Evidentialism and the Will to Believe 9781623560171, 9781472594105, 9781780936642

Citation preview

Evidentialism and the Will to Believe

Also available from Bloomsbury Advances in Experimental Epistemology, edited by James Beebe Epistemology: The Key Thinkers, Stephen Hetherington Pragmatism: A Guide for the Perplexed, Robert B. Talisse and Scott F. Aikin The Continuum Companion to Epistemology, edited by Andrew Cullison

Evidentialism and the Will to Believe Scott F. Aikin

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc

LON DON • OX F O R D • N E W YO R K • N E W D E L H I • SY DN EY

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc 50 Bedford Square London WC1B 3DP UK

1385 Broadway New York NY 10018 USA

www.bloomsbury.com BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published 2014 Paperback edition first published 2015 © Scott F. Aikin 2014 Scott F. Aikin has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as the Author of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. No responsibility for loss caused to any individual or organization acting on or refraining from action as a result of the material in this publication can be accepted by Bloomsbury or the author. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: HB: 978-1-6235-6017-1 PB: 978-1-4742-6583-6 ePDF: 978-1-7809-3664-2 ePub: 978-1-7809-3644-4 Library of Congress Cataloging-in-Publication Data Aikin, Scott F. Evidentialism and the will to believe / Scott F. Aikin. – 1st [edition]. pages cm Includes bibliographical references and index. ISBN 978-1-62356-017-1 (hardback) – ISBN 978-1-78093-664-2 (epdf) – ISBN 978-1-78093-644-4 (epub) 1. Evidence. 2. Knowledge, Theory of. 3. Faith and reason. 4. Clifford, William Kingdon, 1845–1879. Ethics of belief. 5. James, William, 1842–1910. Will to believe. I. Title. BD161.A355 2014 210–dc23 2013045814 Typeset by Integra Software Services Pvt. Ltd. Printed and bound in Great Britain

Contents Acknowledgments Copyright Acknowledgments

viii x

Introduction 1. The objectives of commentary 2. Three themes 3. Five evaluative theses

1

1

9

2

Reading William Kingdon Clifford’s “The Ethics of Belief ” William Kingdon Clifford and the Metaphysical Society Section I – The duty of inquiry 1. The ship owner case 2. The island case 3. Beliefs and actions 4. Beliefs and their consequences 5. Ethics and belief 6. Endorsing evidentialism Section II – The weight of authority 1. Anti-skepticism 2. Testimonial evidence 3. Miraculous testimony 4. The publicity requirement 5. The sacred tradition of humanity Section III – The limits of inference 1. A burnt child dreads the fire 2. Regulative principles 3. Three norms Reading William James’s “The Will to Believe” William James and “The Will to Believe” Preamble

1 4 6

9 13 13 22 27 32 42 47 50 50 53 56 59 64 69 69 72 78 79 79 81

vi

Contents

Section I – Hypotheses and options 1. Introduction and definitions 2. Live and dead hypotheses 3. Forced options 4. Momentous options 5. Religion as a genuine option Section II – Pascal’s Wager 1. Four stages of “The Will to Believe” 2. Voluntarism and its limits 3. The wager 4. Clifford’s veto Section III – Psychological causes of belief 1. A concession to evidentialism 2. Truth and other useful ideas 3. Pascal is a regular clincher Section IV – The thesis of the essay 1. A thematic transition 2. The thesis Sections V and VI – Absolutism and empiricism 1. Two forms of faith 2. Objective evidence and its discontents 3. Truth for empiricism Section VII – Two different sorts of risks in believing 1. The two commandments 2. The case for the Truth Norm 3. Two critical points Section VIII – Some risk unavoidable 1. Applying the meta-epistemology 2. Interested inquiry 3. Two analogies Section IX – Faith may bring forth its own verification 1. Moral and scientific questions 2. Moral skepticism 3. The argument from friendship 4. The argument from social coordination 5. Doxastic efficacy and the Will to Believe

86 86 87 91 95 96 98 98 100 101 102 103 103 106 108 109 109 111 114 114 118 121 123 123 125 128 131 131 133 135 136 136 137 140 145 148

Contents

Section X – Logical conditions of religious belief 1. The overall form of James’s argument 2. Religion’s dual essence 3. Religion as live and momentous 4. Religion as forced 5. The conversion fallacy 6. Religion as doxastically efficacious 7. Evidentialism as irrational 8. Religious tolerance 3

The Ethics of Belief and Philosophy of Religion Question 1: Must evidentialism be an ethical doctrine? Question 2: Can practical reasons trump theoretical reasons? Question 3: Can religion be pragmatically reconstructed? Question 4: What about the power of positive thinking?

Notes Bibliography Index

vii 154 154 160 163 167 170 172 175 178 181 181 184 190 192 195 203 210

Acknowledgments This is a book that has, in some form or other, been brewing for a long time. Since my first exposure to the ethics of belief debates in graduate school, I wanted to know the issue inside and out. One of the real benefits of falling into this matter at that time in my education was that I had a cadre of other like-minded students with whom to discuss the issue. Vanderbilt in the late nineties and early aughts provided a rich variety of attitudes. My two regular partners in conversation on the Clifford–James debate were James Bednar and Allen Coates. My understanding of and views on the ethics of belief are all descendants of the ideas that survived our joint critical scrutiny. Others at Vanderbilt who contributed mightily to my understanding of the issue were Jason Aleksander, Erin Bradfield, Carolyn Cusick, Jeff Edmonds, Paul Ford, Lenn Goodman, David Miguel Gray, Michael Hodges, John Lachs, Mason Marshall, Emily McGill, Jose Medina, Jonathan Neufeld, Brian Ribeiro, Aaron Simmons, Robert Talisse, Jeffrey Tlumak, Derek Turner, and Dylan Wittkower. Additionally, another Vanderbilt alum, Micah Hester organized a session on “The Will to Believe” at the 2012 William James Society, and he was good enough to invite me to present a short version of my main critical line against James. The session was incredibly rewarding. Micah and the other panelists, Jeff Kasser and Michael Slater, provided me with very helpful feedback and justly critical pushback. Others who have given me very useful thoughts on the issue at various conferences and over email have been Jonathan Adler, Guy Axtell, Bryan Baird, Antonio Bendezu, Robert Brandom, John Casey, Mylan Engel, Richard Gale, Cynthia Gayman, Russell Goodman, Charlie Hobbs, Christopher Hookway, Henry Jackman, Ralph Johnson, Jason Kawall, Peter Klein, Mary Magada-Ward, Cheryl Misak, Ted Poston, Harvey Siegel, and Brian Zamulinski. Another very hearty word of thanks is reserved for Colleen Coalter at Bloomsbury for contacting me with the idea of pursuing this book. She, Andrew Wardell, and the rest of the editorial staff at Bloomsbury have been a pleasure to work with. One of my main motives for writing this book was that I thought that Clifford’s view and his argument for it has been all too ignored in the critical literature on the ethics of belief. Beyond the careful work of Brian Zamulinski (2002 and 2004) and Tim Madigan (2012), there have not been really any tight-

Acknowledgments

ix

focus readings of Clifford’s essay or sympathetic replies to James’s criticisms. I concede that Clifford’s argument, ultimately, doesn’t work. Its final move is a confusion of contraries with contradictories—he takes it that if one holds that it is false that All events are parts of the uniformity of nature, then one must hold that No events are uniform. Despite this error, Clifford’s argument fares much better than for what James gives him credit. It’s worthwhile to clarify the view— and to extend and supplement Clifford’s argument where it can be. Evidentialism is the correct attitude, I think, and Clifford’s arguments are the right ethical lines to take when they are properly framed. I, further, believe that far too many of the close readings of James are altogether insufficiently critical. They neither challenge James on his interpretations of Clifford nor do they subject James’s substantive philosophical claims to much scrutiny. That, I think, needs fixing, and I’ll have some very critical lines of argument against James’s position here. My main charge against James’s case is that it relies on an illicit conversion of the plausible All belief directs action to the implausible All action is directed by belief. James’s argument depends on the thought that only by holding a positive belief can the actions comprising and making possible successful living can be performed. But it is far from clear that one must believe in order to act. One most certainly can act without thinking—people do it all the time. A final word of thanks to my wife, Susan Foxman, and our wonderful daughters, Madeleine and Iris. I’ve been blessed in having a supportive and thoughtful family. Both daughters have provided many good critical thoughts on the project, as they’ve regularly been game to discuss the examples from Clifford’s ship owner to James’s social coordination cases. My presentation has improved in clarity because of my practice in explaining the matter to eight- and twelve-year-olds.

Copyright Acknowledgments Short verbatim extracts reprinted by permission of the publisher from The Works of William James: The Will to Believe and Other Essays in Popular Philosophy, Frederick Burkhardt, General Editor, Fredson Bowers, Textual Editor, Ignas K. Skrupskelis, Associate Editor, pp. 13–22, 24–33, Cambridge, MA: Harvard University Press, Copyright © 1979 by the President and Fellows of Harvard College.

Introduction

1. The objectives of commentary William Kingdon Clifford’s “The Ethics of Belief ” and William James’s “The Will to Believe” are yoked together in the story of philosophy.1 The two essays are taken as the classic starting point for reflection on the norms governing responsible belief. Clifford captures his view, evidentialism, with the stark pronouncement that “it is wrong always, everywhere, and for any one, to believe on insufficient evidence.” Clifford, thus, stands poised as the spirit of intellectual honesty, as the one who follows the arguments where they lead, as a thinker who spurns comforting fictions. In contrast, James’s doctrine of the will-to-believe is exemplified by the claim that “our passional nature not only lawfully may, but must, decide an option between propositions, whenever it is a genuine option that cannot by its nature be decided on intellectual grounds.” James offers a defense of the sentiments’ role in intellectual life, he stands as the Romantic resistance to the demands of bloodless reason, he defends belief in the face of withering skepticism. Clifford and James are iconically opposed. Their respective essays are anthologized, taught, discussed, and cited alongside each other as the natural introduction to the ethics of belief. Yet, surprisingly, these essays have never received a book-length joint commentary. That fact is particularly surprising in light of the wide and varied discussions of them in epistemology, American philosophy, and philosophy of religion. The clash between evidentialism and the will-to-believe has spawned many conference essays, journal articles, and book chapters. But, again, none have been devoted to a complete reconstruction of and commentary on the two essays. With this book, that changes. Commentaries are useful scholarly tools, and for influential works that are touchstones for progressively divergent traditions, they are necessary. First, commentaries can show how the deep differences arose from shared background views. With this commentary, I will highlight how much Clifford and James

2

Introduction

share in practical vision and philosophical aspiration. Despite the fact that the two make cases for radically different directions in thought, they share a common set of values of and commitments to philosophy contributing to real practical good. Those who stand as torchbearers of the traditions Clifford and James founded can use a reminder of their common ground. The second reason why commentaries on influential works such as these are useful is that landmark books and essays are too often read in the light of the traditions arising from them and applications to which they are put, but not in the light of their own stated goals and arguments. This happens often, and a literary case may highlight this phenomenon. Upton Sinclair’s The Jungle is a story about how, in the turn of the twentieth century, people working in the Chicago stockyards and meat-packing plants were exploited, but the public response to the book was to the depictions of disgusting conditions of food preparation. Its role in the American story is that it precipitated the Meat Inspection Act, but very few reference its role in the development of the labor movement. In a way, the book is known more for what seems a misinterpretation (if not misplaced emphasis). This is not to say anything as drastic as what happened to Sinclair’s book has happened to Clifford’s and James’s, but that the effects and consequent ways of conceiving of a work can be very different from what actually happens in it. Once a work has that iconic status, it can be difficult to interpret it in anything but the mode of contributing to or making sense of that icon. Again, nothing quite so much is the case with Clifford and James, but there are some distortions that the essays’ reception and status in the received story of philosophy have precipitated. I’ll be working to correct those. In commentaries, especially on iconic works, it is useful to provide historical contextualizations of such intellectual iconography. The historical contextualizer may supplement a reading of a text with other materials the author wrote or was reading. And so we may make sense of what the author said in essay X by finding something similar in book Y, or we may look to a book Z by some other author. In such a process, we place the author’s work in a web of relations, and we see the work as it relates to the rest of the author’s work and that of others. Such is historical commentary, and its demands are high—both in the scholarship to do it and in the patience to read it. I have no intention for this commentary to be heavy historical scholarship. I will reference and explain other works where Clifford and James do, and I will try to briefly set out the stakes for their essays. But otherwise, the rest of the other voices in their milieu will be mostly silent here. The reason is that both Clifford and James wrote their essays for popular consumption, for the

Introduction

3

educated but lay audience. These two essays were not written for scholars. They were written for people who could listen and read closely. Both of these authors wrote their essays to stand on their own, without the need of importing a grand philosophical system to make their claims seem plausible, and without the need of heavy historicizing to make their prose intelligible. As such, this book is an exercise in internal interpretation. The historical and larger philosophical approaches to commentary are devoted to making sense of the texts in terms of how they hook up with things external to them (other essays, other authors, the traditions they motivate). I wish to read these essays on their own terms, as essays that were presented to educated, but not philosophically advanced, audiences. As such, I take them as essays that articulate philosophical positions on the basis of arguments they themselves provide. And as a consequence, my commentary will have relatively sparse scholarly apparatus, as I do not believe one must wade through the various others who have made similar observations or not in order to state the basic argument in the text. The arguments are there, with their markers for premises and conclusions, and we just need to open our eyes and read them. This said, there are places where scholarly disagreements are significant enough to require me to lay out a few interpretive options. I will do my best not to let my own preferences tilt the presentation excessively, but I will be clear about where I believe the evidence leads us. There are certainly those who will make postmodern scoffing noises at my naïve contention that we can go back to the texts themselves. We always read with an agenda, through a lens, or with a bias. That’s how the story goes. However, it’s clear that attitudes like this are self-defeating. Here’s how: when someone who’s read too much French philosophy says “It’s all interpretation” or some such fashionable nonsense, I would nevertheless misinterpret her were I to take her be ordering a cheeseburger, and whereupon hand her one. And that misinterpretation would be because of the fact that I should have read or attended to what the words she spoke or wrote meant. Once we’ve fixed my misinterpretation, we see she was wrong about it all being interpretation. If it’s all interpretation, my interpretation of her sentence isn’t wrong. But it was. So it can’t all be interpretation. Otherwise, there wouldn’t be mis-interpretation. There is interpretation, but interpretation is out to get something right. There is a fact of the matter of what was meant. So as it turns out that what the postmodern critic meant was also false. At least someone got a cheeseburger out of the deal. My point is that this project of internal commentary is, on the one hand, modest. Many of the going scholarly disputes are at most going to have brief exposure. Those disputes are to be worked out piecemeal once we are clear what

4

Introduction

the arguments actually are. On the other hand, though, internal criticism is where it all should start. A close reading of the text, even if the ultimate objective is to connect it up with other works or to interpret according to some program, must start with the text itself. It seems a truism, but in philosophy, saying things that are obviously true is necessary. In this discipline, it’s easy to get lost to the point where you’ve forgotten good sense. It’s best to mark the path out of nonsense clearly. Finally, it is important with these two essays specifically that we approach them with this comportment of internal reading. Since they were presented and published for popular audiences, such modesty is necessary. Neither James nor Clifford expected us to read the essays through the lens of scholars (especially scholars of their own larger corpus), but as thoughtful people. Further, given the traditions the two essays have spawned, it is crucial that we see what of those traditions now are traceable back to these essays and what are later accretions. This book only fixes what’s in the essays. The contrast with the traditions requires more work beyond this commentary, but my final chapter on what I call the lingering questions from the two essays will lay out what I take to be the intellectual terrain one encounters in light of their arguments.

2. Three themes My objective with this commentary is to present the core arguments of W.K. Clifford’s “The Ethics of Belief ” and William James’s “The Will to Believe.” The two essays are written in sections, and my commentary will proceed section-by-section through them. I will explain what happens in each section, how it builds on what happens in previous sections, and what might be said further about the case presented. I will pause here and there to untangle the odd interpretive issue or state (and sometimes answer) what seem to be the most challenging objections. Further, James’s essay is presented as a response to Clifford’s evidentialist viewpoint, and so with James, I will additionally discuss how the case proceeds against Clifford and how the Cliffordian evidentialist might reply. Three themes will emerge in this commentary. The first is that Clifford and James, despite their differences, share a core commitment to a broad form of pragmatism. Their views on how beliefs are acquired, how they function, and why it is important to responsibly hold them both take the form of a concern for the role of belief in practical life. Both philosophers take the view that beliefs

Introduction

5

have not only a representational side, that of presenting the world as one way or another, but also an actional side in being plans for action. Further, both make their respective cases on the basis of the practical consequences that follow from believing one way or another. Both of their cases are practical and ethical arguments for certain norms of belief, and as such, both share a core pragmatic vision. The second theme of this commentary, in light of the authors’ shared pragmatism and ethical aspirations, is the question of how practicable the two views really are. On the one hand, there is the question of how demanding Clifford’s evidentialist requirement really is. Clifford, himself, was aware of this concern, as he notes at the beginning of the second section of his essay that he may be flirting with an untenable global skepticism. He argues that he is not a skeptic, but this protest regularly falls on deaf ears. James’s interpretation of Clifford’s evidentialism certainly contributes not only to this concern but also to the view that it actually is far too stark. A running question, then, is just how practically wise would it be to be an evidentialist of Cliffordian stripes? Is James’s case that it wouldn’t a fair representation of Clifford’s stated views? On the other hand, we have the same question of practicability posed for James’s will-to-believe doctrine. Is it a sound attitude for practice? Is it wise to have confidence in some propositions when there is no evidence to support them? James, like Clifford, sees the extremes to which his view could be taken if untempered and put to use improperly. He holds that when properly deployed, the will-to-believe is not mere self-confidence or wishful thinking run amok. It is a responsible and useful pattern of belief. But then the question is what those conditions for proper deployment are and whether James is right that the willto-believe really is that wise. The third theme is laying out the stakes for Clifford and James’s ultimate philosophical quarry: religious belief. It is important to note that both philosophers agree that the traditional natural theological case for God’s existence has failed. Both agree that the evidence for God is weak at best, certainly not enough to justify religious belief on its own. Clifford’s case proceeds along two lines. The first is simple: because the evidence is not sufficient to show that belief in God is true, one should not believe. That’s just Cliffordian evidentialism. Clifford’s second line is that the evidence also shows that belief in God under these conditions is conducive of other intellectual and moral errors. So it would be no mere one-off case of breaking a rule, one with no consequences if you harbor religious belief without the requisite evidence. To the contrary, Clifford holds, religious belief brings with it a host of other intellectual vices, too.

6

Introduction

Alternately, James’s will-to-believe doctrine is considerably more conciliatory with religious belief. Again, James has conceded that we face the question of belief in the face of an absence of sufficient evidence for the belief. What happens to religious belief when its standard intellectual defenses fail? And, again, James holds that they have failed. James is nevertheless committed to the proposition that religious belief may be responsibly held even after the proofs for it have come crashing down. But, again, only under very specific conditions that James lays out at the beginning of the essay (those of genuine options) and under very specific descriptions of what James calls the “religious hypothesis.” These descriptions of the religious hypothesis are what are of import, as the beliefs that survive scrutiny here and are held responsibly have been transformed in the process of willing-to-believe. It seems that as the traditional arguments for the traditional conception of God fall, so falls the traditional conception of God. Belief in God, in the wake of James’s defense, is reconstructed. The religious hypothesis is less a view about ultimate reality and God’s place in it and more a view about the place of hope in our lives. The lesson is that part of James’s defense of religious belief is to change it into something defensible in light of the lack of evidence for God’s existence. And so religious belief in James’s hands is no longer about God, Jesus, Heaven, Hell, angels, or miracles. It is simply that “the more eternal things are best” and that believing that is good. This is what of religious belief that the will-to-believe doctrine can salvage. The question for religious believers coming to James’s program as a source of support for their antecedent beliefs in God is whether James is really much of an ally in the defense of faith traditionally conceived. Is Jamesian reconstructed religious belief much better an option than Cliffordian nonbelief?

