2 Copernicus, N, 7–​8n7 Copy Principle, 96, 111n3 Coventry, A., 143n10, 150n21 Cover Principle, 4–​9, 15n20, 121n10, 219–​20,  221 and ethics, 6 and human nature, 74–​75, 111, 150, 227 and morality, 227–​28 and psychology, 5, 150 Cowherds, 20n27, 246n26, 276–​77n2, 277–​78n4 Custom, 10, 12n15, 14–​15, 18, 25, 28, 31–​32n3, 32–​33, 35–​46, 76, 78, 142–​43, 145–​46, 159–​60, 164–​65, 164n14, 165n15, 170, 206, 208, 214–​15, 222–​23, 243n22, 249n31, 270–​71, 274–​76,  278–​79 Aristotle on, 248 and authority, 45 Baier on, 19, 26n35, 129, 223n15 as biological, 274–​75, 279 and Buddhism, 20n27, 279n6 Candrakīrti on, 20n27

( 292 )  Index Custom (Cont.) and causation, 150, 182–​83, 214–​15 centrality of, 17–​18, 28, 115n5, 125–​26n14,  222n14 concealed influence of, 96–​97, 98–​99 as convention, 17–​18, 19–​20, 106–​7, 115n5, 170–​71, 208, 227, 275, 279 Enquiry on, 18–​19, 153n3, 249n32 and ethics, 19 and fiction, 213–​15 as habit, 17, 22n29, 33–​34, 46, 82, 115, 115n5, 139, 150, 239, 248 and human nature, 219 Hutcheson on, 33 and justification, 43n13, 45, 153n4, 182–​83, 195, 200, 201n19, and law, 21, 28, 35–​46, 78n4, 157, 226–​27,  230–​31 and knowledge, 34 Mandeville on, 33, 94–95n5, 244 and meaning, 22–​23 and morality, 227–​28, 230, 239–​ 40n18, 243–​44, 250–​51, 256, 257–​59,  258n41 and moral sentiments, 244, 255, 258n41, 278 and naturalism, 26n35, 45–​46 and nomicity, 35 and normativity, 22–​23, 28, 35, 36–​ 39, 45–​46, 227, 235n13, 247 and ontology, 113–​14 and passions, 82–​83, 224, 236 and persons, 273–​80 and Pyrrhonism, 15, 18–​19, 125–​ 26n14, 152n1, 165–​67, 219–​24, 227–​28,  233n9 and reason, 201, 220–​21 science of, 109–​10 Sextus Empiricus on, 15 Shaftesbury on, 33, 57n10 and virtue, 241n20   Darwall, S., 239–​40n18, 248n30 Dauer, F., 22n29, 129n3, 130, 135n7 Davidson, R., 72 Davies, J., 39n11 de Barentane, Marquise, 239n17 Dennett, D., 276 Descartes, R.

on conceivability-​possibility link, 64n20 on epistemology, 219 on external world, 150–​51, 210 on human nature, 106 on idealism, 210n5 on ideas, 62–​63 on imagination, 97–​98 on knowledge, 155 Meditations, 125n12 on mind, 8 on perception, 33 on reason, 34n5, 116–​17 Desideri, I., 20n27 Determinism,  85–​87 Dialogues Concerning Natural Religion, 13 Dignāga, on causation, 61–​62n19 Investigation of the Percept, 61–​62n19 Discourse Untangling the Thought, 189n11 Dissertation on the Passions, 30, 60n16, 232n7, 246n26 Duckworth, D., 61–​62n19   Ecological validity, 108 Ekman, P., 72 Elliot, G., 200n16 Elsky, S., 37, 39n11, 40n12 Em, A., 59–​60 Enquiries, 28, 29–​30, 116n7, 194n13 on artifice, 242n21 on causation, 121n10, 132n5 on Copernicus, 7–​8n7 on custom, 18–​19, 153n3, 249n32, 279 on human nature, 15n20, 18–​19 on induction, 121n10, 132n5 on justice, 240n19, 251n35 on morality, 92n3 on necessary connection, 143n12 on Pyrrhonism, 278n5 on reasoning, 147n17, 157n6 on skepticism, 14n18, 18–​19 on sympathy, 249n32 Epochē, 10–​12, 14–​16, 113, 142–​43, 200, 211 Ethics, 6, 90–​93, 225–​59, 272n5 and aesthetics, 253 Baier on, 170n18, 226–​27, 227n2 and Buddhism, 246n26, 247n27