3. Five evaluative theses I’ve always admired philosophers that make clear what it is that they take to be at issue in a discussion, what they will argue, and where the twists and turns in the argument will be. Doing philosophy well depends largely on doing it clearly. In the service of this, I want to make clear the central theses of this commentary. There are five main evaluative theses I will be presenting: 1. Clifford’s case for evidentialism must be supplemented with an explanation of why false belief is bad and credulity is unacceptable. Clifford’s slope arguments perform this task better than they are regularly appreciated.

Introduction

7

2. Clifford’s defense of the Assumption of the Uniformity of Nature ultimately suffers from the fallacy of confusing contraries with contradictories. 3. James’s will-to-believe cases fail as counterexamples to Clifford’s evidentialist rule. Instead, they are better posed as cases for a belief-dependent notion of evidence, what I call doxastic efficacy. 4. James’s will-to-believe doctrine as a practical attitude depends not only on the plausible view that All belief is a guide to action, but also on its implausible converse, All action is guided by belief. This consistent confusion I will call the conversion fallacy. It is rampant in James’s argument. 5. James’s reconstructed account of religious belief fails to salvage recognizably religious belief. The takeaway is that Clifford’s argument needs some tweaking and that James’s argument is widely misconstrued as having refuted Clifford’s evidentialist requirement. Moreover, I believe that even were James’s critique of Clifford successful, the positive program for James is insufficient as a defense of the religious life he portrays it to be. In fact, I believe that religious believers who take Jamesian willing-to-believe as a means to defend their commitments have made a serious error assessing what the content of the commitments supported is. As such, I will be presenting Clifford as having come out better in the dialectic than James. This, by the way, is a very unpopular view of the exchange in the received story of philosophy. It is a much more widespread opinion that James got the better of Clifford. But in philosophy, as it is in many other areas of our lives, popularity is rarely a reliable guide to truth.

1

Reading William Kingdon Clifford’s “The Ethics of Belief ”

William Kingdon Clifford and the Metaphysical Society William Kingdon Clifford died of tuberculosis on March 3, 1879. He was 33 and left Lucy, his wife of three years, and two daughters. Before his illness he was a Fellow in the Royal Society, an honor reserved only for the most accomplished academics. He was Professor of Applied Mathematics at University College, London, and before that, he had been a fellow at Trinity College, Cambridge. Clifford was born on May 4, 1845, in Exeter, England, to a bookseller, and he went early to his undergraduate education at 15. First, to King’s College in London, but after excelling, his studies took him to Cambridge. Clifford translated works on non-Euclidean geometry into English, and he defended a view of geometry as an experimental and applied science in his essay, “The Aims and Instruments of Scientific Thought.” He showed, further, that three parallel lines define a ruled second-order surface, now named “Clifford Surfaces,” and parallel lines in “elliptic space” are also called “Clifford Parallels,” after him.1 Clifford was interested in the application and the accessibility of mathematics. He, early in his career, travelled on an eclipse expedition to the Mediterranean. The ship he was on struck rocks outside Catalina. It was lost, but all the passengers and scientific equipment were saved. The shipwreck was full of portent, as will be seen in Part I of “The Ethics of Belief.” Clifford was also an advocate of mathematical education, and his Common Sense of the Exact Sciences was written with an eye to explaining the state of the art in scientific and mathematical research to the layman. Scientific literacy, Clifford thought, was a key to social change. Connected to his activism in popularizing scientific findings was Clifford’s advocacy of religious freethinking. At Cambridge, Clifford was, at first, an avid

10

Evidentialism and the Will to Believe

reader of St. Thomas Aquinas and he was an enthusiastic supporter of Catholic doctrine. But then he read Charles Darwin’s On the Origin of Species, the regular thought that science was to be pursued within the confines of a religious perspective no longer seemed defensible. In fact, Clifford came to see much of Christian doctrine and the organizations of churches as positive impediments to the progress of scientific understanding and the improvement of society. In his firey “The Ethics of Religion,” Clifford concludes: “If men were no better than their religions, the world would be a hell, indeed” (L&E 218). Importantly, Clifford holds that most people are not as bad as their religions, because they have a conscience. This conscience, an inborn capacity to see the reasonable connection between facts, the rational response to challenges, is the fount not only of our capacity to understand the world, but also human morality. In short, science and morality share the same cognitive foundation. Clifford calls it a “natural ethics,” in “Right and Wrong: The Scientific Ground for the Distinction” (L&E 156), and in “The Ethics of Belief,” he terms it “the sacred tradition of humanity” (L&E 197). And so, according to Clifford, religion has no role except that of confusion and error in science and ethics. These views made Clifford a lightning rod for the religious concerns of the day. This was the time of what’s regularly called The Victorian Crisis of Faith. Darwin’s works and Thomas Huxley’s public avowals of agnosticism brought many of the intellectuals, religious leaders, and politicians of the day to ask whether England was on the wrong track. In essence, the question was whether the world was intelligible and whether people could be moral without God’s guidance. God, in both ethics and science, is the ultimate explainer. He made the universe and He made the rules for our behavior. In ethics, this concern is felt more acutely, as many worried (and still do today) that God is not only required to explain ethics but also provide motivation to be ethical. Without a God who punishes misdeeds in the life to come, it would be nothing but immorality down here, for sure. Fear of everlasting punishment keeps society together. And so the Victorian Crisis of Faith was not just an academic or intellectual concern. It was a discussion about what kind of society England would be. “The Ethics of Belief ” was Clifford’s signal contribution. “The Ethics of Belief ” was originally presented at the April 11, 1876, meeting of The Metaphysical Society. The society was a gathering of England’s top intellectuals from across various disciplines. Alfred Lord Tennyson, then the Poet Laureate, was a member. As were Arthur Stanley, Dean of Westminster; the current prime minister, William Gladstone; Arthur James Balfour, the later prime minister; the biologist (and “Darwin’s Bulldog”) T. H. Huxley; the Catholic

Reading Clifford’s “The Ethics of Belief ”

11

Archbishop of Westminster, Henry Edward Manning; the Anglican Archbishop of York, William Thomson; and other philosophers, scientists, lawyers, and mathematicians. Another member and prominent architect, Sir James Knowles, edited the popular Contemporary Review. He organized the Metaphysical Society meetings for the purpose of addressing the fact that disbelief seemed rampant among the intellectuals of the day. There needed to be a forum where the best cases for (and against) belief could be heard. He regularly published the papers presented and many of the discussions at the meetings in his journal. Clifford had a chance to revise his essay in light of the discussions, and Knowles published it in the January 1877 issue. The Metaphysical Society, leading up to Clifford’s 1976 presentation, had a lecture series specifically focused on miracles, and Clifford’s essay was an extension of that discussion. The core question regarding miracles related by testimony is: How trustworthy must the testimony be in order for anyone to reasonably believe solely on the basis of this testimony that the miracle occurred? Many other lectures in the series were critical of believing on the basis of miracle testimony. James Fitzjames Stephen, the famous jurist, opened the discussion of Miracles for the Society in 1874. He took on the perspective of a judge’s assessment of evidence for the event. Would testimony in the form of what we have for miracles be sufficient in a court of law? Stephen’s view was that without any other corroborating evidence, mere testimony is not sufficient for conviction. No jury would convict Mr. A for the murder of Miss. C merely on the basis of Mr. B’s testimony that he’d seen, say, A push C off a cliff. Even were B otherwise trustworthy, we know that we “cannot leave this weighty matter only to his say-so.” And so it goes for miracles.2 Along different lines, W. B. Carpenter approached the issue of testimony for miracles under the purview of psychology. Because all experience requires interpretation, Carpenter reasoned, those wanting to see or who are properly cued to see miracles will see them. There is always a “subjective” element to what we see, and as a consequence, mere testimony about what was witnessed is not much evidence without further investigation. It was not that those reporting the miracles lied or intentionally distorted anything, but that their experiences, being affected by their presuppositions, will not be reliable.3 T. H. Huxley then gave an address titled “The Evidence of the Miracle of the Resurrection,” which became widely notorious. Cardinal Henry Newman, who was not a member of the Metaphysical Society, heard about the paper, and he wrote to R. W. Church (then Dean of St. Paul’s), who planned on attending the meeting:

12

Evidentialism and the Will to Believe I hear that you and the Archbishop of York … are going to let Professor Huxley read in your presence an argument in refutation of Our Lord’s Resurrection. How can this possibly come under the scope of a Metaphysical Society? I thank my stars that, when asked to accept the honor of belonging, I declined. (Ward 1912: 333; cited by Brown 1947: 140)

Clifford’s essay came after these presentations and consequent discussions, and his view was unique in that the essay was not devoted to the analysis of some piece of evidence or other or over some miracle’s appeal. Rather, Clifford was focused on whether it matters at all that there is or is not evidence. That is, it seems a natural response to Stephen, Carpenter, and Huxley’s essays that the miracles reported are insufficiently supported by the standing evidence with the reply that the miracles reported are matters of faith, that we accept that they happened. They prove that God exists, that He is good, and He has had interaction with us. But they do so in a less direct fashion than simply by being established to have occurred in the way that any old historical event is established. Instead, they and the religious believer’s belief in them are part of a life of commitment to God, and they make sense only once one is in the right relation to God. To push on miracles as matters of evidence is, so goes the reply, to misunderstand what role they play in creation and in the life of faith. Clifford’s essay is a full-throated defense of the importance of our cognitive lives dictated by evidence, that this fideistic reply to the evidential challenges to claims of miracles and revelation is not only wrong, but also morally bad. Evidence matters, and it is not something to forego or deny lightly. Miracles are more than just unlikely events. They are supposed to be virtually impossible ones. Laws of nature must be broken for them to occur. The challenge with regard to miracles can be captured by the old Humean line: What is more likely, the miraculous event actually occurring or the person telling you about it is in error? Let’s take a very simple case to feel the weight of the Humean parity challenge. Imagine that you are told by your good friend that a Little League baseball team just beat the New York Yankees. In fact, they didn’t just beat the Yankees, they schooled them, smoked them, took them behind the woodshed. It wasn’t a publicity stunt. It wasn’t a stupid feel-good movie. It was a bunch of fifth- and sixth-graders throwing strikes against and hitting home runs off the best team in baseball. For real. No way, you’d say. And reasonably so. For this reason: it’s more likely that your friend is mistaken, was tricked, is lying, or is consistently misspeaking than that would happen. Seriously, the Baltimore Orioles are grown men and professional baseball players, and they can’t beat the Yankees, for Pete’s sake.

Reading Clifford’s “The Ethics of Belief ”

13

There are a few things to note about the Little Leaguers versus Yankees case to note. The first is that such an event isn’t technically a miracle. It is just something highly unlikely. Miracles are those sorts of things, but on steroids (we are talking baseball). Miracles are supposed to be things that not only just don’t usually happen or are highly unlikely things, but instead, they are things that don’t happen, can’t happen. Why? Because miracles are supposed to be proof of the divine, supernatural intervention in the world. If they are mere coincidences or improbabilities, then they nevertheless are natural events and not cases of supernatural entities inserting themselves into the ebb and flow of events down here. Miracles, then, likely require better testimony than people can give. The second thing to note about the Little Leaguers versus the Yankees case is that how one believes is determined by the evidence about the situation, not how one feels about one’s friend, Little League teams, or even the much-hated Yankees. So, despite the fact that you hold your friend dear, have a soft spot for the can-do attitude of 12-year-old baseball players, and hate those Yankees with every sinew of your body (which are all perfectly reasonable attitudes), that would not change the fact that the reasonable conclusion is: No Way. That it would be awesome if a Little League team blew out the Yankees does not make it any more so than the horribleness of a Yankee baseball dynasty makes it any less so. It is precisely here, again, that faith might be invoked. One should trust one’s friends, one should have a little faith in scrappy middle schoolers. Without this element of taking a risk, doing something with one’s heart every once in a while instead of just with one’s head, of feeling a deep pull to some thought or commitment, we wouldn’t be having this discussion about miracles. The question is: What would be so wrong about believing the Little Leaguers story? Where’s the harm in it? For sure, one can concede that the evidence doesn’t support the commitment, but because it would be nice to believe, what’s so wrong about that? And it is here that Clifford’s essay makes a case that such thoughts as these are not only in error, they are immoral.

Section I—The duty of inquiry 1. The ship owner case Clifford opens “The Ethics of Belief ” with a story of a ship owner about to send ship to sea. He has reason to doubt that the ship will make the voyage and bring

14

Evidentialism and the Will to Believe

its emigrant passengers to their new land safely. Even though it has survived many rough trips, it has come through. But it’s needed repairs afterwards. He focuses on the positives—the ship’s done just fine so many times, surely it won’t fail these people in need. Moreover, it’s untoward not to trust the ship’s builders or not to trust Providence. So he brings himself to believe the ship is seaworthy, and he happily watches her depart with those families on board. When the ship goes down in the middle of the deep sea, he tells no one of his earlier worries. Oh, and he collects the insurance money to cover his losses. The ship owner story seems simple enough, but it’s worth making explicit what all the ship owner’s considerations are. The story unfolds in six stages. They are as follows. 1: The ship owner (SO) has the following evidence (e): (e1): The ship is old; (e2): it was not particularly well-built; (e3): it has many miles on it; and (e4): it often needs repairs. 2: On the basis of 1 above, SO comes to doubt that the ship is seaworthy. There are two consequences: (a) SO considers having the ship overhauled, and (b) considering (a) makes him unhappy. Note that so far SO has been proceeding reasonably. SO has surveyed his evidence, seen where it takes him, and has a plan in the face of that evidence— since the boat’s likely not seaworthy, get it fixed. It’s reasonable, even, for SO to be troubled by the expense of fixing the ship, too. He’s a business man, and these expenses pile up quickly and take what would otherwise be money for him and his family. This is where things turn. 3: SO turns his attention to the following counter-considerations (cc’s): (cc1): The boat’s done well in other storms, it should make it back this time; (cc2): Providence will not fail to protect these families; (cc3): it is not appropriate to have doubts about the quality of work put in by those honest craftsmen who built the boat. 4: As a consequence of 3, SO comes to believe that the ship, without further repairs, is “thoroughly safe and seaworthy.” 5: As a consequence of 4, SO books the trip on the unrepaired ship and bids the exiles (and his hired crew!) good journey. Let us pause to consider what’s happened so far. Given 1–2, SO has evidence to believe the following conjunction: NOT-Seaworthy: The ship is not seaworthy and is in need of (costly) repairs.

Reading Clifford’s “The Ethics of Belief ”

15

In light of 3–5, SO comes to believe and act on: Seaworthy: The ship is seaworthy and is not in need of (costly) repairs. The first question to ask is whether e1–4 justify SO in believing NOT-Seaworthy. This seems reasonable, at least given the information, so far, especially in light of e4 in conjunction with e1–3. That is to say, each of the pieces of evidence individually support NOT-Seaworthy, but together, they support it more strongly than they would on aggregating their individual support. This is because if a boat is old, not well made, and is heavily used (e1–3), then in light of the fact that it has needed repairs in the past (e4), it will need repairs more often and more substantively in the future. Consider the case on analogy with an old car: Sam has an old car. It was a cheap-o, even when he bought it new. It’s needed many trips to the shop, and he’s been driving it pretty hard.

In light of this information, it’s well within reason to believe that the car is ready to fall apart. It may be in need of some minor repairs here and there, but it’s likely ready to have the motor fall out at the next pothole or have the axle break in the midst of a hard turn. Sam better not let his grandma borrow the car for a long drive through the Arizona desert without having checked it out. The point is that e1–4 look to provide excellent reason to believe NOTSeaworthy. The trouble for SO is that holding NOT-Seaworthy puts him on a pricey path, and one he does not like the idea of marching. So he at first considers what seems to be some relevant undercutting evidence; that the boat has weathered many storms and made many voyages (cc1). The question is: what’s the relation between cc1 and e1–4? In many ways, cc1 is just e3 restated—the ship’s gone through a lot. Now, in this case, SO then believes “…it (is) idle to suppose she would not come back safely from this trip, also” (L&E 163). The trouble is that this doesn’t follow, especially in light of the connection between e1–3 and e4, which is that even though the ship does come back safely, it also comes back needing repairs. And those repairs will likely increase in substantiality as the boat ages. It is most certainly not idle to have this worry. SO supplements with cc2 and cc3. Each plays a different role. The counterconsideration cc2 is an invocation of the divine will on behalf of the emigrants, whom Clifford even goes so far as to call “exiles,” who “seek for better times elsewhere.” We can safely presume that Clifford’s emigrants are members of a religious minority seeking to escape from persecution in their homeland. One of the later lessons from Clifford is the deleterious moral effects of (credulous)

16

Evidentialism and the Will to Believe

religious belief—particularly toward those who believe differently, and such a story would be consistent with the practice of gouging those seeking asylum. Regardless of the religious beliefs of the passengers, it is the religious beliefs of SO that matters. SO believes in Providence and that it will offer protection of these people. The first thing to note is that cc2 is an implicit acknowledgment that cc1 hasn’t really done its job. Why must Providence provide “protection” unless what was happening was not thoroughly safe? SO must invoke cc2 precisely because cc1 is an exceedingly weak rebuttal to the support e1–4 provide for NOT-Seaworthy. SO needs a new evidence set, one not appealing to the facts about the boat at all, but about the emigrants and God’s plan. The trouble with Providence is that we find out what God’s plan is afterwards, not beforehand. Invoking Providence is a remarkably unreliable way of predicting the future. Too many coaches of good athletic teams have insisted that they get to call themselves a “Team of Destiny” only after they win the championship. Providence, as often as it provides us the boons of fortune, provides us misery and failure, too. SO then turns to cc3, which, again, is a wider consideration than what is under scrutiny when we look only at the boat. This counter-consideration seems to be a kind of principle of etiquette, roughly: One is ungenerous when one suspects that the work of contractors or builders is insufficient. One ought be generous, or at least not be ungenerous. So, one should not suspect their work to be insufficient.

Now, for sure, one ought to be generous. But is it ungenerous to double-check the work and to take precautions against the failures of those who would build and repair the ship? There is a reason why homes sold in the United States have independent inspectors to look over the house to check its foundations, the electrical wiring, and any work that had been done on it since it had been built. We do so precisely because we want to take the owner’s word for it, but we know how they can, ahem, forget about the infestations or recent trouble with the wires. Similarly, we buy our boats, houses, and cars from people we trust, but it, minimally, is not a breach of generosity to kick the tires, get a second opinion, or take a hard look. Here, cc3 is an invocation of another ethics of belief, another model for evaluating beliefs according to an ethical standard. The thought behind cc3 is that beliefs are evaluable in terms of their content, that beliefs, because of what they are of are good or bad. We express a kind of moral character when we are suspicious of those who work for us, of those in our service. Consider

Reading Clifford’s “The Ethics of Belief ”

17

the suspicious business owner who constantly is convinced that her employees are stealing from her, or the unhappy customer who is always concerned that at every meal, there’s some way that he is being squeezed for money. “That’s how they get you!” they say. And they exhibit an unhealthy attitude toward others. Here’s how SO would not be like the suspicious business owner or the unhappy customer: SO has evidence about his boat and its track record. And since it’s old, he has reason to expect that things will be getting worse, not better. The suspicious business owner and the unhappy customer don’t have evidence about the specific cases—they have a general suspiciousness to them. That’s ungenerous. It’s not ungenerous, though, to believe the boat will need repairs given e1–4, but it would not only be ungenerous but also unreasonable to believe NOT-Seaworthy as a default. Now, again, cc3 is an ethics of belief—one’s beliefs must be managed in a way that one may express some interpersonal virtue. In this case, it is generosity, but the thought may be broadened—perhaps to goodwill, and so on. A version of this view will be developed as a case for James’s will-to-believe doctrine later (see Sections IX and X of the James essay). For now, let us concede that the reasoning in cc3 is not evidential, and thereby it is not a reason to believe NOT-Seaworthy is false and Seaworthy is true. So far, we have progressed through five of the six stages of the ship owner case. What we’ve seen is that SO in 1–2 reasonably considers the proposition NOTSeaworthy, and the financial costs consequently upset him. So he embarks on a project of rationalizing in 3–4, which yields his “overcoming these melancholy reflections,” and he resultantly believes Seaworthy. SO’s belief, then, yields his action of allowing the ship to sail without repair in 5. What comes next is: 6: (a) The ship goes down mid-ocean, and (b) SO collects his insurance money and tells no tales. Now, Stage 6 has two parts, and what’s relevant for the ship owner case is, really, only 6a. 6b is an important rhetorical flourish on Clifford’s part, but not directly relevant to the broader case for evidentialism. When the ship goes down, we see the direct result of SO’s methods of deliberating about the seaworthiness of his ship. The families (and crew) die. That’s bad, and a direct result of SO’s false belief that Seaworthy. But what of 6b? Again, I’m inclined to think that it is unnecessary for Clifford’s main argument, but it serves the purpose of engendering our moral outrage. That is, Clifford follows the tale of the ship owner with the question: “What shall we say of him?” He’s set things up conveniently for the kind of answer he’s

18

Evidentialism and the Will to Believe

looking for—that SO is blameworthy. But part of our moral reaction is posited on that last line, that SO collects the insurance money and tells no one of this earlier doubts. First, why would SO do that, then second, why is that a moral issue? Why would SO perform 6b? Well, because in this case, the insurance money and his doubts that make him unhappy are all financial matters. One reason it is appropriate to react strongly to this version of the ship owner case is that it seems SO has formed his beliefs entirely on an economic basis. It is more costly to fix the ship than to send it out without repairs. If the ship goes down, he’s lost the ship, but the insurance money allows him to buy a new one. No great loss, really. Clifford’s inclusion of 6b in the ship owner story, then, is a rhetorical addition to the case, one that ensures we hold that SO has done something contemptible. What’s important is that with 6b, Clifford has provided us with a stand-in for SO’s moral reasoning bearing on how much scrutiny he gives to deliberations about the ship. Because he does not take himself to have a very demanding moral reason to scrutinize his conclusions, and because he fails to take into consideration the losses others would incur if he is wrong, we hold him morally awful. But Clifford will later bring his audience’s considerations of what the morally relevant errors would be, and he will have to supplement his account. Clifford, when he turns to evaluate SO, holds that “he was verily guilty of the death of those men” (L&E 164). We might call such guilt criminal negligence. SO had every reason to believe NOT-Seaworthy, but instead worked himself into believing Seaworthy, and then he sent those people to their watery graves. Now, the fact that SO sincerely believed the ship was seaworthy is not any help. The issue is not whether he lied to the emigrants. Rather, the issue is whether he exercised the proper care in forming his belief: … the sincerity of his conviction can no wise help him, because he had no right to believe on such evidence as was before him. He had acquired his belief not by honestly earning it in patient investigation, but by stifling his doubts. (L&E: 164)

What allows us to hold SO responsible, then, is that he should have known better. Consider other important cases of intellectual negligence that cause harm. We hear of them regularly enough, especially when we watch the local news. A young mother gives her sick baby a dose of bleach to cure a cold. A man leaves his schnauzer in his parked car for hours in a hot parking lot. A father who, when it begins to rain in the midst of the back yard barbecue, brings the grill inside. They, simply, should know better. And so when the baby is poisoned, the

Reading Clifford’s “The Ethics of Belief ”

19

dog baked, and the family asphyxiated, these terrible results are on their hands. Ignorance in these cases is culpable because the evidence of the relevant facts is overwhelming and easy to assess. And so these errors are morally blameworthy because of the fact that others suffer for one’s sloppy thinking. Clifford then asks us to consider a parallel case. Imagine a doppelganger for SO, SO*. SO* proceeds in deliberating about his ship in the same way SO does. And so SO* goes through stages 1–5 just as SO did. Consequently, SO* has believed contrary to the standing evidence, having rationalized his way to believe Seaworthy. But now, as Clifford says, “alter the case a little,” and allow, instead of 6, the crucially different outcome: 6*:The ship completes the voyage and many more after that. SO* is not in error, at least in the sense that he was right that Seaworthy was true. Crucially, because SO* believes contrary to his evidence, he is lucky that the bad consequences did not befall him, his crew, or the emigrants. It is a regular thought that luck is not a morally relevant category. Consider the drunk driver case. Driver A gets blotto drunk at the bar and drives home. On the way, A briefly falls asleep at the wheel and runs into a pedestrian. Driver B gets blotto at the bar and drives home. B also briefly falls asleep at the wheel, but B doesn’t hit anybody. B got lucky, compared to A; A was unlucky, compared to B.

On the basis of what do we evaluate their moral difference? For sure, A’s legal punishment is going to be much greater than B’s, but this is because there are no direct victims of B’s moral wrong. What A and B did, regardless of their consequences, was wrong. They needlessly endangered other people. A’s endangerment yielded actual harm; B’s didn’t. But, again, that’s a difference of luck—surely we wouldn’t say: B was lucky, so less morally bad. Nor would we say: A was unlucky, so morally worse. Again, luck does not seem to be properly morally relevant—what’s relevant is the fact, again, that A and B endangered others. This line of thought extends, now, to the ship owners. SO is unlucky and SO* is lucky. But they both, as Brian Zamulinski notes (2002: 439), endanger others in holding the belief Seaworthy. Clifford reasons: Will this [the fact that the boat makes the journey] diminish the guilt of her owner? Not one jot. (L&E 164)

The form of the reasoning leading to the belief, the attention given to it, and the potentially morally relevant consequences are what determines the moral

20

Evidentialism and the Will to Believe

evaluation of the ship owners. And Clifford explicitly denies that the luck element distinguishing SO and SO* is morally relevant: When an action is once done, it is right or wrong forever. No accidential feature of good or evil fruits can possibly alter that. (L&E 164)

This is to say, because SO and SO* have 1–5 in common, and 1–5 are the only things in their respective stores that are exclusively up to them, their difference in terms of the outcome, lucky 6* or unlucky 6, is not morally relevant. It is not that SO was factually wrong that makes his belief morally wrong, but that SO formed the belief in a way that put others at risk. That is what the argument from the second ship owner, SO*, shows. Clifford concludes from this comparison, then, that the moral evaluation of beliefs is not a matter of their content, but rather of the methods one uses in getting and maintaining them. Contrast this with the briefly considered ethic behind cc3, that SO should think well of his fellows. There, it is the content of the belief being evaluated—you’re obliged to believe certain propositions because of their moral value and how they express your sociable virtues. Clifford’s ethic of belief is markedly different. It is not a matter of evaluating the belief in terms of being true or false (as with 6*) or that it expresses the right moral sentiment or not (as with cc3), but, again, with whether it was a belief he should have known better to hold or not, given his evidence. Clifford makes these contrasts: The question of right or wrong has to do with the origin of the belief, not the matter of it; not how it was, but how he got it; not whether it turned out true or false, but whether he had a right to believe on such evidence as was before him. (L&E 165)

The upshot is that because both SO and SO* believe contrary to their evidence on the basis of rationalized counter-considerations, and they thereby put people at risk, they are both wrong and to the same degree. Before we move to Clifford’s second case, we must pause to consider how the ship owner case relies on a significant moral principle, one that takes the matter of consequent luck to be morally irrelevant.4 In short, you can’t be held responsible for things not your fault, nor can things not up to you improve your moral esteem. This principle is an old stoic view, namely that the moral value of an action or one’s deeds is exclusively a matter of what is wholly under one’s control. Take, for example, Epictetus’s criterion for moral relevance in Encheiridion 1: Some things are up to us and some are not up to us. Our opinions are up to us, and our impulses, desires and aversions—in short, whatever is our own doing.

Reading Clifford’s “The Ethics of Belief ”

21

Our bodies are not up to us, nor are our possessions, our reputations, our public offices, or that is, whatever is not our own doing.

I will argue later (in Section I-6) that Clifford’s moral philosophy is a form of neo-stoic ethics, but as it stands, the parallel with the stoics is only for the sake of clarifying Clifford’s view that moral evaluation is restricted to what is under the agent’s control. One question, though, is whether Clifford is right that whether given that SO and SO* are wrong or right is a matter of luck, the truth or falsity of their views is morally irrelevant. A perfectly reasonable challenge to Clifford can be: sometimes being right or wrong is what matters, not how one arrived at the view. I’d presented Clifford’s view with the slogan that “luck is not a moral category,” but despite the fact that the principle sounds right, there are cases that test its correctness. Again, to return to the drunk driving case earlier, we think of the crimes we would punish the drivers for as different. A is guilty of vehicular manslaughter, while B is guilty of driving while impaired. Or consider the difference between two hit men, Lucky Luke and Bad Luck Barry. Luke takes the shot at his target and kills him. Barry takes his shot just as his target stoops to tie his shoelaces. Were the two tried and found guilty, again, there would be two different crimes on the docket: premeditated murder for Luke, but only attempted murder for Barry. Again, the moral wrong they intended to do was the same. But the consequences are different by luck, and we treat them differently under the law. That’s a problem for the anti-luck view, isn’t it? For sure, Clifford’s background principle has both its appeal and its drawbacks. There are two options for the Cliffordian here. On the one hand, the Cliffordian can maintain the view on the moral irrelevance of luck by noting that there is a difference between moral luck and the legal evaluation of the event. We punish some more severely for the sake of the deterrent elements. It’s clearer to the onlooking public that the bad action with bad consequences was bad, but less so with the one with good (or less bad) consequences.5 So for the sake of clarity, we satisfice our moral censure with the civic role of punishment. So even though luck is irrelevant to moral blame, it can be relevant with legal censure. On the other hand, the Cliffordian can concede a minimal element of luck here, one that tracks a further natural attitude toward luck, namely that we feel the wrong more intensely when the resultant consequences of the wrong are before us. It is allowable that we treat the two morally similar cases differently, so long as they are subject to the same kind of moral objection—drunk drivers unjustifiably endanger others, the hit men try to kill others, the ship owner endangers his passengers. That is the baseline for our moral censure, and the

22

Evidentialism and the Will to Believe

differences in punishment depends on what needs to be done to set things right. That is, punishment in some cases must not only be proportionate to the wrong done, but also to the harms done that must be restored. So the restorative function of punishment may be invoked in cases where the moral wrong may be equivalent but punishment different, as one case may need more restorative work than another. In the lucky cases, there’s no harm done, so nothing to be restored; in the unlucky cases, there’s harm done, so something needs to be set right. I will proceed as though one (or both, as they are consistent) of these two lines is right, but I, myself, prefer the first. In essence, the Cliffordian view so far is a form of anti-luck moral assessment yielding a form of anti-luck epistemic assessment.

2. The island case Clifford’s second case shares a number of similarities with the first, but there are important differences. It is worthwhile laying it out formally in order to demonstrate both and to draw out the consequences. Clifford asks us to imagine an island with a religious minority on it, one that denies the central Christian doctrines of original sin and eternal judgment. People on the island, perhaps sharing the same attitude many hold toward those who do not ascribe to the Doctrine of Hell, hold those of this religious minority in suspicion. The assumption is: if one does not believe in Hell, then why should one have motive to be moral? In light of this, the other islanders suspect the religious minority has been kidnapping children to convert them. Some men are so bothered by this thought, they begin to publish these accusations so as to agitate the wider populace. A commission is formed to investigate the matter, and the commission finds that not only the accusations are baseless, but also that the exonerating evidence was easily available. The agitators are then held in universal moral contempt as those who are “no longer to be counted as honorable men” (L&E 166). Similar to the ship owner case discussed earlier, the story seems simple enough, but laying out its stages will help discussion. In all, I see four stages to start. They unfold as follows: 1: There is a suspicion on the island that members of the religious minority (a) manipulate the laws to teach their views to children, and (b) kidnap children to indoctrinate them.

Reading Clifford’s “The Ethics of Belief ”

23

2: A group of agitators (GAs) form a society to raise awareness of 1a and 1b, and they widely publish these accusations. 3: As a consequence of 2, a commission is formed and finds: insufficient evidence for 1a or 1b, and further, evidence of innocence, which was easily found. 4: After 3, all the GAs are held in disrepute. First, a quick note that the island case is entirely stipulative in terms of the evidence and considerations bearing on the conclusions. In the ship owner case, we have details about the ship, what goes through SO’s mind, and so on. In the island case, there is no presentation of the evidence, only reportage of whether there is any and whether it is sufficient and easily found. This is not a complaint yet, but it is worth noting, and I believe it speaks to the broader objective of Clifford’s essay, which is not to present a first-order theory of evidence assessment, but a second-order account of why caring for sufficient evidence is important. The consequence is that the details of the evidence needn’t be of primary concern for the present argument, but of what to do in the face of having it or not. So let’s first look at the transition from 1 to 2. In essence, the move is one from hearsay and prejudice to belief and full-throated advocacy. Likely in the background here is an ugly tendency for folks not only to outgroup others, but also to believe the worst about those in the outgroup. In fact, once one group has progressed to a point of distance from another group, it is all too easy to not only suspect something bad of them, but also to positively expect and anticipate it.6 The consequence is that the GAs are religious bigots with control of the presses—they not only have prejudiced beliefs, but they also promulgate them widely. The transition from 2 to 3 is intuitive, too. If there is a public uproar, officials will be called upon to do something. And in this case, there is an investigation. But the investigation is (lucky for the religious minority) not stacked with the bigots from GA. The commission finds, again, stipulated by Clifford, the following: Finding 1: There is insufficient evidence for the accusations. Finding 2: There is sufficient evidence of innocence, and it was easy to obtain. It seems right that those who are aware of these findings no longer see the GAs as honorable men. The fact that they genuinely believed their accusations makes them less liars, but nevertheless wrong in their accusation. Not just making a false accusation, but for making a baseless one. Clifford frames it:

24

Evidentialism and the Will to Believe For although they had sincerely and conscientiously believed the charges they made, yet they had not right to believe on such evidence as was before them. Their sincere convictions, instead of being honestly earned by patient inquiries, were stolen by listening to the voice of prejudice and passion. (L&E 166)

The problem is that the GAs believed and acted on their preconceptions and biases, and this, again, needlessly endangers members of the religious minority. As with the ship owner case, Clifford proposes a different outcome to highlight what’s truly wrong with the GAs. He proposes a revision, and so we change the case so that a further “still more accurate investigation” uncovers more evidence. But this time, the investigation uncovers evidence of the accuracy of the charges. The GAs might say “Now you see we were right after all; next time perhaps you will believe us.” And so, the second island case is an extension of the first, one with two further stages: 5*: A more accurate follow-up investigation finds more evidence. This time the accused members of the religious group really are guilty. 6*: Members of the GAs revel in being right and bristle at having been doubted in the first place. Does being right make the accusations from the GAs acceptable? Surely in their own eyes, they’ve been vindicated. They had a hunch, made an accusation, and the evidence they just knew was out there all along finally came to light. They just had to keep from doubting themselves, they say to each other in selfcongratulatory tones. But are they really vindicated? Clifford’s answer is that they are not; and, again, his reasons are based on a view with regard to what the appropriate moral categories are, and he is consistently out to deny the moral relevance of luck in these matters: …the question is not whether their belief is true or false, but whether they entertained it on the wrong grounds. (L&E 166)

What was up to the group of agitators was whether they harbored (and promulgated) beliefs “they had not right to have.” That is, what makes the GAs morally similar, both in stage 3 and stage 5* (the two versions of the island case), is that they believed on insufficient evidence. The only difference is that in one they happen to be found out as having done so, and in the other they are vindicated as right (and so not revealed to form their beliefs merely on prejudice and supposition).

Reading Clifford’s “The Ethics of Belief ”

25

It’s worthwhile to pause and note the similarities between Clifford’s two cases and take note of a few dissimilarities. The first and most important similarity, again, is Clifford’s anti-luck ethics, which drives him to a form of anti-luck epistemological evaluation. Agents are not morally responsible for what they have no control over, and so the ship owners and the members of the agitator groups get no credit for being right when they’ve formed their beliefs in morally irresponsible ways. Clifford’s assumption, of course, is that how subjects form their beliefs is largely up to them and whether these beliefs come out true or false is not. The agent is morally and/or epistemologically responsible for the methods of management—the ways the agents react, motivate themselves to attend to evidence, direct their attention, and so on. What they are not responsible for is whether their beliefs and plans come to fruition. We take care of our beliefs, and truth takes care of itself. Note here that this dichotomy will be challenged by James’s doctrine of the willto-believe with what I will call them doxastically dependent facts—truths substantially dependent on whether a subject believes them. (I will discuss these in Chapter 2, Section IX-5.) Now, for the differences between the two cases and their significance—there are two important differences. First is that the ship owner (SO) believes contrary to his evidence; but the agitators (GAs) believe on insufficient evidence. Second is that SO had most of the relevant evidence before him, which warranted him in at least contracting for inspections and perfunctory repairs. The GAs did not have any of the relevant information before them, but it (given 3) was easily obtained. The consequence is that even though the broader vice of epistemic negligence is true of both SO and the GAs, they have significantly different forms of negligence. Let us call the negligence of SO, who rationalizes so as to believe against what his evidence supports, epistemic insolence. Those exhibiting this tendency will believe what they will, even in the face of every reason they are wrong. They have a tenacity of mind that can block out and even pervert (as with e3 becoming cc1 for SO) counterevidence into evidence for their views. Alternately, let us call the negligence of the AG’s epistemic sloth. The GAs are happy with their views and once they are convinced they are right, they are satisfied in a way that those who believe they are right always are—they smugly look at those who disagree as benighted or fooled in some way; they take themselves to have an insight that even if hard to prove, is nevertheless right. And so the simplest tasks of inquiry are ones that never occur to them to undertake or seem worthwhile for them to consider.

26

Evidentialism and the Will to Believe

Given the two different kinds of negligence, we see two norms of assessment arise. The first, regarding standing evidence, which seems to run: If a subject (S) has evidence at time t that supports not-p, then if S were to form a belief at t with regard to the question whether p, S does not have the right at t to believe p.

I’ve intentionally put this norm in a relatively weak form, as it doesn’t yet seem clear that Clifford holds that S must believe that p or that S has a right to believe that p (though later, Clifford will argue so). The point is that, so far, S does not have the right to believe against the evidence that not-p. The second norm for evaluation, the one behind the island case, seems to be two lines: one synchronic, like with SO; the other diachronic. First, the synchronic norm. Given that the GAs believed on the basis of insufficient evidence (or, no evidence at all, but prejudice), and Clifford holds that they had not “right” to do so, it seems he is committed to: If S has insufficient evidence at t to support p, then if S were to form a belief at t with regard to p, then S would not have the right to believe that p at t.

Now, with the island case, we have the GAs believing that members of the religious minority are kidnappers and legal weasels, but they do so only on the basis of their nonevidential prejudices. They have no right, given the risks to others incurred, to do that. Now consider the problem that because the GAs formed their beliefs on the basis of insufficient evidence, they became slothful with regard to searching out further evidence bearing on the matter. Let me try to get the background norm out in two ways. First, consider the judgment of epistemic sloth (and hence, no right to belief), given how easy the counterevidence was to be found (Finding #2, from Stage 3 of the island case): If S believes that p at t, but S could have very easily uncovered evidence that not-p, then S has no right to hold the belief that p at t.