Index  ( 293 ) and Cover Principle, 6 and custom, 20 and empiricism, 23n31 and epistemology, 26n35, 170n18 Laird on, 227n3 and naturalism, 234n11, 250–​51n34 and passions, 247–​48n29 and psychology, 249n31, 249n32 and Skeptical Inversion, 10 Evidence, 155, 157n7, 158, 159–​60 Externality and causation, 124–​25 and internality, 198 External world, 10, 119–​23, 147n18, 177–​81, 186, 189–​94, 195–​202, 210,  276–​77 Allison on, 199n15 Baxter on, 201n18 belief in, 121–​22, 129, 129n3, 172–​ 89, 218–​23, 247 Berkeley on, 186 and Buddhism, 189n11 and causation, 118–​27 denial of, 196–​97, Fogelin on, 174n3 and imagination, 179–​81 Locke on, 186, 194, 194n14 Malebranche, 177n6, 180n7 and Pyrrhonism, 172–​73, 195, 198–​202,  215–​19 and reason, 178–​79 and senses, 176–​78 and skepticism, 10, 186, 195–​202, 210–​11,  218–​23 Wright on, 177n6, 180n7, 194n13 Externalism, 132n5 Falkenstein, L., 119n9 Feyerabend, P., 110–​11 Fiction, 213–​16,  276–​78 moral, 235n13 and truth, 277–​78 Fogelin, R. on belief, 174n3 on external world, 173n1 on Hume’s language, 155 on identity, 186–​87n10 on necessary connection, 130n4 on perception, 264 on phenomenalism, 175n5 on probability, 162n12

on senses, 174n3 on skepticism, 3, 6n4, 13n16, 130n4, 153n3, 167n16, 173n1 Fortesque, D., 36, 38n10 Fourfold prescription, 12–​13, 14–​15, 16, 19, 113–​14, 169, 221, 221n13 Frede, M., 16n21, 17n24 Freedom and agency, 23, 82n8, 85–​89 Baier on, 23n32 and causation, 31, 85–​89 and morality, 87n11, 88n12 Free rider problem 245n24 Freud, S 148n20   Gamboa, S., 116n7 Game theory, 245n24 Garrett, D. Copy Principle, 96, 111n3 on epistemology, 158n9 on freedom, 23n32 on imagination, 98, 111n3 on induction, 132n5 on missing shade of blue, 111n3 on morality 249n31, 250–​51 on perception, 264–​65 on probability, 158 on psychology, 117n8 separability principle, 190–​91 on skepticism, 11n14, 125–​26n14,  152n1 Title Principle, 15n20, 219–​20,  221n13 on virtue, 249n31 Gibson, J., 109 Gopnik, A., 20n27, Govind, R., 36–​37, 38n9 Green, T., 22n29, 193n12, 211n7 Greig, J., 20n27 Grünbaum, A., 110–​11   Habit, 45, 51–​52, 118, 217 Aristotle on, 82–​83 custom as, 17–​3, 22n29, 33–​34, 45–​ 46, 62–​63, 109–​10, 114–​16, 115n5, 131, 139, 171, 227, 247 and justice, 241n20 Hale, M., 37–​38 Hallie, P., 12n15, 16n21 Hardin, R., 97n8, 116n7, 249n31

( 294 )  Index Harvey, G., 40n12 Heidegger, M., 277 Hermeneutics, 29n1 History of England, 35, 45, 251n35 Hobbes, T., 58n12, 246n25 Home, H., 112 Hull, C., 8n8, 34, 192–​93 Hutcheson, F., 25–​26, 29, 237n15, 250n33 on belief, 170 on custom, 32–​34 on mind, 263–​64 on morality, 93, 93n4, 225, 225n1, 226–​27, 227n3, 232n7, 234–​35,  238 on naturalism, 278–​79 on passions, 33–​34, 50–​51, 52–​53, 54, 56, 81n7, 81–​82, 91n1, 93, 170,  236–​37 on sentimentalism, 278–​79 on social nature, 20   Idealism, 119–​20, 210n5, 279–​80 and Berkeley, 119–​20, 210n5 Descartes on, 210n5 Ideas, 22–24, 110–​18,  133–​35 Allison on, 121n10 Baxter on, 67n25, 115n6 Berkeley on, 62–​63, 122 Descartes on, 62–​63 feigned, 67n25 and impressions, 22n29, 34, 100–​1, 110–​18, 133–​35, 140–​41,  203–​4 Locke on, 26n35 of necessary connection, 133–​35 pseudo, 23–​24, 118–​21, 129 relative,  119–​20 Identity Baxter on, 119n9, 186–​87n10, 267n1, 271n4 and Buddhism, 212n8 and convention, 203–​4, 214n11,  214–​15 diachronic, 123, 190, 208–​10, 211–​12,  217 and fiction, 213–​15 Fogelin on, 186–​87n10 intensionality of, 271n4 numerical, 119–​20, 119n9, 184–​85, 185n9, 208–​10, 211–​12, 265