Epistemic sloths don’t have a right. Now, I think the norm can be diachronically stated, as it really is an evaluation of performance: S has a right to believe that p at t only if S has investigated the matter whether p thoroughly enough prior to t so as to ensure either (a) there is no easily discoverable evidence either that not-p or that p is not supported by the present evidence, or (b) that if there is such evidence as (a), there is sufficient evidence to rebut that evidence.

The anti-sloth principle here is simply that any responsible epistemic agent has checked out the usual suspects and done due diligence. No easy cases are

Reading Clifford’s “The Ethics of Belief ”

27

left unchecked and if the easy cases have gone against the view, the responsible inquirer has either rebutted that evidence or has revised the target belief. That’s what happens with the second version of the island case, clearly, as the easy rebutting evidence is itself rebutted; and all of this evidence collection is performed by the right kind of believer, “the more accurate investigat(ors),” as opposed to the GAs. Clifford even attributes a moment of reflection to the GAs, a case where they are before their own conscience, and they must admit (even if but to themselves): Every one of them, if he chose to examine himself in foro conscientae, would know that he had acquired and nourished a belief, when he had no right to believe on such evidence as was before him; and therein he would know that he had done a wrong thing. (L&E: 167)

Clifford holds, then, that even the epistemically slothful and insolent, if given a moment to reflect on it, will concede that they are in the wrong.

3. Beliefs and actions Clifford, before continuing with this argument, specifically to generalize from his cases, pauses to consider an objection: In both the ship owner and island cases, we are speaking as though we are evaluating beliefs, but it would be more appropriated to be evaluating actions. The ship owner failed to have the ship examined and fixed. The island agitators made unsubstantial accusations. It, so goes the objection, is in the action that we have the harm of endangerment of others, not in the believing. To this objection, Clifford has two replies. The first is to concede that it was in the failure of actions that we can see the wrongness of the cases, specifically to follow up on evidence. The second is to deny the dichotomy of belief and action the objection requires. I will take the two answers in turn and show that they yield a form of cognitive pragmatism that can explain why beliefs (and the means of their formation) are morally evaluable. Clifford’s first move is to concede that “so far as it goes, this view of the case (distinguishing beliefs and actions) is right and necessary” (L&E 167). However, the ship owner (SO), even once he has the belief that Seaworthy, still has an obligation (given his evidence) to inquire. The same goes for the island’s GAs— even though they sincerely believe the members of the religious minority are guilty, they are epistemically slothful if they don’t do their due diligence. And so, each case has the subject in question still “with a choice in regard to the

28

Evidentialism and the Will to Believe

action suggested by (his belief), and so cannot escape the duty of investigation on the ground of the strength of his beliefs” (L&E 167). This is to say: given the evidential situations, both SO and the GAs had duties to perform certain actions, those of inquiry. They did not, and so they are blameworthy. Now, so far, we can see Clifford’s first concessive move is to place the emphasis on the norm of prohibiting what I called earlier epistemic sloth—failing to inquire where and when one ought to. This is actional and diachronic. And note Clifford’s concession—moral evaluation should be on the action. This is where the move to the second reply is important, as Clifford then argues: No man holding a strong belief on one side of a question… can investigate it with such fairness and completeness as if he were really in doubt and unbiased. (L&E 168)

Beliefs, then, are objects of moral evaluation, too, because they impel us to perform our actional duties of inquiry or inhibit them. Clifford’s argument then turns to the consequences of holding a positive belief, and as a consequence, foregoing inquiry: Not inquiring in cases where and when it is required is morally wrong. Prior belief not based on fair inquiry inhibits further and fair inquiry. So prior beliefs not based on fair inquiry “unfits a man for the performance of his necessary duty.” (L&E 168)

Beliefs not based on sufficient evidence or formed without a proper attempt at inquiry prevent further fair inquiry. So even if we are evaluating actions, and they are the primary locus of moral censure, beliefs must also be morally evaluated, because they may impel or impede further duties of finding more evidence. Now, look at this premise, that belief impedes further inquiry. Is this true? Clifford offers a case, based on a view of belief that can be easily classed as pragmatic. For it is not possible so to sever belief from the action it suggests as to condemn the one without condemning the other. (L&E 168)

Beliefs are guides to actions, they direct our lives. And so they, as direction for action, are evaluable by the same lines our actions are. They are tied. Nor is it truly belief at all which has not some influence upon the actions of him who holds it. (L&E 168)

In fact, Clifford holds that to hold a belief “truly” is to look upon the action in a way that he terms as “lust(ing) after it,” to commit the act “in his heart.”

Reading Clifford’s “The Ethics of Belief ”

29

I’ve called Clifford’s view a proto-pragmatist commitment, as there was, at the time of Clifford composing “The Ethics of Belief ”, no substantive pragmatist tradition yet to speak of. C. S. Peirce’s seminal essays on Pragmatism, “The Fixation of Belief ” and “How to Make our Ideas Clear,” were published in Popular Science Monthly in 1877 and 1878, respectively. These were basically out in the same time frame as “The Ethics of Belief.” It is possible that Peirce and Clifford met and talked, and Clifford most certainly was familiar with Peirce’s work on logic, as he apparently favorably compared Peirce’s contributions to logic to Aristotle’s. And he was in regular contact with Peirce on matters regarding stellar photometry (Fisch 1986: 126). Whether there is an uncharted Peircan influence here or not (or, for that matter, a Cliffordian influence on Peirce), Clifford’s view of belief is of the family of pragmatism (one needn’t be inspired by a pragmatist to be one … else the first wouldn’t have been one). It is important to note that the argument of “The Ethics of Belief ” is that certain patterns of belief formation and maintenance are ethically objectionable, and the best case to be made for that view is that this is so because the beliefs in question are prime contributors to one failing to perform some morally obligatory action, or alternately, performing some morally objectionable one. Thus a pragmatic case for evidentialism. This point, that Clifford’s epistemic norms for why one should care about evidence depend on ethical norms, is taken as a reason to consider Clifford’s view a form of epistemic pragmatism. As Veli Mitova puts it, “the sacredness of knowledge … merely piggybacks on the sacredness” of doing one’s practical duty for humankind (2008: 474). Clifford’s norms are all posited on the practical consequences they either avert (in our cases, not endangering others) or encourage (in our cases, fulfilling our duty to humankind in maintaining our correctness of mind). For Clifford’s case, it is necessary to keep a levels distinction in place, which is that between asking the first-order question of what evidence should determine what I should believe and the second-order question of why I should care about the evidence in directing my beliefs. Clifford, again, proposes no answer to the first question, as he has no special views in the essay about evidence to propose (but will have a view on constraints on evaluating testimony later in Section II). But his answer to the question of why care about evidence is that because beliefs formed in an inappropriate evidential fashion (either slothfully or insolently, as we’ve seen before) are cases of morally objectionable negligence. So we have moral reasons to care about epistemic norms. It is to this line of reasoning that a tu quoque is leveled against Clifford. Remember that Clifford had earlier held that moral reasons for a belief (say,

30

Evidentialism and the Will to Believe

that it contributes to us being charitable with others—cc3 in the ship owner case Stage 3) are not appropriate. But the structure of Clifford’s argument here is similar, and so risks being invalidated by Clifford’s own rule. Mitova states the challenge as follows: A Clifford-style pragmatist justification of epistemic norms … cannot support, but undermines evidentialism. (2008: 487)7

This is a serious challenge. But I believe it can be answered. First, it’s possible to concede that Clifford’s argument is from ethical to epistemic norms without stating what the natures of the ethical norms are. For now, let’s simply acknowledge that the ground floor of Clifford’s ethics of belief is ethics, and the endorsement of the epistemic norms depends on them. I will address the nature of those ethical norms later (Chapter 1, Section II-5). However the tu quoque will be answered shouldn’t depend on any specific reading of Clifford’s metaethics. Second, two different forms of rule endorsement must be distinguished. That is, there is a difference between the question: Is X or Y the better rule for evidence?, and the question: Why follow the best evidential rules? The criteria for the first needn’t reference the second, either. So evidential rules needn’t be reducible to moral rules, but they, given Clifford’s case, are binding because we have moral rules to follow them. A more direct answer is in order. If the question is Why be bound by evidential rules?, then an answer must also provide reasons answering a normative question. Other normative considerations bearing on the truth or falsity of an answer are pieces of evidence. So, if someone asked if it is wrong to, say, dangle a baby out a 10th story window, I might say: it’s wrong because it endangers the child’s life for no reason. When I’ve given this reason, it is a reason that established the truth of the claim that it’s wrong to dangle babies from high-up windows. This is certainly a different evidence from the evidence discussed in Clifford’s essay (specifically that of empirical evidence), but insofar as there are moral truths and reasons that bear on them, those kinds of reasons count as evidence. What’s necessary, at this stage, is to acknowledge that Clifford’s view depends on a substantive metaethics. I will once again address this point about Clifford’s ethical views in more detail once I have made it through his full line of ethical reasoning later in Section I. For now, if we can see that insofar as there are normative truths and reasoning that bears on them, let us take those reasons as the kind of evidence that holds off the tu quoque version of the self-refutation problem for Clifford.

Reading Clifford’s “The Ethics of Belief ”

31

Before we move to Clifford’s general case, we must pause to consider a related objection, but one more widely held. The objection runs that Clifford wrongly treats belief with the same scrutiny as and with the same rules as action. Actions are up to us, but beliefs are not fully so. We usually just find ourselves with beliefs. We don’t decide what to believe, but usually just end up with beliefs. In a sense, consider this argument to be an extension of Clifford’s own anti-luck moral view: we should be held morally accountable only for what is under our control. But we do not have much control over our beliefs. They do not respond to our will in the way, say, that raising a hand will. I cannot change my beliefs in the way I might change my posture or change my expression. Call this the voluntarism problem for Clifford’s ethics of belief. I believe the voluntarism problem is posited on an overstatement. We do have a measure of indirect control over our beliefs, a kind of control we see in clear display with the SO case earlier.8 SO finds himself, given his evidence, considering costly repairs to the ship, but then refocuses his attention to other counter-considerations that psychologically overturn his previous inclination to believe. He does so not for the truth but for the convenience of this belief. So he exercised his control over his beliefs, not by changing them directly, but by exercising his control over where his attention was directed. He, you might say, forced himself to look toward the kinds of considerations that, regardless of their connection to the truth of the matter, would eventually direct his belief where he prefers it. Pascal, at the close of this famous wager (which I will discuss in more detail in reply to James’s presentation in Chapter 2, Section III-3) concedes that he cannot generate the genuine article of belief just with the wager. Even if it is a lock-certain good bet, that does not yield belief. What does Pascal do? He must put himself into a position where his belief will come naturally as a product of his environment—so he goes to church, thinks often and longingly of the Lord, stops talking to atheists. Again, he does not control the belief directly, but he controls the factors he knows to be conducive of belief or disbelief. So we, when we evaluate our beliefs, are evaluating how we came to them, how we maintained them. Call this view indirect doxastic voluntarism, which concedes that the analogy between action and belief isn’t perfect, as we do not have direct control over our beliefs. But because we do have indirect control over them (in what actions we take arriving at and in maintaining our beliefs), we can evaluate them as things up to us, and so they are matters of moral consideration.

32

Evidentialism and the Will to Believe

4. Beliefs and their consequences Clifford’s cases of the ship owner and the island agitators are very specific kinds of examples. How do they show a generalized, universal ethics of belief? That is, they, for sure, show that evidence is important to beliefs when in those kind of circumstances (namely, deliberating about fixing a boat before a long voyage and investigating whether members of a religious minority are kidnappers). For sure, one can concede that in those cases, attending to evidence is morally obligatory. But it doesn’t mean that it’s obligatory for all cases. We can develop this objection. I will call it the generalization problem. The objection proceeds from the following observations of what the ship owner and island cases share: they are cases wherein, because the error costs for falsepositive beliefs are so high, the burden of proof is very demanding for positive beliefs. But not all cases are like that, and so the rule we see in them does not generalize. Let me explain. Consider the ship owner, again. SO is asking whether Seaworthy. The boat is or is not, and SO can believe it is or not. This yields four possibilities: (i) SO believes Seaworthy, and is right; (ii) SO believes Seaworthy and is wrong; (iii) SO does not believe Seaworthy and is right; and (iv) SO does not believe Seaworthy and is wrong. Those are the four possibilities, and here is the practical reasoning about the evidential standards for the case: SO had better be really sure that he’s right when he believes Seaworthy, since if he’s wrong (case ii), he’s got a sunken ship, lost crew, and dead families. Though not relevant really to the moral estimation of the situation but certainly relevant to the man, SO does have insurance so at least he’s defrayed the costs of the ship. That makes him out to be all the more morally horrible, since that’s the only real precaution he’d taken. Consequently, when SO believes Seaworthy contrary to his evidence, SO has believed wrongly because positives under these circumstances should have very high standards for evidence. High error costs drive up scrutiny. The same goes for the island case, and this is a familiar commitment, which is that if you’re to hold someone guilty of a crime, you need excellent evidence. This is because if you’re wrong, not only have you charged an innocent person, you’ve not got the real criminal. Another high error cost, and so another case wherein we have moral reasons to drive us to search for excellent evidence. Clifford’s cases, because they are high error cost cases, have demanding evidential standards built into them. In SO’s capacity as a person in charge of the ship’s safety, he has a role-responsibility to ensure the ship really is safe. When

Reading Clifford’s “The Ethics of Belief ”

33

he doesn’t do that, he’s grossly negligent. Even when he’s lucky and right. The question, then, is whether all cases are like that. They aren’t, and a very simple case can show this. Consider my buddy, Karla. She’s a fan of the football team, the Cowpokes. Every season, she’s on the phone with me telling me, in real sincerity, that this is a championship year for the Cowpokes. This year, they can beat the Hornblowers, and their new recruits look to be the best yet. She’s already planning on going to the Championship Game. Karla may have some skimpy evidence here and there, but it’s the usual early season team affirmation. Redneck, Hornblower, Cowpokes, and even the lowly Detroit Lions fans do it every year. Now, I see this as a low error cost situation, at least compared to the ship owner and island cases. Again four possibilities arise: (i) Karla believes the Cowpokes will win out, and they do; (ii) Karla believes, and they don’t; (iii) Karla doesn’t believe, but they do; and (iv) Karla doesn’t believe, and they don’t. Neither of the errors, (ii) or (iii), seem particularly costly. In (ii), Karla’s disappointed, and may need to sell her Championship Game tickets. That’s bad, but no biggie—there will be a very happy Hornblower fan to pick it up. In (iii), Karla’s a pessimist, but she gets pleasantly surprised. Maybe she can pick up the tickets from a disappointed Hornblowers fan. Notice that this Karla the Cowpokes fan case would work for either epistemic insolence or sloth. That is, the basic error structure stays the same whether Karla believes contrary to very good evidence that the Cowpokes couldn’t even beat the lowly Lions or if she believes on the basis of insufficient inquiry, where some crucial bit of evidence was a simple Google search away. The point of this example is to show that the structural features of the cases Clifford presents as exemplary of the general evidential norm aren’t so general. Clifford’s cases are all high error cost cases, and that drives up the demand for evidence. But in low error cost cases (like Karla), it looks like the evidential demand peters out. Heck, Karla might even be found blameworthy by her fellow Cowpoke fans in case (iii), as they might hold her to be a “fairweather fan,” which means not really a fan. And so, as the objection goes, Clifford’s move from his cases to a general ethics of belief is a hasty generalization. Before he can proceed, Clifford must answer the generalization problem. The basic strategy is to notice that Clifford’s initial cases were exemplary of a demand for evidence, because the failure to attend to and believe in accord with one’s evidence in those cases was clear moral negligence. However, given the case of Karla, it seems there are cases where believing on the basis of insufficient evidence (either contrary to the evidence or without fair inquiry) is not morally negligent in the same sense. Clifford’s answer is that these cases only seem not to

34

Evidentialism and the Will to Believe

have massive moral negligence, because the costs with these cases are not to be found in the immediate error costs incurred, but in their broader consequences. Clifford acknowledges that his previous examples were ones where “the belief held by one man was of great importance to other men” (L&E 170). But for his view to be properly general, the duty of inquiry and evidence we see in the prior cases must extend even to the “trivial belief.” Clifford gives two forms of argument from consequences against what we might call seemingly innocuous overbeliefs. Karla the football fan is exemplary. Let us start with the term overbelief, which we will borrow from William James. It is a general term for a belief that goes beyond what the subject’s current evidence supports. From our previous cases, SO had an overbelief that Seaworthy was true, the GAs had an overbelief that members of the religious minority are kidnappers, and Karla has the overbelief that the Cowpokes will have a good season. The generalization problem is that what’s clearly objectionable about SO’s and the GAs beliefs isn’t objectionable about Karla’s, so Karla’s is a seemingly innocuous overbelief. Clifford’s argument is that such seemingly innocuous overbeliefs, again, only seem to be innocuous, because their costs are not immediately visible. Instead, they become larger, systematic costs. As a consequence, all overbelief is costly. Even the ones that seem innocuous. Clifford’s two forms of argument deserve separate treatment. The first I will call the content argument, and I will call the second the habit argument. The content argument is made in an epistemic holist fashion. In essence, Clifford’s view is that when we add overbeliefs to our set of beliefs, we make them part of the set of commitments we use to test further commitments. So when an overbelief is allowed in, it not only brings one credulous belief in, but it also allows in more over time in the form of their being consistent with the initial overbelief. Consequently, credulous belief is not just a one-off case, but once the belief is added to our commitment set, it undermines further evidential assessments: [The belief] goes to make a part of the aggregate of belief which is the link between sensation and action at every moment of all our lives, and which is so organized and compacted together that no part of it can be isolated from the rest, but every new addition modifies the structure of the whole. (L&E 168)

Clifford’s line of thought is a form of a conformational holism, such that once we start uncritically adding beliefs to our system of beliefs, our capacity to critically assess cases (in high error cost cases, as it turns out) is controverted.9 The argument proceeds according to three assumptions:

Reading Clifford’s “The Ethics of Belief ”

35

The background thesis: Our means of evidence assessment is dependent on our broad set of background beliefs and knowledge. Holism: This background is an organic whole, so as we change, add, or subtract parts, the whole functions differently. Coherentism: The whole functions in a way that if there are beliefs (q, r, s) like p in the background, then the coherence of p with q, r, and s is a reason to think there is evidence that p. The consequence, then, is that when we uncritically believe, we add those uncritical beliefs to our background. When we add them to the background, we change our system of evidence of assessment to favor beliefs of this sort.10 If we change our system of beliefs to favor a belief we’ve uncritically adopted, then we’ve taken a good deal out of the power of critical evidence assessment. And so, we might think, when we do come to a high error cost case, we may do our evidence assessment, as required, but the tools we use for that assessment are corrupted and ineffective at detecting truth. And so our seemingly innocuous overbelief yields, down the road, morally objectionable negligence: No real belief, however trifling and fragmentary it may seem, is ever truly insignificant; it prepares us to receive more of its like, confirms those which resembled it before, and weakens others. (L&E 169)

Once these overbeliefs are in the system, they retard the responsible evaluation of other beliefs, and so there are never any innocuous overbeliefs. Return to the Karla case. She holds the sincere overbelief that the Cowpokes will win the championship. How does the content slide happen? Well, perhaps with the fact that teams that are of championship caliber should get media coverage. But the Cowpokes don’t get any. Well, there must be some reason for that, thinks Karla, and she then hypothesizes that the media is ignoring the Cowpokes. There’s likely a conspiracy against this team—they clearly are contenders, but get not even a sniff. The fix must be in. Nobody wants the Cowpokes to win, so the media will ignore them with hopes they’ll go away. Every expert who holds the Cowpokes have no chance, then, is part of the conspiracy, and just empty talking heads. And so Karla has gone from one overbelief born of enthusiasm to now a belief that she can’t trust most of the experts on a matter of circumstance to her. If this sports case isn’t moving, consider this analogous to politics and one’s favored candidate. Clifford’s move here is to observe that this holistic content slide from seemingly innocuous overbelief to systematic irresponsibility does not occur merely in one person, but among communities. We speak and pass on our

36

Evidentialism and the Will to Believe

commitments, reasons, and so on. And they become public property. Clifford holds that “no man’s belief in any case is a private matter which concerns him alone,” but because we share our commitments with others, they become society’s property. We, given the consequences that can follow from subverting the system of belief, have a responsibility to maintain this system’s integrity, to improve it, instead of degrading it. Into this, for good or ill, is woven every belief of every man who has speech of his fellows. An awful privilege, and an awful responsibility, that we should help to create the world in which posterity will live. (L&E 169)

Belief is a kind of “heirloom” we can pass along to those around us, a “precious deposit” that we collectively share. It should be used for truths, and these provide stability and “bind men together, and strengthen their common action.” (L&E 170) Remember that we’d noticed that the intellectual duties of inquiry are indexed to the social role of the believer.11 The ship owner has certain responsibilities of safety, and so has specific cognitive responsibilities of ensuring it (not just insuring against it). Same goes for those who make accusations and take it upon themselves to publicize them—they have the responsibility of fact checking before they go to print. But is there a general obligation of epistemic responsibility? We’ve seen all the cases proceed from special responsibilities, but is there a role responsibility common to all humans? Clifford’s answer is in the affirmative: It is not only the leader of men, statesman, philosopher or poet, that owes this bounden duty to mankind…. No simplicity of mind, no obscurity of station can escape the universal duty of questioning all we believe. (L&E 171)

From the special cases, we see the seeds of a universal duty of critical reflection and inquiry. This is, again, because we all partake in and contribute to the common benefits of humankind. In the same way as we ought not subvert our own individual capacities for critical reflection by adopting a belief uncritically, we owe an obligation to the rest of humankind not to subvert our shared resources. An objector may challenge along the following lines: Why can’t we have oneoff exceptions? Why must every one of our moments of fanciful belief or exercises in wishful thinking poison our capacity to think critically later? Surely you can be an effective critical thinker when it’s needed even after you’ve performed some wishful thinking about your favorite football team. Enthusiastic fans don’t just start to walk in front of busses or eat glass for breakfast.