and passions, 68n26, 69–​70, 73, 208n4 Pears on, 185n9, 265 personal, 68n26, 69–​70, 73, 119n9, 176, 203–​16, 208n4, 265, 269–​70 and Pyrrhonism, 186–​87n10, 203 specific, 119–​20, 184–​85, 185n9, 208–​10,  213–​15 Illusion, 61–​62n19 cognitive, 8–​9, 23, 62–​63 and meaning, 113 of self 65n21 Imagination, 83–​84, 98–​100, 116–​17, 124–​26,  196 and belief, 172, 174–​75, 175n5,  265–​66 Berkeley on, 98 and causality, 124–​26 confidence in, 218–​19 and custom, 235n13 Descartes on, 98 and external world, 179–​81 Garrett on, 111n3 Kant on, 98, 116–​17, 117n8 Laird on, 83n9 Locke on, 98 Malebranche on, 83n9 moral, 254–​55, 254n38 and passions, 83–​84, 224 and Pyrrhonism, 125–​26n14 and reason, 34n5, 125–​26n14, 217–​18 and sympathy, 97–​100, 246–​48 Wright on, 174n4 Impressions, 55, 100, 120, 180–​81, 205 and ideas, 9n11, 26n35, 110–​14, 111n3, 131, 133–​34, 203–​4 passions as, 26, 53–​54, 75–​77 of reflection, 26, 52–​54, 75, 97, 131, 139, 144 Induction, 130n4 and causation, 29n1 Enquiry on, 121n10, 132n5 Intensionality of identity, 271n4 Intentionality of passions, 51–​52 Inverse square law of passions, 77–​78,  242   James I, 38n10 James, W., 106n1 Judgment

Index  ( 295 ) aesthetic, 253, 257 and causation, 252 disinterested, 55 moral, 45n15, 91n2, 232n7, 251–​53,  257–​58 Justice and artifice, 240–​44, 248n30 Baier on, 32, 239–​40n18, 245n24, 249–​50,  251n35 Cohon on, 250n33 and convention, 239–​40n18, 241n20, 245n24, 250n33 Enquiry on 240n19, 242n21, 251n35 and the imagination, 246–​47 Mandeville on, 249n32, 250n33 and natural sympathy, 32, 98–​99 and passions, 58–​59 and social contract, 243 and virtue, 240–​44, 249n32, 249–​59 Justification Allison on, 221n13 Baier on, 153n3 and custom, 43n13, 45, 153n3, 169–​ 70, 181–​83, 201n18, 208 and law, 42, 45 Loeb on, 43n13, 153n3 mathematical,  159–​65 moral, 43n13 Nāgārjuna on, 155n5 and skepticism, 10–​12, 159–​65, 169–​ 70, 173, 182, 201n18, 221n13 and stability, 43n13, 153n3   Kail, P., 129n3, 132, 148n20 Kant, I., 6, 80, 118–​19, 122, 124, 145, 220, 270, 270n3, 272, 277n3,  279–​80 Critique of Pure Reason, 117n8, 210n5,  279–​80 on ethics, 87–​88 on imagination, 83–​84, 98, 116–​17,  117n8 theory of synthesis, 73 transcendental unity of apperception, 268 Katalepsis, 201n18 Kemp Smith, N. 3, 5n2, 8–​9n9, 25, 28, 30–​31, 91n2, 129n3, 130n4, 165n15, 263–​64,  278–​79

Knowledge Descartes on, 154–​55 Locke on, 155, 159n10, 186 Malebranche on, 159n10 mathematical,  154–​55 and probability, 156–​58, 159n10, 160 of self, 205 Korsgaard, C., on free rider problem, 245n24 on morality, 231n6, 272n5 on normativity, 43n13, 233n8 on reason and passions, 81n7 on trust, 243n23, Kripke, S., 16n22, 115n5 Kume, A., 148n19 Kuzminksi, A., 4n1   La Fleche, 20n27 Laird, J., on ethics, 227n3 on imagination, 83n9 on passions, 50n2, 57–​58, 81n7 on phenomenalism, 21 on Pyrrhonism, 13n16, 112n4, 152–​53 on reason, 81n7, 125–​26n14 Language and abstract ideas, 114–​18, 275 acquisition of, 116n7 Ainslie on, 115n6 Baxter on, 115n6 Berkeley on, 115, 115n6 and custom, 17 and self-​knowledge, 271 Sellars on, 115, 116n7 Wittgenstein on, 115n5, 115, 116n7 Law, 35–​41, 201n18 common, 36–​40, 41–​42, 78n4 and custom, 21, 28, 38–​41, 78n4, 157, customary, 36–​42, 78n4 and justification, 45–​46 and morality, 241n20 natural, 35 unwritten vs written 41 Leibniz, G., 34n5, 219 Leiniz’ Law, 206 Lewis, D., 229 Locke, J. on epistemology, 219 on external world, 177n6, 186, 194 on ideas, 26n35

( 296 )  Index Locke, J (Cont.) on imagination, 98 on knowledge, 155, 159n10, 186 on moral sentiments, 58n12 on naturalism, 279–​80 on obligation, 243n22 on passions, 26n35, 58n12 on perceptions, 117n8 on probability, 158, 159n10 on rights, 243n22 on senses, 194n14 Loeb, L on epistemology, 132n5, 165n15 on externalism, 132n5 on justification, 43n13, 153n4, 165n15 on missing shade of blue, 6n3 on moral sentiments, 256n39 on passions, 55, 55n8, 256n39 on reason, 165n15 on skepticism, 132n5 on stability, 43n13, 153n4 on sympathy, 55n8   Mackie, J., 254 Malebranche, N. on external world, 177n6, 180n7 on imagination, 83n9 on knowledge, 159n10 on passions, 50n2 on perception, 159n10, 177n6, 180n7 on probability, 158 on psychophysiology, 8, 82–​83, 109, 189n11,  192–​93 on reasoning, 31–​32n3 on self-​knowlege, 8, 147n16   Malherbe, M., 200 Mandeville, B. on artifice, 255 on custom, 32–​33, 244 Fable of the Bees,  31–​33 on justice, 250n33, 250–​51n34 on moral sentiments, 58n12, 241n20, 244 on morality, 94n5, 244 on naturalism, 278–​79 on passions, 32, 33, 58n12, 81n7, 244 on psychophysiology, 8 reason, 81n7

on social nature, 20, 32 on virtue, 255 Marusic, J., 159n10 Mates, B., 10n12, 12n15, 16n21 Matilal, B., 276–​77n2 McArthur, N., 44n14 McEvilley, T., 4n1, 20n27 McIntyre, J., 206n2, 265–​66 Metaphysics, 5n2, 24, 64n20, 106, 119–​20, 150, 201–​2, 206 Baxter on, 5n2, 6n3 Berkeley on, 6 and epistemology, 115n6, 173–​74,  174n3 Pears on, 5n2 and Pyrrhonism, 150 Strawson on, 115n6, 174n3 Wright on, 5n2, 174n3 Mill, J., 10 Miller, D., 231n5 Millican, P., 121n10, 132n5 Missing shade of blue, 110–​11, 110n2, 111n3, 112n4 Garrett on, 111n3 Loeb on, 6n3 Montaigne, M., 13, 20, 20n27 Montesquieu,  36–​37 Moore, G., 24n33 Morality  225–​42 Aristotle on, 233–​34n10 and belief, 92 and Buddhism, 56n9, 95n6, 246n26, 247n27, 253n37 and character, 252–​53 Cohon on, 58n13, 227n2, 235n13, 247n28, 253n37 and convention, 11, 91–​92, 227–​28,  247–​50 and Cover Principle, 227 and custom, 11, 227–​31, 233, 239–​ 40n18, 243–​44, 248–​51, 256, 258, 258n41 and education, 91–​92 Enquiry on, 90–​91 and fiction, 235n13 and freedom, 89n13 Garrett on, 249n31 Hutcheson on, 92n3, 93, 106n1, 225, 226–​27, 227n3, 232n7, 234–​35 and imagination, 238, 246–​48​, 254–​55