Reading Clifford’s “The Ethics of Belief ”

37

This challenge leads Clifford to his second reply to the generalization problem, what I call the habit slide. First, Clifford observes something about humans: we like thinking, we enjoy the thought that we know things, that’s why we prefer belief to doubt. Belief is satisfaction, doubt dissatisfaction. Belief provides assurance, doubt does not. It is the sense of power attached to a sense of knowledge that makes men desirous of believing, and afraid of doubting. (L&E 172)

Belief, then, is a temptation. That’s why we are so keen on finding exceptions to the evidence norms, why we’d like for Karla’s wishful thinking to be innocuous, why the odd moment of credulity can be OK. We want to know, and even if we don’t really know, believing we know is still pretty satisfying. But that satisfaction, holds Clifford, is inappropriate: If the belief has been accepted on insufficient evidence, the pleasure is a stolen one. (L&E 172)

Now, so far, this point sounds more like foot stomping from Clifford. The question is: are there cases where we can hold a belief without evidence, but insulate it from the rest of our other beliefs so that it does not poison our broader capacities for critical scrutiny. And Clifford’s case is that such a belief would be “sinful, because it is stolen in defiance of our duty to humankind” (L&E 172). That’s not much of an answer, so far. Now consider another overbelief case. Consider Ben, my middle-aged buddy. Ben likes sausage but doesn’t like exercise. Trouble is, at 40, Ben needs more exercise than sausage if he plans on fitting in his clothes. Actually, Ben should just knock off the sausage entirely. So Ben’s now eating healthy meals. But then there’s Thursday night at the Cowpokes game. They have great sausages at Cowpokes games. So Ben says to himself, “Just this one.” Well, anyone who has been on any kind of diet or tried to quit drinking or smoking knows something about that sentence. It doesn’t mean what we think it means when we utter it. It really means, “I’m now off the wagon, but I’ll struggle for a while to keep up appearances.” And the same goes for anything that’s a genuine temptation—if we make the exceptions and do our best to pretend that they are simple one-off cases, we are really just rationalizing. We’ve just given into temptation, and we are developing the habit of doing so. Clifford takes this exception thought, now, to belief: Every time we let ourselves believe for unworthy reasons, we weaken our powers of self-control, of doubting, of judicially and fairly weighing evidence. (L&E 173)

38

Evidentialism and the Will to Believe

The consequence is that the habit of credulity begins to stick. Believe on the basis of enthusiasm Just this once? Engage in wishful thinking with the thought This time and no more? Clifford expects those not to be one-offs, but the beginnings of a habit: But a greater and wider evil arises when the credulous character is maintained and supported, when a habit of believing for unworthy reasons is fostered and made permanent. (L&E 173)

This permanency, then, becomes social, just as we’d seen with the content argument. We are constantly conversing and sharing our views with others. We do so both verbally and nonverbally. We are good mind readers. And as a consequence, it is not only the contents we share, but also our habits of critical reflection. Some set good examples, others not so good. And consider this, again, on analogy with Ben the dieter. Surrounded by others who make the exception makes him so much more inclined to get the double sausage dog with extra cheese. Just this once. Clifford makes an analogy with theft. There may be times when theft causes no immediate harm—you may steal a bauble from someone who will never miss it. But the harm isn’t to the one stolen, reasons Clifford, it is done to the thief and the rest of humanity: But I cannot help doing this great wrong towards man, that I make myself dishonest. What hurts society is not that it should lose its property, but that it should become a den of thieves; for then it should cease to be a society. (L&E 173)

The wrong done is that, in stealing, the one who steals becomes a thief and can pass this habit to others. The habit then runs amok—we become those who merely give into temptation, and perhaps no longer even see it as temptation at all. “The credulous man,” Clifford darkly intones, “is father to the liar and the cheat” (L&E 174). The basic structure of both of Clifford’s replies to the generalizing problem is that of slippery slope arguments. Slippery slope arguments take the following form: P1: Doing a seems acceptable. P2: Doing a sets a precedent for and is thereby a causal contributor to a series that results in z. P3: The result z is unacceptable. C: Therefore, doing a is unacceptable.

Reading Clifford’s “The Ethics of Belief ”

39

Clifford’s answer to the generalizing argument is that there are no exceptions to the rule that one should not allow overbeliefs. For all cases, the replacement of a, with overbelief, and in both arguments, z, the unacceptable outcome is the general inability to critically assess evidence and consequently morally objectionable negligence.12 There are only apparently innocuous overbeliefs. What distinguishes the content and habit answers to the generalizing problem is how P2 reads for the two. Here, I believe, are their rough and ready versions. P2-Content: Overbelieving that p adds p first to the believing subject’s background beliefs for critical testing and then to the community’s. This results in critical thinking being undermined, and this results in morally objectionable negligence. P2-Habit: Overbelieving is a case of giving in to temptation to believe, which easily becomes an individual’s habit of credulity. This habit is shared with others, and this undermines critical testing, which results in morally objectionable negligence. The key is that everything about Clifford’s argument depends on the slopes with P2 being genuinely slippery, instead of bumpy staircases, where something may be a prerequisite for a bad outcome, but does not guarantee or make it more likely than not. Are these plausible causal sequences? Do we, in Clifford’s spirit, have evidence to believe they are? The first issue is to determine what the relevant evidence would be. Surely we’ve all had experiences where what was supposed to be a one-off exception brought our resolve down on a matter, and we’ve all seen groupthink steer many in wrong directions. Things that are tempting can easily become addictions, and peer pressure is often surprisingly strong. So there is evidence for both P2Content and P2-Habit, as many of our experiences bear this out. The trouble is that this isn’t likely sufficient evidence for these views—anecdotal evidence is evidence, but it’s exceedingly weak, and often subject to selectional distortions like confirmation bias. Now Clifford’s case is one that appeals to the role of our cognitive behavior yielding broader consequences for the group in which we deliberate. His thesis with either version of P2 is that bad contributions to group deliberation tend to be amplified in the group. This is a testable hypothesis, and it turns out there is evidence for this view. In short, the two central problems for group deliberation are Error Amplification and Group Polarization. Error Amplification is that people use what are called availability heuristics to (a) search for evidence for views in question, and (b) assess that evidence. The

40

Evidentialism and the Will to Believe

trouble is that when there are familiar or significant events (for example a terrorist attack or a case of unfair hiring practice), we tend to think these things are more common than they likely are. As individuals, because we react strongly to these cases, we have a consequent bump in how often we think they occur. The trouble is that groups magnify that effect—when individuals have a bias, the group is more biased than any individual, when they, say, vote on a matter.13 A further consequence of this phenomenon, that biases get magnified by the group (likely because we, when we share information, present it as representative), is that we put a premium on what we count as shared or common knowledge. What this, in turn, yields is not just that errors are amplified, counterevidence and other considerations get drown out by confirming evidence. This related phenomenon is called the common knowledge effect—once we share information, it has an influence on how the group treats it as though it is central to understanding an issue, and consequently, many things are not disclosed to the group. Why? Because if an individual has information running counter to what’s considered common knowledge, the individual will not share that information (or feel pressure not to), because she is concerned that she is likely mistaken. Disagreeing with the group is not just socially, but also cognitively difficult.14 The second, related phenomenon is that of Group Polarization, which is that members of information-sharing groups have a tendency to have stronger versions of their views after engaging with the group.15 Now, this sometimes is because new information comes in from the rest of the group. But not all the time. In fact, the polarization phenomenon is that if you put people with the same information together, when they share that information with each other, you get a common knowledge effect, which then has an amplified effect on the individual members, who, even with no new information, hold stronger versions of views they had before. As Cass Sunstein puts it: People may polarize because they are attempting to conform to the position they see as typical within their own group. (2006: 95)

If arguments come from those one identifies with, they turn out to be especially persuasive. They will seem just right. And those from the out-group will consequently look less and less apt.16 So there is pretty good evidence that Clifford’s two slopes are slippery. This is most certainly the case with the content version of P2, as it looks like our contributions of assertion have the consequences in the content dynamics of the group to influence later evidence assessment. The Habit version of P2 still seems plausible, at least in the sense that it seems clear that we’ve all got tendencies

Reading Clifford’s “The Ethics of Belief ”

41

of credulity and even group settings set up for deliberation nevertheless will yield bad habits and amplify them. And less deliberative communicative circumstances have more profound distortions.17 But the challenge can be rephrased: does this evidence support the strength of Clifford’s conclusion? Here, it’s clear that the answer is that it doesn’t. Clifford’s thesis, given the slopes, is that: No real belief, however trifling or fragmentary it may seem, is ever truly insignificant. (L&E 169)

All the evidence points to is that we have a tendency to those slopes, but it does not show that every overbelief will have such consequences Clifford foresees.18 But I believe that Clifford’s view can be modified to be acceptable both to the Cliffordian and to the objector. The case for it goes as follows. Given the evidence that we have both Content and Habit slopes, it is unwise to overbelieve and give into credulity. This is not because every case of giving in will have bad consequences, but because we know that the bad consequences are likely in light of the slopes, it is simply negligent to roll the dice with “just this one.” Even if it, contrary to our evidence, does not yield a form of error cascade we might see with the Amplification Problem, that doesn’t mean it was fine to have the belief. That, again, means we were just lucky. Given what we know about how we deliberate and how our habits of mind are formed, it is epistemically insolent and thereby negligent to overbelieve even if the belief doesn’t yield the bad consequences. My solution here is to push a version of our objection to the second version of the ship owner case (SO*). Recall that the problem with SO* was not that the ship actually went down, but that given the evidence before him, SO* had every reason to expect trouble, but was lucky that the ship made the voyage. SO* endangered the crew and emigrants. So it is not the case, even with immediate consequences for overbeliefs, that the belief must have bad consequences to be bad. Rather, it’s that the belief must needlessly endanger others, and even if it all turns out fine, it’s in the needless endangerment we see the problem. The same I’m arguing works on the analogous case with the seemingly innocuous overbelief. Given our evidence of human and group psychology, overbelief is dangerous business. Even if it turns out that some are inert with regard to the bad consequences, the fact that so many have the bad consequences makes it objectionable to allow any. And so Clifford may generalize to the moral estimation of overbelief from these cases to all, but it is not because (as he believes) of the consequences they all have.

42

Evidentialism and the Will to Believe

5. Ethics and belief So far, we’ve been able to see that Clifford has made a case for the ethical evaluation of beliefs, specifically overbeliefs. A question lingering in the background, if we concede that beliefs can be ethically considered, is what mode of ethics bears on beliefs? So far, most of the reasoning has been broadly casuistic, in that I’ve baldly moved from showing that overbelief endangers others to the fact that it is wrong. The question is: what’s Clifford’s take on this connection, and is he right about that? On this matter, there is little to no interpretive agreement, so it’s worth our time to untangle a few threads. Now, to be clear at the outset, I’m not convinced that Clifford’s overall project lives or dies with any of the interpretive options we take, whether we see him as a consequentialist, deontologist, or other. The argument works, I think, with any of the metaethics on the table, and it may be a virtue of the essay that its conclusion that it is wrong to overbelieve depends uniquely on no substantive view of the right or the good. But, again, there is a question about Clifford’s metaethics in the essay, so we must pause to address (and perhaps resolve) the issue. My plan in this section is to present the cases for the deontological and the consequentialist readings and show that the balance favors the deontological theory, specifically, a form of stoic deontology with regard to human capacities. This said, I will then make my case for an ecumenical reading of the essay, so that, again, consequentialist or deontological, the argument is still reasonable. To begin, the case for Cliffordian deontology is focused mainly on the prevalence of the term “duty” being so widely applied in the cases of belief evaluation. Deontology is marked by the primacy of duty to moral evaluation, the demarcation of a set of rules for action (like a set of categorical commands), a focus on the primacy of the right to the good, and a form of moral absolutism that determines the rightness or wrongness as a matter internal (and hence timeless) to what is evaluated. Consequently, the normativity being applied when one invokes a cognitive duty shirked is that one’s moral outlook is that duties and their being performed is what matters. Clifford’s track record on this matter is quite impressive. Here is a sample: – The “Duty of Inquiry” (The title of section I) – A “universal duty of questioning all we believe” (L&E 171) – The “duty of investigating the ground and strength of (one’s) convictions” (L&E 167)

Reading Clifford’s “The Ethics of Belief ”

– – – – –

43

The fact that the agitators “had no right to believe…” (L&E 167) A “necessary duty” of fair inquiry (L&E 168) A concession that “the duty is a hard one” (L&E 171) A “duty to guard ourselves” from overbelief (L&E 172) Those who have “done their duty” of “scrupulous care and self-control in the matter of belief.” (L&E 177)

I could go on. The point is that Clifford’s overt language is heavily deontological when he speaks of the norms bearing on belief management. Duty, duty, rights, duty. And when he’s not using that vocabulary, he’s waxing into high church rhetoric of command language. He calls the failures of duty: “sinful” (L&3 172), “sin(s) against humankind” (L&E 174), and “sacrilege” that “desecrates belief ” (L&E 170). In the same vein, he invokes a “sacred tradition of humanity,” and calls belief “a sacred faculty” (L&E 170) that with knowledge yields a “pure and holy temple” (L&E 198). My presumption is that a good deal of the evidence used by those who take Clifford to be a deontologist is in the Duck Test category, which is the humorous joke to capture the thought behind inductive reasoning: If it walks like a duck, swims like a duck, and quacks like a duck, then it’s probably a duck.

Here, the point is that Clifford is quacking like a deontologist. This may be weak evidence, so some supplemental reasons are worth bringing forward: Clifford’s view is that the evaluation of actions are static, once the action is performed, no consequences or contingency can change their evaluations. – Overbelief is a “sacrilege which centuries shall never be able to blot out” (L&E 197) – It is a “stain which can never be wiped away” (L&E 174) – The credulous are “cast into darkness and oblivion forever” (L&# 198) – “When an action is once done, it is right or wrong forever; no accidental failure of its good or evil fruits can possibly alter that.” (L&E 165) That’s swimming like a deontologist. And finally, there’s the starkly stated deontological-sounding statement: [I]t is wrong, always, everywhere, and for anyone, to believe anything on insufficient evidence. (L&E 175)

And that’s deontological walking. Vorstenbosch calls this a kind of “radical deontological stance” (2000: 99), and it is in the fact that it is “always” and for “anyone,” that it seems a particularly uncompromising deontological line.19

44

Evidentialism and the Will to Believe

The case for a consequentialist reading of Clifford must proceed along two lines. The first is that of showing Clifford’s consequentialist reasoning in the essay, and the second is addressing all that duty-talk from Clifford. If it quacks like a duck and you’re saying it’s a dog, you’ve got to explain not only why you hold it’s a dog, but also what’s up with all the quacking. The consequentialist reading begins with the case from the consequences of belief. In essence, Clifford’s reasoning is that beliefs matter because they are the antecedent causes of action; they are premises in our practical syllogisms. Because our beliefs are what drive us to act in one way or another, they are morally evaluable in terms of the moral qualities of the actions they give rise to. The crucial thing, then, is that the actions of endangering others, as we saw in the ship owner and island cases, is what explains why the overbelief in the cases were unacceptable. And the consequences of apparently innocuous overbeliefs are those that lead to the conditions for other beliefs that themselves yield endangering actions. As Brian Zamulinski frames the argument: What ultimately matters are what we do and how we affect others through our actions. (2002: 439)

That’s consequentialist reasoning par excellence, and for sure, that’s how Clifford’s case for the “duty” to investigate in all cases is grounded. He uses the term “duty” only to highlight the fact that it is a moral obligation which, given the consequentialist argument, has no exceptions. That looks deontological, but it only looks so. Zamulinski extends the case for consequentialism in that he takes it that “non-consequentialism provides no grounds to justify rejecting rather than relabeling any propositional attitude” (2004: 811). Have a cognitive attitude about the truth of a proposition, but it’s not supported by evidence? Well, the trouble isn’t with the lack of evidence, it’s with the belief. That’s just a verbal solution, Zamulinski holds. The trouble, he argues, is that the consequentialist, but not the deontologist, can explain the transition from the hypothetical imperative: If you are seeking truth, then proportion belief to evidence.

To the categorical imperative: Proportion belief to evidence.

The consequentialists can, as they have the Content and Habit slope arguments from before on their side—that is, arguments from consequences of not following the rule. You should seek truth, because doing so prevents the kind of moral errors we see in the ship owner and island cases, and so you should proportion belief

Reading Clifford’s “The Ethics of Belief ”

45

to evidence. The deontologist is in a tough spot to explain why one categorically should pursue evidence, without appeal to getting other goods or achieving some goal or preventing bad things. But that’s not categorical, then. There is no “objective moral obligation” with belief, except through consequences in action. I think it is hasty to call the debate in favor of the consequentialists, though it is clear that Zamulinski’s case is compelling. But there is another option, which is a kind of mixed view, a stoic ethics. My case for the stoic interpretation will proceed in two stages. The first is to show that Clifford’s writing has a distinctive stoic stamp on it, and even Clifford’s initial ship owner example is in the form of stoic standard fare for philosophical reflection. The second stage is to make the case that stoic deontology is more a mixed view, one that integrates consequences of actions as ways of determining role duties. As a consequence, I think a defense of the deontological interpretation can be made that can answer Zamulinski’s challenge. First, Clifford was, given other aspects of his life, a neo-stoic. In “The Ethics of Religion,” he approvingly quotes Seneca, noting that were people morally equal to their gods, antiquity would have been morally horrible from top to bottom (L&E 217). In “Cosmic Emotion” (L&E 399–400) and “The Ethics of Belief ” (L&E 167), Clifford invokes the Ciceronian notion of standing before one’s conscience. Famously, Clifford’s epitaph was from Epictetus: I was not, and I was conceived; I loved, and did a little work; I am not, and grieve not. (From Madigan 1999: xx)

And Clifford goes so far in “Right and Wrong” to claim that stoicism is “one of the most important expressions of moral sense of all time” (L&E 133). That Clifford was an admirer of the stoics is one thing, but applying their views to the ethics of belief is another. But it is not so far off, as Cicero’s On Duties opens with a short discussion of the first duty, knowledge of the truth. It is at the core of all other duties, and in fact, the primary error of humans is “treating the unknown as known too readily” (1913: I.vii.18–19). Cicero’s later famous examples also match Clifford’s form for the ship owner and island cases. Consider the Ciceronian case of the house for sale, where the current owner suspects it is infested, but sells the house anyway (1913: III.xiii.55–56). Another is that of two Romans named cobenefactors of a large will with some men they suspect have forged the will, but they, instead of investigating, go along with the petition (1913: III.xviii.73). And in Paradoxa Stoicorum, Cicero even has a ship case parallel to Clifford’s.

46

Evidentialism and the Will to Believe Whether a helmsman capsizes a ship with cargo of bullion or a barge laden with chaff makes some difference in the result, but none in respect to the helmsman’s incompetence. (1942: III. 20)

Cicero holds all these men’s actions fraudulent, as they forego their duties for convenience. The point is that the Ship owner and Island cases are structurally similar stoic problem cases, ones that direct and clarify the views to be articulated. And, as we see with Cicero’s helmsman case, it is specifically in the incompetence of the case, not the outcomes, we have the evaluation. Clifford’s overall view, that of all humans having a kind of role obligation of intellectual honesty, given their membership in the human race, is a centrally cosmopolitan stoic view. Again, Cicero held that it is simply in virtue of being a human being, with the social bonds we have, to make oneself at home in the world by being a conscientious thinker (1913: I. vi.17), and the first of Epictetus’s domains of perfection was, “In judgment, avoid error, control one’s assent, to strive to see and understand only truth” (1928: D.3.2.1–2). And Marcus Aurelius, whom Clifford refers to as the enlightened stoic emperor (L&E 133), insists that “what (truly) harms us is self-deceit and ignorance” (2002: M.6.21). The Marcan view is that we contribute to the good of humanity in the rational directions we take “to work with others” (2002: M.8.12). Further, Clifford’s commitment to the completeness of the error of overbelief is also classic stoic ethics, as the stoics hold, as reported by Cicero, all good acts are equally good and all bad acts are equally bad, that “transgressions are not measured by their results but by the vices of the persons transgressing” (1942: III.20). Clifford’s view of the completeness of wrong is parallel, as he holds that “whoso shall keep the whole law, and yet offend in one point, he is guilty of all” (L&E 175). The point here is that stoic ethics integrates elements of consequentialist reasoning to identify duties. Why might one have role-related duties, to begin with, unless the role itself had certain objectives that would be retarded unless one performed those duties? And in this respect, Clifford even invokes the utilitarians in “Right and Wrong” to identify why veracity is a virtue, more truthfulness is conducive of stability (L&E 157). Clifford was: first, familiar with; second, amenable to; third, regularly invoking; and fourth, in large parts deploying the techniques of the stoic tradition. And finally, given the integrated nature of the tradition, conceiving of duty as contributing to one’s own virtue and to the harmony of humankind, the stoic ethical tradition is well placed as a compromise system between the stark deontologists and consequentialists.20 Now, to return to Zamulinski’s challenge. The challenge was that only consequentialist ethics could make the case for an ethics of belief. I think the

Reading Clifford’s “The Ethics of Belief ”

47

stoic tradition’s integrated approach to ethics, one that takes the theoretical objectives to be coordinate with practical objectives in the good life, can provide the requisite ethical justification. So long as, with Cicero and Epictetus, we take the perfection of judgment to be part of the good life and right living, it has ethical import. It is worth making a final point about Clifford’s rhetoric of presentation. It has, in William James’s estimation, “too much robustious pathos in the voice” (WB 18). Richard Gale jokes that “Pomp and Circumstance” should be played in the background for “maximum effect” (1993: 355). Clifford speaks of sin, desecration, and endless oblivion, and he invokes a sacred faculty of belief, a pure and holy temple of knowledge and truth. Let us start by noting at least a kind of irony in the fact that this vocabulary is being used to argue against believing in miracles and revelation. This is the rhetoric of the church, the vocabulary of vehement religious belief. The common use of such KingJamesified vocabulary is in the services of eliciting assent to religious doctrine. Clifford’s use of the vocabulary, then, could be a kind of subversion of the vocabulary of religious belief, a hearty needling religious believers with their own words. But I think there is a different reason, running parallel. Clifford is not merely invoking a religious rhetorical tradition to subvert it, but to harness the moral insight behind it. That is, Clifford’s strategy was not merely “polemical,” as Bernard Lightman has argued (1987: 120), but rather it was a move to take the core moral commitment behind the religious tradition to direct its critical gaze on the cognitive excesses of the tradition. Clifford’s use of religious vocabulary is a form of internal critique. He uses the moral norms of the religious believers as the tools to criticize their methods of forming beliefs about the basic features of reality. This is why he, further, quotes Milton, Colerige, and Abijah (from I Kings 14: 16, who rebukes Jeroboam’s indiscriminate conferrals of priestly power). It is to show that the Christian tradition has a rationalist strain, that of responsible pursuit of truth. It is no mere rhetorical ploy, but rather a form of argument that shows that even without belief in miracles and other fanciful things, there are nevertheless sins, duties, sacred faculties, and holy things. But these are natural things.

6. Endorsing evidentialism The concluding paragraphs of Section I of “The Ethics of Belief ” contain the main target for the essay’s argument: the statement of the norm that one should

48

Evidentialism and the Will to Believe

care for evidence and have one’s beliefs directed by it. Let’s call it Clifford’s Evidentialist Norm (CEN): [I]t is wrong, always, everywhere and for any one to believe anything on insufficient evidence. (L&E 175)

The CEN is strict; Clifford makes no exceptions. There, in light of the two slope arguments discussed earlier, are not one-off or inconsequential cases of overbelief. They lead either directly to or create conditions for actions that are irresponsible, and so they themselves are irresponsible to hold. Notice a few things about the evidentialist norm in Clifford’s hands. CEN has belief seeming to be ON or OFF—one believes or does not. Evidence, too, has an absolute element, as it, even as degreed, is sufficient or insufficient. Compare this conception of belief and evidence with other statements of evidentialist norms in the Enlightenment tradition. Descartes states a version of the norm in the Fourth Meditation with the clarity and distinctness rule: If… I simply refrain from making a judgment in cases where I do not perceive the truth with sufficient clarity and distinctness, then it is clear I am behaving correctly and avoiding error. (1996: AT VII. 59)

Of course, the Cartesian notion of evidence is very strict, but it is analogous to Clifford’s, as it runs that one withholds belief until it reaches an appropriate threshold. Locke has a version of the evidentialist commitment in An Essay on Human Understanding The mind, if it proceed rationally, ought to examine all the grounds of probability, and see how they make more or less, or against any probable proposition, before it assents to or dissents from it. (1996 [1690]: IV.xv.5)

Hume, in “On Miracles,” states his evidentialist norm: A wise man … proportions his belief to his evidence. (1993 [1748]: 73).

Hume’s version has belief as degreed, and evidence as degreed with no cutoffs for belief or suspension. Credence may, then, vary according to the valences of the strength and direction of the evidence. We see two elements of the norm in contrast, as, again, there may or may not be a degreed notion of belief— whether belief is considered on a model of assent, abstinence, or denial. The second is in terms of what grounds the norm. In the case of Descartes, it is a truth norm (or better, falsity-avoidance norm), as it is that one should follow the clarity and distinctness rule to ensure one never has errors. Hume invokes wisdom and Locke rationality, both of which give the impression that they take

Reading Clifford’s “The Ethics of Belief ”

49

the justification for their respective version of the evidentialist norm to depend on what the cognitive enterprise is. That is, wisdom is about intellectual insight, and rationality is about making the right theoretical decisions when in matters of belief. Practical considerations may be further down the road and relevant, but they are not what yield the norms. Clifford’s view is unique from all of these, then, in being a moral case for the evidentialist norm. Again, I see this view as part of an older inherited stoic tradition, but it is worthwhile to make the contrast with the modern period and its progeny. Even in the twentieth and twenty-first centuries, divisions between those who hold that the case for the evidentialist norm must come exclusively from the norms of theoretical rationality and those who hold the norm depends substantively on practical rationality (specifically, ethics) still obtain. On the one hand, there are the contemporary ethicalists like Wood (2008) and Zamulinski (2002 and 2004), and on the other hand, there are the cognitivists like Feldman (2000 and 2006), Adler (2002), and Shah (2006 and 2008). In the final chapter, I will work out what I see as an ecumenical evidentialism, but one that is considerably more amenable to the cognitivist’s camp. Regardless, it’s clear that Clifford is in the ethicalists’s camp, and it seems he is, perhaps, its most prominent representative. This said, it could be objected that Descartes’s model is itself ethicist, instead of cognitivist, as I’ve portrayed it. He holds cognitive error as analogous to sin and the abuse of free will. If so, Clifford’s comes in second place. Still, not too shabby. Clifford’s rule, CEN, however, is stated as a synchronic norm, amounting to the injunction to never be epistemically insolent. But what about epistemic sloth? That is, Clifford’s two identified cognitive errors were insolence and sloth, but the main summary focuses only on the synchronic sin of insolence. Clifford addresses sloth with the follow-up rule: It is never lawful to stifle a doubt; for either it can be honestly answered by means of the inquiry already made, or else it proves that the inquiry was not complete. (L&E 176)

This is the prohibition against sloth. These are two faces of a core commitment that requires they be coordinated. And so, we have a joint statement of evidentialism, which can be captured by the broader Integrated Evidentialist Rule. (IER): If any subject (S) believes any proposition (p) at any time (t), then S has properly done so only if: (i) S has sufficient evidence at t that p is true, and (ii) all doubts S has had (and should have) regarding p’s truth or falsity have been

50

Evidentialism and the Will to Believe investigated so that there are no truths S could have easily discovered that would have affected S’s evidence.

In short, according to IER, to believe responsibly, subjects must have sufficient evidence and have done due diligence in checking things out. Clifford considers two objections. The first comes from those whom he classes as “have been brought up from the cradle with a horror of doubt” (L&E 175), and the second comes from those he calls “busy” people, who want to believe, but have no time for inquiry (L&E 176). His answers to these two are telling. To those who are taught to fear doubt, Clifford clearly sees them as those who don’t fear just any doubt, but specifically religious doubt. His answer from the religious poetry of Milton and Coleridge, which invokes the value of truth over belief, has twofold effect. On the one hand, these are, as we noted earlier, religiously founded reasons to care for demanding cognitive norms. One might say, using the rhetoric of those who might fear doubt, Clifford’s charge is: true religion does not encourage credulity, but intellectual honesty. On the other hand, the force of the content of both Milton and Coleridge’s lines are that the profound religious errors (heresy and loving oneself more than God) are expressed more markedly in one’s hasty beliefs, instead of one’s responsible inquiry and doubt. The contrast is that even if one fears doubt, one should fear the sins of heresy and self-love more. And consequently, the pious should prefer the IER. The second objection is that one is too busy to take up “the long course of study” that is required to avoid the errors of cognitive sloth. There are many things we would like to have opinions on, but we do not have sufficient time for developing informed opinions. To these, Clifford holds that if one does not have the time to inquire, then “he should have no time to believe” (L&E 176). That’s pretty harshly put, but it’s certainly the responsible attitude—if you don’t know the facts of a case, even if you’d really like to have an opinion, if you haven’t the resources to be informed, you haven’t the right to have those beliefs.

Section II—The weight of authority 1. Anti-skepticism Clifford sees the most pressing objection to his version of evidentialism, which we can see in the IER that one may believe only on the basis of sufficient evidence and only after one has thoroughly investigated the issue, is skepticism.

Reading Clifford’s “The Ethics of Belief ”

51

Remember that Clifford’s harsh reply to the busy believer to close Section I was that if one does not have time for fair inquiry, then one cannot responsibly believe. Trouble is, the objection goes, we have lots of cases where our actions require we act before we have perfectly clear evidence or before we’ve answered all our doubts. That’s, like, what it is to be busy. The consequence is not just skepticism, but apraxia: Are we then to become universal skeptics, doubting everything, afraid always to put one foot before the other until we have personally tested the firmness of the road? (L&E 176)

Clifford’s image is the joint story of taking the connection between belief and action seriously, and enforcing the requirement that we not believe on the basis of insufficient evidence or incomplete inquiry. It’s a humorous picture: unsure the ground in front of him is solid, the Cliffordian then cannot act. Even walk. Skeptical, as we see we must suspend judgment we do not know; apraxia, as we, having suspended judgment, have no beliefs to direct action. Clifford takes the purportedly skeptical consequences further, as it seems he sees the potential objector then taking it that until we have powerful evidence that what seems obviously correct moral norms are right, we have no reason to follow them, either: Shall we steal and tell lies because we have had no personal experience wide enough to justify that it is wrong to do so? (L&E 177)

Importantly, this objection is substantively different from the skeptical apraxia objection, as it seems that we have endorsed default practical attitude in the face of what seems to be insufficient evidence. That is, it seems that Clifford’s anticipated objector is taking, now, a skeptical immoralism line—that until moral rules have sufficient evidence, we are not bound by them, and so may steal and lie. The apraxia and immoralism lines with skeptical charges are, in form, reductio arguments. That is, they are arguments that show a view in question entails some unacceptable logical consequence, and so must be false. In both cases, the arguments run: P1: IER yields a form of skepticism, in practical matters. P2: Skepticism in practical matters yields, on the one hand, apraxia, and on the other hand immoralism. P3: Apraxia and immoralism are unacceptable. C: Therefore IER is unacceptable.

52

Evidentialism and the Will to Believe

The form of the argument as reductio seems fine, so Clifford must, to save IER, reject one of the premises. His main target is P1, but he will also make a case against P2. Clifford’s argument against P1 is that we do have sufficient evidence for many practical matters, and that we have no ground for skepticism in the kind of cases that drive the two reductios: The beliefs about right and wrong which guide our actions in dealing with men in society, and the beliefs about physical nature which guide our actions in dealing with animate and inanimate bodies, these never suffer from investigation; they can take care of themselves, without being propped up by “acts of faith,” the clamor of paid advocates, or the suppression of contrary evidence. (L&E 177)

Clifford’s point is that we, when we reflect on these issues, find “certain great principles,” which direct our lives and order our thoughts, and the more we think about them and investigate them, the better supported they are. This reply sounds more like bluster or mere optimistic cheering. Skepticism about induction is very hard to answer, as the Humean challenge is hard enough, but Goodman’s new problem seems downright befuddling. And moral skepticisms and the proliferation of substantive disagreement over both the nature of moral norms and their application seems more troubling the more one investigates it, not less. For now, let us allow the challenge and Clifford’s reply to be tabled. Clifford will return to the defense of the principle of the uniformity of nature in Section III, and we will allow the full case to be made before we pass judgment. The trouble is, though, that Clifford defends the inductive principle with explicit argument, but has no full section devoted to the defense of the moral principles under scrutiny. He will mount a short defense later in Section II, but it’s incomplete. This, of course, is the shortcoming of any essay—that it cannot be a defense of all the commitments comprising its argument—but this lacuna seems both glaring and appalling, given the centrality of Clifford’s moral case for the IER. That is, it should be the moral skeptic that should be Clifford’s prime target for criticism and correction here—one who demands evidence, but has no moral commitments. Surely this person, given the heavy breathing in reply to the immoralism reductio that these commitments don’t need to be “propped up,” should be keen to call Clifford’s bluff. Regardless, we, again, will table our analysis of Clifford’s case against P1 for the commitment to the uniformity of nature until Section III. Clifford’s case against P2 is that belief is certainly tied to action, but it is not the only source of action. The objection, in P2, confuses the order of conditions in the belief–action connection. All belief yields coordinate action. But the converse does not follow: that all action is a consequence of belief. Call this the

Reading Clifford’s “The Ethics of Belief ”

53

Conversion Fallacy (it will be important later, with James, as one of my critical lines with James will be that his argument proceeds on consistently committing this fallacy—see Chapter 2, Section X-5). We may act, holds Clifford, in many cases we do not have sufficient evidence to determine a belief, and moreover, those actions may yield further evidence: Moreover there are many cases in which it is our duty to act upon probabilities, although the evidence is not such as to justify present belief; because it is precisely by such action, and by observation of its fruits, that evidence is got which may justify future belief. (L&E 177–8)

So Clifford’s answer to P2 is that there is a class of morally responsible (and sometimes obligatory) action that does not require belief, but nevertheless yields evidence that can justify later belief. Such actions are hypothetical and tentative, but actions nonetheless. So apraxia is not a consequence of the skeptical upside of IER. Now, it is not difficult to see what the hypothetical-tentative actions Clifford’s view would enjoin. Return to the ship owner case, where SO has plenty of evidence that his boat needs an overhaul, but that may not be enough for him to believe it does. But SO still has practical responsibilities, given what the error costs are. Consequently, what SO should do is get the boat checked out. In so doing, he does his duty and collects more evidence in the process. In the island case, the members of the agitator group (the GAs) did not have sufficient evidence to believe members of the religious minority were kidnappers, but they had a worry, and it was worth making some inquiries, perhaps looking into recent abductions. When the GAs do that, they don’t have a belief, but nevertheless act to inquire, and in so doing, gather more evidence. And in both cases, there are some actions that duty, because of the error costs with the actions, requires they not perform. SO doesn’t suspend judgment, but then sends the unfixed boat out with the immigrants as an experiment to see if it will float, and the AG’s don’t just publish anything they dream up with the hope that maybe something will stick. Duty requires we not only have care with how we believe, but also requires that our actions, when we don’t know the facts, are directed in some ways and not others. So even if P1 is true (which it nevertheless may not be), P2 is false.