Index  ( 297 ) and justification, 43n13 Korsgaard on, 231n6 Locke on, 58n12 Mandeville on, 32–​33, 58n12, 94n5, 241n20, 244 and naturalism, 227n2, 227–​28,  249n31 and normativity, 227n2, 247 and obligation, 272n5 and passions, 58–​59, 58n12, 90, 146, 234, 234n12, 235, 241–​42, 256n39 and perception, 92n3 and phenomenology, 92n3 and Pyrrhonism, 10, 90–​91, 223, 227–​ 28, 229–​30, 232n7, 232–​33, 233n9 Qu on, 227n2 and realism, 233n9 and reason, 232, 235n13, 247–​48n29 and responsibility, 88–​89, 89n13, 252n36 Shaftesbury on, 10n12, 225, 226–​27, 227n3,  234–​35 More, T., 40n12 Morris, W., 31–​32n3 Müller-​Lyer illusion,  69–​70   Nagasena, 206n3 Nāgārjuna Fundamental Verses on the Middle Way, 161n11, 276–​77n2 Vigrahavyāvartanī—​Reply to Objections, 155n5 Naturalism 3, 5n2, 6–​8, 6n4, 24, 26n35, 31–​32n3, 38n9, 71, 87, 88, 89n13, 114–​15, 130n4, 153–​54, 165n15, 171, 172–​73, 174n4, 185–​86, 195, 201, 215–​16, 251n35, 256, 258, 258n41, 270–​72,  278–​80 Baxter on, 271n4 Cohon on, 227 and custom, 45–​46 and ethics, 225n1, 249n31 Mandeville on, 233n9 and mind, 271n4 and morality 227n2, 227–​28 Norton on, 233n9 Pears on, 31–​32n3, 34n5 and Pyrrhonism, 153n4, 226–​27, 233n9, 278 Qu on, 9, 16n22

and skepticism 3, 6n4, 8n8, 16n22, 160–​61, 169, 201, 220, 233n9, 269–​70,  278–​79 Wright on, 174n3 Naturalized epistemology, 5, 5n2, 165n15, 219 Nature and artifice, 58–​59, 238–​44, 247–​51, 253–​57,  276–​80 Necessary connection, 121n10, 128–​29, 130n4, 131, 133–​51 Allison on, 144n13 Enquiry on, 143n12 Fogelin on, 130n4 Nāgārjuna on, 143n12 Pears on, 144n14 Wright on, 121n10 Newton, I. 7, 77–​78, 89, 98–​99, 117n8, 170, 242, 263–​64, 269–​70 Enquiry on, 7–​8n7 Nietzsche, F., 10 Nihilism, 10–​11, 16, 113–​14, 210–​11 and Berkeley, 10 causal, 129 epistemic, 16 and Pyrrhonism, 10–​11, 16, 152n1 Nominalism, 22, 62–​63, 65n21, 105, 114–​17, 215–​16,  228 Baxter on, 115n6 and custom, 45–​46 and ethics, 228 semantic,  22–​23 and skepticism, 45–​46 Normativity, and custom, 28, 36–​40, 240n19, 247, 273 and epistemology, 31–​32n3, 240n19, 247 Korsgaard on, 43n13, 233n8 and morality, 240n19, 247 naturalization of, 246–​50 and necessary connection, 128–​29 source of, 28, 226–​27, 273 Norton, D., 26n35, 58, 91n2, 233n9, 234n11   Obligation, 243n22, 243n23, 258n41 and convention, 243n22 Kant on, 272n5 Locke on, 243n22 moral, 272n5 and rights, 243n22

( 298 )  Index “Of the Original Contract,” 38n9 Ontology, 11, 113–​14, 208–​9 and convention, 11 and custom, 113–​14, 208–​9 and skepticism, 11, 113–​14 Owen, D., 34n5, 158n9   Panfictionalism, 276–277 Passmore, J., 21n28, 175n5, 268n2 Passions, 5n2, 7–​8n7, 31–​32, 49–​89, 164n14, 170–​71, 207–​8,  215–​16 Ainslie on, 68n26 Alanaen on, 58n12 Baier on, 164n14, 170–​71, 207–​8 calm and violent, 25, 52, 54–​56, 58–​59, 81–​83,  258n40 cause of, 71–​79 as cognitive, 52–​54 Cohon on, 58n12, 235n13, 258n40, 258n41 and custom, 81–​83, 150–​51 direct and indirect, 49–​54, 56 Hutcheson on, 33, 50–​51, 52–​53, 54, 56, 81n7, 81–​82, 91n1, 93, 170, 237 as impressions of reflection, 26, 52–​54, 75–​76,  77 indirect, 235n14 intentionality of, 51–​52 as involuntary, 180n7 Laird on, 50n2, 57–​58, 81n7 Loeb on, 55, 55n8 Malebranche on, 50n2 Mandeville on, 33, 58n12, 81n7 and morality, 33, 58n12, 90–​101, 235–​38, 241–​42, 247–​48n29 and moral sentiments, 235–​38,  256n39 object of, 59–​62 and personal identity, 215–​16 pride and humility, 56–​57, 59–​61, 67–​69, 75–​76, 78, 215 Qu on, 65n22 and reason, 53–​54, 79, 164n14,  170–​71 and self, 215–​16, 215n12, 265–​66 as social, 57–​59, 170–​71, 230–​31,  231n6 Shaftesbury on, 33, 50, 52–​53, 58, 81–​82, 91n1, 234n12