2. Testimonial evidence The question, if the evidentialist rule as IER is that one believes only on the basis of sufficient evidence and one seek out further evidence, is what counts

54

Evidentialism and the Will to Believe

as evidence? Now, since the issue for the Metaphysical Society’s lecture series is miracles, the focus of Clifford’s attention is on testimony. The reasoning in the background is this: we don’t witness miracles that often, but our religious traditions abound with them, and they take the fact of miracles to be evidence of supernatural forces, gods and angles, making themselves and their preferences known. We hear about these miracles, are impressed by them, and then come to believe religious truths on the basis of their testimony. That’s the target reasoning for Clifford’s criticism, and as I outlined earlier, this was the target reasoning criticized by Huxley, Stephen, and Carpenter—all of whom challenged whether these events could happen and whether we can trust the testimony of those who tell us it did. Clifford, now that he’s made a moral (and, arguably, a religiously moral) case for the duty to believe only on the basis of adequate evidence and to inquire into the adequacy of one’s evidence, turns to the question: is testimony of miracles sufficient evidence of their occurrence and of their supernatural causes? The question of when testimony provides sufficient evidence is tied to the question of when we can see it fails to provide evidence. Clifford’s view is that testimony fails when it fails either of two requirements: Veracity: The testimony is given by someone who is not lying, but trying to speak the truth. Knowledge: The testimony is given by someone who has satisfied the IER, that is, has sufficient reason and has exercised the right judgments of inquiry regarding it. The thought is, if we have grounds to believe that the testimony is a lie or not the product of conscientious inquiry (or as Clifford says, “is ignorant or mistaken” (L&E 178)), then we should not take the testimony as evidence. And consequently, Clifford holds that the case for adequate evidence requires that we must have reason to believe the testimony satisfies these conditions. In order that we may have the right to accept his testimony as ground for believing…, we must have grounds for trusting his veracity…; his knowledge …; and his judgment… (L&E 178)

And so with testimony, Clifford holds that it itself is not an independent source of evidence, but that for it to transmit reasons to us, we must have reasons to treat the testimony as reasons. We, with regard to many facts in the world, are epistemically dependent on others for information. The question is whether that dependence and the information that comes to us through communication has default evidential status for us or not. Clifford’s view is that it does not, but rather

Reading Clifford’s “The Ethics of Belief ”

55

that we each, individually, must exercise our own judgment to find whether each source is reliable or not, and only after we’ve answered affirmatively with regard to both the veracity and the knowledge conditions can we rightly treat such testimony as reason to believe. Clifford’s view here is a form of what can be called reductionist epistemology with testimonial evidence—namely, that one must have positive reason to trust in order to treat testimony as evidence. Hume, famously, stated a version of the reductionist view as a “general maxim” in “On Miracles”: [O]ur assurance in any argument of this (testimonial) kind is derived from no other principle than our observation of the veracity of human testimony, and of the usual conformity of facts to the reports of witnesses. (1993 [1740]: 74)

Note Hume’s reductionism: humans are generally reliable, so we will treat this or that piece of testimony from this or that human as reliable. Clifford’s is more demanding, as it seems he’s not just thinking one applies an a posteriori generalized rule to individual cases, but that we evaluate the individual cases individually. Clifford turns to articulate the two questions: Is he dishonest? May he be mistaken? (L&E 179)

And he holds that we must ask them in the cases of trusting each piece of testimony. Contrast this general reductionist view, if but briefly, with antireductionist views that assertions are credible evidence until shown otherwise. This view has its modern roots in Tomas Reid’s famous Principle of Credulity, which runs that since God created us to be social creatures, He gave us, on the one hand, a proclivity to speak the truth, and on the other hand, “a disposition to confide in the veracity of others and to believe what they tell us” (1983: 94–95). A test case may clarify the difference. Suppose you’re driving in a new town, and you’re lost. You just need to get back to the downtown area, and you’ll be able to re-orient yourself. But you’re really lost and can’t even find downtown. So you stop a person walking on the sidewalk and ask how to get to downtown. They tell you where to go, and off you go.

The difference between the two views is that the reductionist’s requirement is: to trust that person’s assertions, you need positive reason to do so. You might have reasoned in the background that the person, if walking, likely lives in the neighborhood, so knows the town, and seems honest enough. So, given these considerations, you’re reasonably trusting her. The antireductionist holds that you didn’t need to do that reasoning, you just had to go on her testimony. Only if

56

Evidentialism and the Will to Believe

you had reason to doubt it would you not trust her testimony, like, if she seemed confused about where downtown was or if she laughed like a comic villain after giving you the directions. The two views have their respective merits and drawbacks. For the reductionist, the good thing is that we are critical thinkers with regard to our sources; we test out those we rely on for information. This seems like responsible cognitive management. The trouble for the reductionist, though, is that it is often way too much intellectual overhead to keep track of people’s records and so on. The benefit of the antireductionist line is that it captures our natural inclinations of dependence. Children seem to automatically believe on the basis of the say-so of others—it seems the Cliffordian would hold such credulity in contempt. And it, plausibly, is here that the antireductionist’s view seems less a norm of what we ought to do to have the right intellectual scruples, but more a description of what we actually do when we want intellectual satisfaction. Regardless of how the debate between the testimonial reductionists and antireductionists shakes out, the burden objection to reductionists is magnified in Clifford’s case, as, again, his view is that we need positive reason in each instance of deference to testimony. Clifford’s case, then, will depend on how the account of the application of the two questions works specifically with miracles, but depends on whether it is plausible in broader application.

3. Miraculous testimony The most common error with testimonial assessment, Clifford holds, is that we take having established the veracity condition for testimony for having also satisfied the knowledge condition. This is most clearly taken to be the case with that in testimony of miracles. The excellent moral character of a man is alleged as ground for accepting his statements about things which he could not possibly have known. (L&E 179)

The fact that our sources of information about miraculous divine revelations are sincere in their reports is taken as evidence of their knowing what they are talking about. But this is an error. Clifford’s driving example is that of Mohammed’s reception of divine insight. The Muslim believer surely holds Mohammed’s moral character to be of the very highest, so he must surely satisfy the veracity condition. And his teachings, too, are full of important moral insights, ones that contribute to the dignity of the human race. This seems to be evidence that he’s right about God’s will.

Reading Clifford’s “The Ethics of Belief ”

57

Clifford’s reply to the Muslim case from revelation runs along two lines. However, a pause is necessary here, as it’s worth asking what the significance of putting the case for revealed truths and miracles in Mohammed’s hands is. Why choose Islam as the paradigm for revealed traditions and their case for the argument from testimony? Here, I think Clifford is genuinely taking advantage of his audience’s substantive religious views—namely, that they hold Islam is not the true religion. Perhaps, even, that it is a pernicious false religion. My supposition is that Clifford’s strategy is to show that the cool evaluation of the Muslim believer’s case for and then from revelation will receive maximum scrutiny from the Englishmen to whom the lecture was addressed. Then that critical eye will be turned toward their own favored views—critical thinking is best learned in seeing others’ faults in reasoning, and then eventually turned on one’s own reasoning.21 The first piece of pushback is not the clear non sequitur from the thought that Mohammed was an honest man to Mohammed had sufficient evidence for what he claimed, but on the veracity of the claim initiating the inference. Clifford notes “we should be tempted to take exception against his view of the character of the Prophet and the uniformly beneficial influence of Islam” (L&E 180). Religions have mixed records, and the believers have the tendency to pay attention to the good parts of the mix. It is a natural tendency (one earlier identified as a form of the availability heuristic of confirmation bias). The second pushback is the non sequitur: He can concede the sincerity of Mohammed’s claims, but that is insufficient to take this testimony as evidence: [T]he character of Mohammed is excellent evidence that he was honest and spoke the truth so far as he knew it; but it is not evidence at all that he knew what it was. (L&E 181)

Mohammed really believed he met an angel on a mountaintop that revealed deep truths to him. He was honest about that when he related it. But honesty does not mean truth, especially under the conditions of most revelation. Clifford makes this point clearer with the case of the Celestial Visitor. Suppose that Clifford goes into the desert, like Mohammed. Suppose further that he fasts, prays, and climbs a high mountain. Now imagine that he has the experience of a celestial visitor, who brings him truths of the universe and of the direction of humankind. Now, given that he’s not eaten, is hot and tired, and has been alone for a very long time, the first question to ask is whether this celestial visitor is real at all.

58

Evidentialism and the Will to Believe It is known to medical observers that solitude and want of food are powerful means of producing delusion and fostering a tendency to mental disease. (L&E 181)

Second, even were the celestial visitor real, the two kinds of things he tells us are of importantly different character. The injunctions to the just direction of society are ones we can test as to whether they are right, but the revelations of God’s nature, His desires, or some deep divine insight into reality is not something we can test and verify. But, now, Clifford is in a position to ask how the celestial visitor knows these things. Is the visitor, additionally, honest? How are we sure the visitor is that trustworthy at all. The most we know of the visitor is that he talks to people fasting on mountaintops, not to scholars, scientists, and researchers. Not a good start. Let us consider the two kinds of information from the visitor. Call one set of injunctions ethical precepts, the other metaphysical doctrines. Now, let’s imagine that, like Mohammed did, Clifford were to come back and believe both the ethical precepts and metaphysical doctrines. Now, if we admit for the sake of argument that the ethical precepts really do contribute overwhelmingly to the improvement of humankind, they offer peace and solace to the suffering, then is that not evidence they were right? Here, Clifford concedes with one move, but objects with another: We are only at liberty to infer the excellence of his moral precepts…. So that… it is the Prophet’s knowledge of human nature, and his sympathy with it, that are verified; not his divine inspiration or his knowledge of theology. (L&E 184)

The point is that we can verify the ethical precepts of the revelations, and (assuming that the celestial visitors are real), that they know these things in the fashion we came to verify them. But does this mean that they are right about the metaphysical doctrines? They impel us to agree to the ethical precepts, they bring us comfort and succor. Clifford demurs: The fact that believers have found joy and peace in believing gives us the right to say that the doctrine is a comfortable doctrine, and pleasant to the soul; but it does not give us the right to say it is true. (L&E 183–4)

The evidence for the two doctrines is different, and the efficacy of the ethical precepts is evidence of knowledge of the human capacity for good, but not of divine revelation. Note, further, that were there to be difficulties with the divinely revealed ethical precepts, that would be reason to question them. So revelation about matters we can investigate can have authority, but that

Reading Clifford’s “The Ethics of Belief ”

59

authority is dependent on how well it can be sustained in light of our critical questions. Clifford has provided two sources of challenge to the revelation strategy for religious doctrine. First, one can regularly challenge the moral character of those who are providing testimony, and second, even were we to admit their veracity in ethical matters, we need not, in the end, defer to them on metaphysical doctrine. We are in no position to attribute knowledge to them. So we are in no need, in the end, of deferring to them, and, in fact, have reason to be suspicious. One further problem for revealed traditions is their multiplicity. As Clifford puts it: If there were only one Prophet, indeed, it might well seem difficult … to decide upon what point we should trust him, and on what we should doubt his authority…. But there is not only one Prophet. (L&E 185)

The epistemic problem of religious diversity is simply that even if testimonies of revealed traditions have initial credibility, the fact that there are many and all with conflicting messages undercuts their individual credibility. The Buddha, too, was someone we’d be inclined to say passes the veracity test, but he says there’s no God, that we’ll be annihilated, and so on. But Mohammed says there is a God, that there’s Heaven or Hell waiting for us after death. Given that we have no means to verify any of these inconsistent claims and we have no reason to give any one of them more credence than another, they cancel out.

4. The publicity requirement Clifford pauses to clarify the thought behind his rejection of Mohammed’s testimony. The charge is that we cannot accept Mohammed’s testimony about the divine because he was not in a position to know the things he is reporting. This is because Mohammed would have had to have reason to hold that the one from whom he was receiving testimony was in a position to know the things attested. Since Mohammed may have had the ability to know that the angel knew the ethical precepts, Clifford is willing to allow this testimony. But because he was not in a position to know that the Angel Gabriel knew the metaphysical doctrines, we cannot hold Mohammed’s testimony as evidential. To be clear, it seems that the requirements of veracity and knowledge be used even with sources of revelation. We can set the chain up as follows, using the celestial visitor case as a model:

60

Evidentialism and the Will to Believe

Stage 1: A relies on testimony from B that p, so A must have reason to believe: (i) that B is honest, and (ii) B is in a position to know that p. We have, by hypothesis, no problem granting B’s honesty. But how can we grant B’s knowledgeability, if B is reporting what the celestial visitor revealed? This requires us to go to a second stage in the testimonial story: Stage 2: B relies on testimony from a celestial visitor, so B must have reason to believe that: (i) the celestial visitor is honest, and (ii) the celestial visitor is in a position to know that p. Again, the content of p matters in this case. If p is an ethical precept, we are in a position to verify that these attesters were in a position to know. But when p is a deep or abstruse metaphysical truth, say, that there’s a Heaven and what happens there, then neither A nor B is in a position to assess whether their attesters are in a position to know that fact, because that’s a fact inaccessible to us. So when the fact’s not one we can have evidence that the attester is reliable about, we have no reason to hold that the attester satisfies the knowledge condition. And so A cannot reasonably believe on the basis of B’s evidence, because A should know that B could not attribute knowledge to the celestial visitor. Neither, then, can A reasonably believe the testimony. Now, notice, again, that Clifford’s epistemology of evidence is a particularly demanding form of reductionism. The argument is running that the justification one gets from testimony is a function of the positive epistemic reasons one has for trusting the testimony. Now, again, because revelation has both the problem of unverifiable metaphysical claims and the problem of wildly inconsistent sources of revelation, there not only is no reason to trust it, we have a positive reason to distrust it. That’s the core of Clifford’s argument. Clifford’s requirements for testimonial justification are very demanding. They might be so demanding, they are absurd. To see the looming problem, consider the following case of phone number: Carrie wants to know Bill’s phone number. She’s in no position to know it, so she asks him. He then tells her. Carrie has testimony from a, we presume, honest attester, but because she does not know Bill’s phone number, how can she be in a position to know that Bill knows his phone number? For her to know that Bill knows his phone number, it seems she’d already need to know it.

Such a situation defeats the whole point of testimony—the more restrictive one is on testimonial evidence, the more reductionist one is on testimony, the less one sees it as testimony at all, but as allowing others to say things you already

Reading Clifford’s “The Ethics of Belief ”

61

know. Now, that’s crazy. And it would be very bad for Clifford if his requirement that we have reason to attribute knowledge to attesters before we can accept their testimony makes it so that we had to already have the knowledge we needed from them. The phone number case demonstrates what I call the attribution problem for the knowledge condition for testimony. It seems that if you say that someone knows something, you’ve got to know it, too. Again, if I were to say of you that you know that Barack Obama is the president of the United States, I’d need to know that myself beforehand. So how do epistemically dependent people judge that attesters live up to the knowledge condition of testimony without simply no longer being epistemically dependent? Again, Clifford’s view comes very, very close to sounding like that paralyzing knowledge-attribution cascade, so it’s very important to be clear whether there’s a difference and what it is. Or else, Clifford’s case against testimony about miraculous revelation pretty much applies to all testimony. This means that Clifford ends right back in the skeptical reductio he was out to avoid. Clifford presents two cases to clarify the knowledge requirement for testimony. We will see if it can answer the attribution problem. I will call Clifford’s two cases the chemistry case and the Arctic ice case. Let’s take the chemistry case first. A chemist tells Clifford that it is possible to produce one substance by putting two other substances together. Clifford holds that he’s “quite justified in believing upon his authority, unless I know anything against his character or judgment” (L&E 187). Regarding the former, Clifford holds that he may rest easy that the chemist’s “professional training is one that tends to encourage veracity and the honest pursuit of truth” (L&E 187). For sure, this seems naïve to us, but let’s let it pass as Victorian enthusiasm. What’s important is how the knowledge condition is met. Here, Clifford invokes a special criterion: I have reasonable grounds for supposing he knows the truth of what he’s saying, I can be made to understand so much of the methods and processes of the science as makes it conceivable to me that, without ceasing to be a man, I might verify the statement. (L&E 188)

Clifford doesn’t have to verify the statement, he just needs to know that the statement was arrived at by what he recognizes as evidentially responsible means. He could, had he the time and inclination, do the experiment. So he may attribute knowledge to the attester, because “the verification is within the reach of human appliances and powers,” and these means had actually been used by his attester. The important thing, first, is that for Clifford to have reason to hold

62

Evidentialism and the Will to Believe

that what the chemist knows is not what Clifford himself must also know, but that he can see that the means for knowing are available and have been used by the attester. The means are public, but as it stands, the special competence of the chemist. So he is justified in attributing knowledge, even when he does not know. Now alter the case a little, with the same chemist, but this time, let the testimony be: an atom of oxygen has existed unaltered in weight and rate of vibration throughout all time (L&E 189). Now, this piece of information is within the topic of chemistry, broadly construed, but it is not the kind of truth that the methods of chemistry can establish. Clifford can’t learn some set of techniques and come to test this second claim. It is not the kind of truth we can have access to, given our means as finite human beings. Clifford, then, puts the point dramatically: “[I]t is a thing which he cannot know without ceasing to be a man” (L&E 189). The contrast between the chemist’s two claims, then, is that certain contents are not available for public testing. The first case is, but the second isn’t. Because we have no way to attribute the process of arriving at the chemist’s second claim to anybody, we cannot attribute knowledge to the chemist. But it is not merely because we don’t know it. Again, we didn’t know the first claim, but we can attribute its knowledge to the chemist, because we can endorse his methods. But this is not the case with the second claim, so not only we do not know it, it’s reasonable to say the chemist, even if he attested honestly and sincerely, does not know it, either. The Arctic ice case is roughly similar, but has one significant wrinkle. First, have an Arctic explorer tell us that it was as cold as x degrees on the pole, that the sea was y meters deep, and the ice was z meters thick at a specific location. Clifford is willing to accept the testimony. The evidence can be collected, the explorer’s companions can corroborate, and “we might, without ceasing to be men, go there and verify his statement” (L&E 190). Again, the crucial element is that we are able to attribute reliable means to the attesters and those reliable means are ones we ourselves could use, were we inclined. The evidence is public in the sense that were humans with the right training and resources to come upon it, they could see the same conclusions. Now, compare this Arctic ice case with the testimony of an old whaler, who holds that the ice at the pole is uniformly 300 feet thick all over the pole. The trouble is that this statement is verifiable by a person (or a large collection of them) in principle.22 The trouble is that the whaler himself is not in possession of the verifying evidence. He doesn’t have the tools, and he doesn’t have the manpower at his disposal to do that. Even though the means are available and

Reading Clifford’s “The Ethics of Belief ”

63

public, we have not only no reason to believe but also have reason not to believe that the whaler availed himself of the resources. So, even if the matter is within the reach of human knowledge, we must also determine whether it is within the reach of this particular human attester. At this point, we can state the attribution norm for the knowledge requirement, and then we can show how Clifford can answer the attribution problem. Given Clifford’s cases, I believe the conditions for satisfying the knowledge requirement and having evidence for attributing knowledge to attesters is as follows: Listener A may hold that attester B has satisfied the testimonial requirement of knowledge for testimony that p only if A has reason to believe that (i) B’s claim that p is the kind of content that has publicly verifiable evidence so that A could access this evidence and (ii) B’s claim that p is the result of B’s sufficient inquiry regarding p and access to sufficient evidence that p.

The point, now, is that for A to hold that B meets the knowledge condition for testimony that p, A needn’t also know that p. A must know that p is the sort of thing that people, given our cognitive resources, can know, and that B has access to those resources and has used them. So the Cliffordian solution to the knowledge-attribution puzzle with the phone number case earlier would be: Carrie knows that the question of people’s phone numbers a matter than can be publicly verified and determined. She knows, further, that Bill not only has access to but has accessed the evidence determining what his phone number is. So she may reasonably hold he passes the knowledge requirement for his phone number, even though she herself does not know.

So our listener would not have to, as Clifford might put it, “cease to be a human being” to test the claim. She would just need to call the number and see if Bill picks up. Simple. The point, here, has been to show that though Clifford’s approach to testimony is a demanding form of reductionism, it is not so demanding that it yields the attribution paradox. That would be very, very bad for Clifford’s view, because it would essentially yield skepticism not just with regard to the testimonial metaphysics in religious revelation (which was Clifford’s target), it would yield universal skepticism with regard to testimonial knowledge as well. That’s a case of proving too much, and Clifford’s view must be able to address it. I think, so far, I’ve shown that he can. This still leaves us with the worry that Clifford’s view is still very, very demanding, as it still has the requirement that we have access to the background conditions for the testimony we receive. But children don’t do that when they learn language; nor do we when we stop to ask for directions;

64

Evidentialism and the Will to Believe

nor does it seem we do it overtly. The question, again, is why must the testimony norm be so demanding in the first place?