virtue and vice, 90–​93, 93n4, 236–​38, 249,  251–​57 Wright on, 7–​8n7, 180n7 Pears, D. on epistemology, 5n2 on identity 185n9, 265 on metaphysics, 5n2 on mind, 265 on naturalism, 31–​32n3, 34n5 on necessary connection, 144n14, 146 on reasoning, 144n14 Penelhum, T., 7, 87, 87n11, 208n4, 215 Perception, 122–​24, 190–​92, 194n13,  196–​97 Berkeley on, 117n8, 180n7 and the external world 176–​78 as involuntary, 180n7 Malebranche on, 177n6, 180n7 moral, 92n3 Wright on, 177n6, 180n7, 194n13 Person, 67, 69–​70, 206, 207–​8, 265, 266–​67,  277 Ainslie on, 208n4 Baier on, 208n4 Candrakīrti on, 66, 66n23, 66n24, 74n2, 211n6 and custom, 273–​80 Hobbes on etymology, 66 Locke on, 66, 117n8 Pears on, 265 reality of, 66–​67 Personal identity, 15n19, 69–​70, 73, 203–​24,  263–​71 Ainslie on, 68n26 as a fiction, 213, 276–​77 and the passions, 207–​8 Phenomenalism, 21 Piece of wax argument, 125n12 Pocock, J., 35–​36, 39 Popkin, R., 3, 10n12, 12n15, 16n21, 152–​53, 152n1, 162n12, 223 Principles of Interpretation, 4 Pringle-​Pattison,A S., 152n1 Probability, 154–​63, 201n17 Allison on, 162n12 calculus, 156, 162–​63 Fogelin on, 162n12 Garrett on, 158, 158n9 and knowledge, 156–​58, 158n9, 160 and the law, 156–​58

Index  ( 299 ) Locke on, 158, 159n10 Malebranche on, 158 Owen on, 158n9 Popkin on, 162n12 Stove on, 162n12 Progress of sentiments, 245–​46 Pseudo-​idea Principle, 22, 64–​67, 85, 87, 111–​13, 123–​24, 129, 147n16, 150, Pym, J., 38n10 Pyrrho, 13, 20n27 Pyrrhonism, 9–​17, 108–​9, 152–​54, 163–​ 67, 173n1, 216–​20, 223, 227–​28, 273,  278–​79 Allison on, 221n13 and atheism, 13n16 Baxter on, 10n12, 13n16, 186–​87n10,  201n18 and belief, 152–​53, 163–​67 and Buddhism, 20n27 and custom, 18–​19, 152n1, 164–​65, 219–​20, 223, 227–​28, 233n9 Enquiry on, 278n5 and ethics, 90–​91, 227–​28, 229–​30,  233n9 and external world, 172–​73, 191–​95,  216–​19 Fogelin on, 13n16, 186–​87n10 Garrett on, 11n14 and identity, 186–​87n10 and imagination, 112n4 Laird on, 13n16, 112n4, 152–​53 Mates on, 10n12 and naturalism, 152–​54, 273, 278–​79 and nihilism, 152n1 Norton on, 233n9 Popkin on, 10n12, 152–​53, 152n1 Pringle-​Pattison on, 152n1 and reason, 152–​54, 163–​67, 216–​17 and skeptical inversion, 9–​13, 203, 211 and Wittgenstein, 278–​79 Questions of King Milinda, 206n3 Qu, H., on epistemology, 16n22, 165n15 on mind, 9n11 on morality, 227n2 on naturalism, 9, 16n22 on passions, 65n22 on self, 65n22 on skepticism, 9, 16n22, 165n15 Quine, W., 110–​11, 201n19, 278–​79

Rawls, J., 258n41 Realism, Beebee on, 121n10, 129n3 and causation, 121n10, 142, 145–​46 and the external world, 121n10,  201–​2 Kale on, 129n3 moral, 38n9, 233n9 skeptical, 121–​22, 129n3, 132 Strawson on, 129n3 Wright on, 121n10, 129n3 Reason authority of, 170–​71 Baier on, 170–​71, 220, 223n15 and custom, 164–​65, 201 Descartes on, 34n5, 116–​17, 169 and external world, 178–​79 and imagination, 34n5, 125–​26n14,  218–​21 as self-​justifying, 159–​60, 163, 217,  218–​21 Korsgaard on, 233n8 Laird on, 81n7, 125–​26n14 Loeb on, 165n15 and necessary connection, 133–​35 Mandeville on, 81n7 and morality, 232, 247–​48n29 and passion, 53–​54, 80–​81 Pears on, 34n5 and Pyrrhonism, 152n1, 163n13,  163–​71 rehabilitation of, 163–​71 Rorty on, 247–​48n29 and skepticism, 152–​71, 181–​83, 198, 201, 223 Reflexivity, and Pyrrhonism, 160–​63, 216–​24 Sextus Empiricus on, 16 Reificationism, 10–​11,  113–​14 Reid, T., 10, 15n19, 130n4, 152n1 Relativism, 43 Mandeville on, 32 Richter, H., 23–​24 Rights, Ainslie on, 115n6 Locke on, 243n22 and obligation, 243n22 of property, 40–​42, 115n6 Roth, A., 128n1

( 300 )  Index Rorty, A., 35n6, 68n26, 247–​48n29,  258n40 Russell, P., 6, 86–​87, 87n10, 88n12, 90–​91,  252n36 Śāntideva, 20n27, 60n18, 74n2, 93n4, 246n26, 247n27 Sawada, K., 12n15, 148n19, 201n17, 231n5 skeptical inversion, 9–​17, 203, 211 Schmitt, R.. 5n2 Schneedwind, J., 93n4, 246n25 Schopenhauer, A., 122, 123n11, 145n15 Self,  203–​24 and Buddhism, 70n27, 74n2 Candrakīrti on, 66, 66n23, 66n24, 208n4 and causation, 61–​62, 203 illusion of, 65n21, 66 impression of, 204–​5 as object of passion, 59–​70, 94–​95 and passions, 59–​70, 75–​79, 94–​95,  215n12 and person, 66–​67 as pseudo-​idea, 22, 64–​67 Qu on, 65n22 Shaftesbury on, 94n5 as social, 94–​95, 274–​77 Sellars, W., 22–​23, 114, 124, 278–​79 on language, 115, 116n7 Myth of the Given, 277 Senses, 98, 172–​202 and Buddhism, 52n5 and external world, 176–​78 Fogelin on, 173n1 and skepticism, 123, 172–​202, 210, 217, Sentimentalism, 225–​27, 234–​39,  278–​79 Sentiments, moral, 55n8, 56n9, 92–​97,  232–​38 and custom, 238–​51, 258n40 Separability principle, 63–​64, 190–​91 Sextus Empiricus and custom, 15 Fourfold prescription, 12, 14–​15, 16, 19, 113–​14, 152–​53, 169, 221 Outlines of Pyrrhonism, 12–​13, 15, 113,  152–​54 and reflexivity, 160–​61

and skepticism, 10–​12, 16–​17, 152–​54, 210, 222, 278–​79 Shaftesbury Alanen on, 58n12 on custom, 33, 57n10 on mind, 7–​8n7 on morality, 225, 226–​27, 227n3, 234–​35,  234n12 on naturalism, 278–​79 on passions, 7–​8n7, 33–​34, 50, 81–​82, 91n1, 234n12 on self, 94n5 sentimentalism, 225, 232n7, 234–​35,  278–​79 and social nature, 58n12, 93–​94 Shaver, R., 227n2 Sidelle, A., 20n27 Skeptical Inversion, 9–​17, 17n24, 113, 128, 169–​70, 203 Skeptical realism, Wright on, 129n3, 132, 174, 174n3 Skepticism, 3–​4, 5n2, 125–​26n14, 216–​24,  278–​79 Academic 152–​53, 152n1, 152n2, 153n3, 160–​61, 216 Allison on, 221n13 Baier on, 153n3, 168, 223n15 Baxter on, 8 and Berkeley, 186 Broughton on, 168–​69 and Buddhism, 3–​4 and causation, 88n12, 118, 130n4, 132n5, 168, 182–​83 and custom, 13–​14, 45–​46, 182–​83 Descartes on, 168n17, 173, 173n1 Enquiry on, 14n18, 18–​19 and external world, 113–​14, 122 Fogelin on, 3, 13n16, 130n4, 153n3, 167n16, 173n1 Frede on., 16n21, 17n24 Garrett on, 11n14, 125–​26n14, 152n1 Hallie on, 16n21 Laird on, 152–​53 Loeb, 43–44n13, 153–54n4 Mates on, 16n21 and morality, 233n9, 228, 258 and naturalism 3, 6n4, 8–​9n9, 9, 16n22, 88n12, 89, 168–​69, 186, 201, 220, 278–​79 and nominalism, 45–​46

Index  ( 301 ) and personal identity, 205–​11 Popkin on, 16n21 Pyrrhonian, 9–​17, 108–​9, 152–​54, 163–​67, 173n1, 216–​20, 223, 227–​ 28, 273, 278–​79 Qu on, 9, 16n22, 165n15 and reason, 16, 108–​9, 152–​71, 220, reflexivity of, 16, 160–​63, 216–​24 and senses, 121n10, 123, 172–​202, 210, 217 Skinner, B.F., 8n8, 51–​52 Skye,  75–​76 Smith, A., 94n5, 99n9, 241n20 Social contract theory, 38n9, 194n14, 243, 243n22 Space and Time, 5n2, 119n9 Spenser, E., 40n12 Stability, 43n13, 55n8 and justification, 153n4 Loeb on, 43n13, 55n8, 153n4 Stewart, J., 125n13 Stove, D., 155, 162n12 Strawson, G., 5n2, 29n1, 121, 127n16, 129n3, 132, 174n3, 174 Strawson, P., 16–​17 Stroud, B., 24n33 Substance, 67n25, 111–​14 Swain, C., 264 Sympathy, 55n8, 94n5, 96–​97, 99n9, 235 Ainslie on, 58n13 Baier on, 32 Candrakīrti on, 95n6 and custom, 32, 245–​47 Enquiry on, 249n32 as force, 99n9 and imagination, 96–​97, 245–​47 Loeb on, 55n8 natural, 32, 57n11, 94–​97, 100–​1, 245–​47, 247–​48n29, 254–​55,  274 Rorty on, 247–​48n29 Shaftesbury on, 94–95n5   Talisker,  75–​76 Tanaka, R., 116n7 Taylor, J., 29n1 Thompson, E., 276 Thompson, E.P.., 35, 39–​40, Title Principle, 15n20, 219–​20 Tolman, E., 8n8, 51–​52, 275–​76n1

Townley, C., 243n23 Traiger, S., 213n9 Truth, and Buddhism, 211n6, 277–​78n4 conventional, 211n6, 277–​78n4 and fiction, 277–​78 fictional,  213–​16 Tsongkhapa, 9, 276–​77n2   Uniformity of Method Principle, 25–​26, 129, 168, 172–​73, 210, 226–​27, 243n22, 252   van Fraassen, B., 110–​11 Vasubandhu, 189n11 Virtue, 59n15, 249–​59 artificial, 249–​51, 253–​58, 275 Cohon on, 241, 254–​55 Garrett on, 249n31 and custom, 241n20 and justice, 242–​51 Mandeville on, 255 natural,  256–​59 social, 249n31 Vygotsky, L., 272, 276   Watson, J., 8n8, 34, 51–​52 Waxman, 270n3 Westerhoff, J., 155n5, 157n7 Whelan, F., 6n6, 18n25, 45n15, 59n15, 251n35 Wigmore, J., 157n7 Willow,  75–​76 Wilson, F., 117n8, 130n4 Winch, D., 241n20 Winkler, K., 121n10 Wittgenstein, L., 16–​18, 16n22, 278–​79 epistemology, 201n18 On Certainty, 121n10, 125n13, 164, 173n2 hinge proposition, 121–22n10, 173n2 on language, 115, 115n5, 116n7 Philosophical Investigations, 280 Tractatus Logico-​Philosophicus, 279–​80 use theory of meaning, 22–​23, 115n5 Wolff, R., 117n8 Wright, J. on epistemology, 5n2

( 302 )  Index Wright, J. (Cont.) on external world, 177n6, 180n7, 194n13 on imagination, 174n3 on metaphysics, 5n2 on mind, 7 on naturalism, 174n3

on necessary connection, 121n10 on passions, 7–​8n7 on perception, 177n6, 180n7, 194n13 on psychophysiology, 126n15, 189n11 on skeptical realism, 129n3, 132, 174n3, 174