5. The sacred tradition of humanity Clifford, given how demanding the requirements for believing on the basis of testimony are, must renew his case for why the norm must be so stringent. Recall that the slope arguments from before were arguments from our moral responsibility to care for the cognitive health of humankind. We have a rolerelated duty, as humans, to contribute properly to the shared common knowledge we all live our lives in light of. A similar hortatory case will be made here, but this time it is not in terms of what we are obliged not to do (namely, be credulous), but what we are obliged to do in the service of this shared cognitive inheritance. Clifford, again, uses his most forceful religious rhetoric: What shall we say of that authority, more venerable and august than any individual witness, the time-honored tradition of the human race? An atmosphere of beliefs and conceptions has been formed by the labors and struggles of our forefathers, which enables us to breathe amid the various and complex circumstance of our life. (L&E 190)

We are socialized into a cognitive culture, one that affords us the tools and resources to understand, research, know, and express ourselves. At first, we are totally dependent on it. This is clear. But the objective is to acculturate those coming into the culture so that they can do the same as we do: question, debate, and inquire. Children and their dependencies are not, then, counterexamples to the demandingness of the Clifford’s argument, but the reason why it must be so demanding. Children depend on, at first, the results of this tradition, but they slowly become participants in it, and that participation is not mere obedience to given norms, but engaging with, questioning, and challenging them. The norms of our cognitive culture are the tools we use to think: It is around and about us and within us; we cannot think except in the forms and the processes of thought it supplies. (L&E 190)

But one of the norms of cognitive culture, one of the norms that keeps it alive and worthy of our respect, is that we are constantly testing and challenging its norms. It, in fact, is our “bounden duty” to ask questions and investigate. It is a misuse of our cognitive tradition to be mere obedient parrots of its contents. Rather, we are enlivened by this tradition to perfect it by being critical of it. This

Reading Clifford’s “The Ethics of Belief ”

65

is all, of course, a standard Enlightenment line (we see a version of it in Kant’s “What is Enlightenment,” for example), but it is a consistent line Clifford takes, which is that we have a duty as humans to humanity’s cognitive improvement. And so we must shoulder the demanding norms of testimony, as to do otherwise would be: Refusing to do our part towards building up the fabric which shall be inherited by our children. (L&E 191)

The point, again, is that we have an obligation to create a cognitive culture that prizes the dignity of critical thinking. Clifford contrasts the ethic of this culture with those of some other cultures, specifically with the case of the African medicine man and the Australian aboriginal. To be sure, Clifford’s examples (like that of Mohammed discussed earlier) are chosen to strike the exact right note with the English gentlemen in his audience, which is simply: here, we have a backwards, nearly stone-age, culture, and the cases presented will part of the explanation they have remained so for so long. Clifford’s cases reek of Victorian colonial racism, and their tone smacks of the entitlement that comes from those enjoying the benefits of the Age of Empire. In short, yuck. This said, Clifford’s philosophical point can be made while bracketing the cultural imperialism in the background. The medicine man case is simply that a mysterious entity must be propitiated, or else the people in the village will lose their cattle. There is no independent means of telling whether the entity is really propitiated. If the cattle survive, we assume it must have been. If not, it wasn’t. These sorts of practices, those without independent means to check, are “founded on fraud” (L&E 191). The Australian aboriginal case is that even after being given hatchet handles with a notch for their axes, the aboriginals still prefer to tie their hatchets on the side. They are so wedded to traditions of doing things a certain way, they don’t see a better way, even when it is literally right before them. Now, the lesson from the two cases (again bracketing Clifford’s objectionable judgment of these cultures) is that there can be norms internal to a culture that are impediments to the culture’s improvement. Clifford’s prime examples are those of the habits of credulity in the face of practices that have no known ways to distinguish successful from unsuccessful performance and the slavish dependence on tradition. These are both cases where parts of a culture drag the rest of the culture down. The duty, then, of those who have inherited a cognitive culture and are now active practitioners of and thereby contributors to it is to perfect it. That is why the duty of questioning is so important.

66

Evidentialism and the Will to Believe [T]he rule… that is stored up and handed down to us is rightly used by those who act as the makers acted, when they stored it up; those who use it to ask further questions, to examine, to investigate; to try honestly and solemnly to find out what is the right way of looking at things. (L&E 196)

The rule, again, is that the aggregate testimony of our cognitive culture and traditions are “subject to the same conditions as any one of them,” its members (L&E 192). What Clifford calls the “sacred tradition of humanity,” then, is not a collection of traditions to ape or creeds to recite, but the practice of questioning. Any tradition worth its salt can stand up to critical questions and survive their scrutiny—that is what makes it worthy of our respect. Any creed that may appeal to us should do so because it is an appropriate answer to a question, one that can be and has been shown to outperform its competitors in being responsive to evidence. The sacredness of the precious deposit imposes upon us the duty and the responsibility of testing it, of purifying it to the utmost of our power. (L&E 168)

We do not take traditions and our standing views to stifle doubts, but rather our doubts are those things upon which our traditions and cultures show their worth and are improved. To prefer that we opt to dial down the requirement of scrupulous questioning is to damage the greater culture, to set it on a path of degradation. Clifford’s doctrine of the “sacred tradition of humanity,” that of critical scrutiny, is posed to answer the challenge that his requirements for believing on the basis of testimony are too demanding. Crucially, the intuition behind the child objection to Clifford (that children can’t perform the knowledge test) gets used for the requirement as we have a responsibility to their cognitive inheritances to be exceedingly critical. One component of this sacred tradition of humanity Clifford pauses to develop is that of critical ethical views. In the same way that our cognitive lives are given interpretive direction by our cognitive inheritance, our ethical lives are given actual direction by our ethical inheritance. In the moral world… it gives us the conceptions of right in general, of justice, of truth, of beneficence, and the like. (L&E 193)

This inheritance builds on a deeper moral sense we have, as Clifford holds, we have “definite instincts” to which our conceptions must answer. We have immediate moral inclinations toward decency, toward beneficence, and our

Reading Clifford’s “The Ethics of Belief ”

67

cognitive traditions give us direction to how to fulfill these inclinations. However, these rules are worthy of scrutiny. We want to be just, but how can we do that most effectively? We want to be generous, but what is the truly generous action? The second is a matter of inquiry, criticism. Clifford’s example is that of indiscriminately giving money to beggars. Tradition had been that generosity required that one do so, but as Clifford contests: [T]rue beneficence is that which helps a man to do the work which he is most fitted for, not what keeps and encourages him in his idleness; and that to neglect this distinction in the present is to prepare pauperism and misery for the future. (L&E 194)

Clifford’s point is that the social safety net is a permanent basement if it does not prepare those at the bottom for improvement or provide opportunities. Any safety net that perpetuates further dependence on it is not true beneficence, and it is a matter to be determined by questioning and inquiring whether one policy is better than another at this function. The consequence is that Clifford holds our “social inheritance” to have two components. On the one hand, there is, “the instinct of beneficence,” which directs us to benefit others. On the other hand, there is “the intellectual conception of beneficence,” which determines what conduct meets the instinct’s direction. The interplay between these two elements, in turn, strengthens and purifies our natural inclinations, but if the intellectual element is not used, we are left with a “mere code of regulations. Which cannot rightly be called morality at all” (L&E 194–5). There are then two important points to take away. First is that Clifford, if put briefly, makes good on the justification of moral judgment, which he saw earlier in danger of skeptical ruin at the opening of Section II. The justification of our inductive cognitive lives will be taken up later in Section III. The lingering question is whether this is a sufficient answer to the skeptical challenge, and I will return to this question shortly. The second take away from Clifford’s discussion of the moral realm is how he connects moral norms with epistemic norms. It is a kind of reversal from what we’d seen before. The original question was what the source of the categoricality of epistemic norms was, and the answer was ethics. Ironically enough, given the present turn in the argument, it seems that if epistemic norms are not properly functioning in our cognitive lives, then the ethical norms have no categoricality, either. They are, again, “mere codes and regulations.” And so, in a twist of sorts, moral normativity has a dependence on epistemic normativity, too. My view is that this reversal should not be surprising in light of Clifford’s thought, which has shown to have strands of holism and

68

Evidentialism and the Will to Believe

stoicism. A lesson of holist epistemology (and broader view of normativity) is that the norms and contents have mutual dependence for both their content and bindingness. We see this in Clifford’s content slope in Section I, and now we see it in the reciprocity of moral instinct and cognitive conceptions of duties. The stoics, too, had an integrated conception of the perfections of the soul. They were divisible in terms of the Epictetan topoi: the perfection of judgment, desire, and action, but these were merely focal features of our overall perfection, and they interpenetrate. One sees right so as to do what’s right, and the practice of doing what is right allows one to see more clearly. The final question is whether this account of the “moral world” is sufficient to answer the skeptical challenges that demanding evidentialist standards bring upon those committed to moral truths. Now, this challenge comes in two forms. The first is in the form we saw earlier as the self-defeat problem, which is that evidentialism is argued for on the basis of ethics, for which Clifford gives no hard evidence. The second is what we might call the Pyrrhonist challenge for any view, which is that the evidential demands of scrutiny are endless. The Pyrrhonist challenge entails the self-defeat challenge, as both are that Clifford’s reasons are insufficient. However, the two are distinct, as the self-defeat challenge is simply that Clifford has not proposed any positive theory of evidence in the moral sphere. The Pyrrhonist challenge is that there are no positive theories of evidence at all. I think it is clear that the self-defeat version of the skeptical challenge can be answered. Clifford has proposed a basic theory of what appropriate ethical reasons consist. So his view is consistent, as his argument does not, by its own standards, run afoul of the IER. We will revisit this point later, with James, as there’s a question whether James is running a version of the self-defeat argument in “The Will to Believe” or something else. I think it’s something else, for two reasons, but that’ll have to wait for Chapter 2, Section VII-2. Whether Clifford’s ethical theory can answer the Pyrrhonist’s challenge is a matter beyond the scope of this chapter. However, I will propose a brief, rough defense. Regress skepticism, whether it be in the form of asking if any empirical view is justified or upon what my moral rule is grounded, is a profoundly challenging philosophical problem. A standard strategy has been to propose foundations to these series of reasons, beliefs with special contents, or sources that allow them to provide reasons independently from support from other reasons. Alternately, the coherentist view has been that the basic unit of justification is the system of beliefs and the mutual support obtaining between the contents of that system. Clifford’s view is clearly a

Reading Clifford’s “The Ethics of Belief ”

69

form of coherentist holism, but notice that his backing requirements are never satisfied simply by the fact that the belief belongs to the system. This is also the source of error and credulity, too. What makes the system so that it can confer justification is that it is maintained internally so that there actually is no final answer, but an ethic of opening new questions and inquiry. And so the regress continues. This sounds like a reversion to skepticism, but it isn’t. The reason why it appears skeptical is that we’ve neglected a further option for the anti-skeptical reply to the Pyrrhonist’s challenge: epistemic infinitism. What it is to be a knower is to be open to answering all the critical questions that can be asked, and to take it upon oneself to answer them. The thing is that critical questions have no endpoint in principle. But that’s what knowers do—they keep investigating, keep looking for and ensuring their answers. A cognitive culture that provides its members with the desire and skill to pursue that project is the properly run culture. “The intellectual part of the heirloom (the sacred tradition of humanity) is to enable us to ask questions” (L&E 195). If Clifford is to have an answer to the Pyhrronist’s challenge, it will be in the form of a mixed epistemic theory—a coherentist infinitism.23

Section III—The limits of inference 1. A burnt child dreads the fire As we saw with Clifford’s justification of moral knowledge, there are two levels to the formation and maintenance of judgments. First, there is the native level of instinct, which gives our thoughts on the matter direction. In the moral case, it is our inclination to do good, be just, and so on. The second level is our collection of cognitive conceptions of how to fulfill those native impulses. This level is always being tested and refined with evidence. The proper formation of judgment is an interplay between the first and second levels with the first giving the second the ends, and the second not only providing the first with the means, but refined feedback as to what is to be good, just, and so on. A similar structural story will be told with regard to the second target for skeptical challenge for the evidentialist—namely, the assumption of the uniformity of nature. The skeptical challenge can be stated in the form of a dilemma. The uniformity of nature is a contingent fact—a world of sheer randomness is not a logical impossibility. We would know of the uniformity of

70

Evidentialism and the Will to Believe

nature either a posteriori, from empirical evidence, or a priori, from pure reason. If we were to know that nature is uniform from pure reason, we would know it as a fact of reason. But facts of reason are necessary truths, so we do not know the uniformity of nature a priori. If we know the uniformity from experience, then we would know it as a contingent truth. But what experience or experiences would serve as such evidence? Without assuming the uniformity of nature, no experience or aggregation of them could be evidence of the uniformity beyond them. We would have to assume that nature is uniform to see our uniform experiences of nature as evidence that the rest of nature not experienced (facts in the future, facts far away, and so on) are like those we have as evidence. Given the dilemma, we do not know that nature is uniform. Clifford’s strategy is to stick to the two-levels model of the sacred tradition of humanity and hold that the assumption of uniformity is, like our moral instincts, native, but nevertheless open to refinement. We have the deep, and non-optional, assumption that nature is uniform, but the nature of that uniformity is in question, and so we must collect empirical evidence to determine the details. So the two-levels story splits the difference, really: it’s a third horn to the dilemma that’s a form of taking both horns. Of course, the problem is that if two horns of a dilemma are sharp, then taking both isn’t much of a strategy of avoiding being skewered. But Clifford’s strategy, as we will see, is designed to mitigate the problems of the two forms of the dilemma. Clifford’s organizing case is that of the burnt child fears the fire, “because it believes that the fire will burn it to-day just as it did yesterday” (L&E 199). The child, in arriving at this belief relies on two pieces of evidence, but must rely on a deep principle about the world to arrive at the belief. The case proceeds as follows: Step 1: The child has two experiences: Exp1: The experience of seeing a fire now. Exp2: The memory of being burned by fire previously. Step 2: The child believes that the fire now will burn it, and so dreads it. Clifford identifies two principles connecting Step 1 to Step 2. The first is that Exp2 must be reliable—that memorial experiences are not only of, but reliable about, real events in the past. As Clifford says, “the belief that the child was burnt yesterday goes beyond the present experience, which contains only the memory of a burning, not the burning itself ” (L&E 199). So the child, to have evidence of the current fire burning from past fires burning, must have a principle of memorial evidence true of it.

Reading Clifford’s “The Ethics of Belief ”

71

Memorial Evidence (ME): S’s memories that p occurred are evidence that p actually occurred.

First, note that Clifford is a consistent reductionist with justification; in this case, memorial justification depends on our assessments, like with testimony, of its antecedent reliability. Importantly, Clifford observes, in order to be justified in holding ME, we must be aware of the fact that memories are generally reliable and that memories are alike in this rough way. But this commitment, then, relies on a deeper epistemic principle, namely that once Exp2 has been able to establish that there was a fire in the past that burned, that fire now will be similar to the fire in the past. To be justified in holding ME, we must be justified in holding that memories in the past being reliable is reason to believe memories now and in the future will be similar. The assumption, then, both for ME to be justified and for the move from facts about the past (justified on the basis of ME) to the present and future to be justified, we must be committed to an Assumption of the Uniformity of Nature (AUN), which allows the transition. Assumption of the Uniformity of Nature (AUN): Nature is uniform in a way that what we do not know is relevantly similar to what we know. The burnt child dreads the fire because she assumes the fire in the past is a guide to fires in the future, and that her memories of fires in the past are reliable guides as to what they really were like. Both rely on the AUN. Clifford supplements the burnt child case with two more illustrative examples: spectroscopes and history. The spectroscope, which measures the wavelengths of light, has shown that there is hydrogen in the sun. The evidence is that the sun’s light has a specific wavelength that is the same as that given off by hydrogen in the laboratory. On the assumption that hydrogen behaves the same way no matter where it is, the similarity of wavelengths of light is, then, evidence. A similar assumption is behind our capacity to reason about history. We know many things about the siege at Syracuse on the basis of Thucydides’s testimony in his history of the Peloponnesian War. Thucydides tells us about the siege and other historians tell us that he was a general in Athens and a later chronicler, so he would have accurate knowledge of the events. Clifford says: [M]en do not, as a rule, forge books and histories without a special motive, we assume that in this respect men in the past were like men in the present; and we observe in this case no special motive was present. (L&E 202)

The point, from the requirements of testimony, is that we have evidence of knowledge, evidence of trustworthiness, and so, assuming these historians are

72

Evidentialism and the Will to Believe

like the people we know now, they are reliable attesters. But, as we see in all of those cases, we must have the AUN applied to the uniformity of human nature in these cases. In short, for us to have evidence of any contingent fact beyond our present experiences, we must rely on the AUN. The AUN is what makes empirical evidence a reason to believe at all. Now, note, what evidence is gathered in these cases is further detail about that uniformity. Again, on the two-levels view, there is the AUN and then the collected evidence of what form this unity takes. And so the principles behind testimony are that humans will behave similarly, and what behavior (as we find out through the application of the first assumption) is that they are generally reliable. From this commitment we derive facts about history from their aggregate testimony. Clifford, then, restates his two-levels view: We may, then, add to our experience on the assumption of a uniformity in nature; we may fill in our picture of what is and has been, as experience gives it to us, in such a way as to make the whole consistent with this uniformity. (L&E 203)

And so, similar to the model for moral knowledge, Clifford has a native ordering commitment that directs and then is refined by its application in specific empirical judgments. But, one may reasonably ask, especially if Clifford has insisted that this is a non-skeptical view consistent with evidentialism, how is one justified in deploying and living one’s cognitive life in terms of that initial assumption. Again, Clifford’s evidentialism requires that we have evidence for the views we use, not just assume them. Is the AUN ever provided with evidence?

2. Regulative principles The skeptical challenge to Clifford’s evidentialism was that the AUN is behind all our inductive inferences, but it itself does not seem to have any evidence in its favor. Clifford’s response has been that it is behind all our inductive inferences. So far, that’s not much of a response to the skeptic, as those who worry that global skepticism would follow if the AUN were not justified are not made less concerned by more examples of how central the AUN is to the proper functioning of a cognitive system. And making the terminological choice of calling it an “assumption” doesn’t really help matters. What positive reason is there to hold that the AUN is true? Calling it an assumption highlights the fact that we don’t have reasons supporting it.

Reading Clifford’s “The Ethics of Belief ”

73

Clifford’s defense of the commitment to and use of the AUN proceeds on two connected fronts. The first is what I will call the non-optionality argument— basically that the AUN is a necessary component of any set of beliefs if they are to be systematic at all. Consequently, there can be no reasons to ever count against the AUN. That’s something. The second is what I will call the practical justification of the AUN, which is that its application is restricted to the facts that concern us, ones relevant to our practical lives, and that any justification for AUN going beyond those bounds is not only hasty, but also unjustified. The non-optionality argument is that, as noted earlier, without the AUN, there are no ways any experience can be taken as evidence for anything else beyond the fact that experiences are currently being undergone. If this is the case, then there can be no experiences that can be evidence against the AUN, as the AUN must be deployed for those experiences to be evidence in the first place. That is, if one were to have some experience that provides evidence that the AUN is false, we must also use the AUN to interpret that experience as evidence. And so, there can never be empirical evidence that the AUN is false. No evidence … can justify us in believing the truth of a statement which is contrary to, or outside of, the uniformity of nature. (L&E 204)

Instead, we should conclude that we’ve received distorted information or that there is a uniformity about which we’ve been ignorant (that is, it is the kind of uniformity we were mistaken about, not that there’s a uniformity). Consequently, even if an event contrary to the AUN were to occur, we would nevertheless be in a position where we would (a) have no right to believe it, except when we had an experience like what had happened, and (b) make no further inferences from our belief that we’d had that experience, if we held the AUN thereby false. The dilemma here seems massively overstated. Surely those who hold that there are miracles, say, don’t hold that the AUN is wholly false, only that there are exceptions to it. Those are the miracles. The AUN is still reliable, and we use it to identify miracles. Clifford’s case is that believing one has a case of nature’s nonuniformity is like a poison pill that destroys one’s whole cognitive system. But surely in order to be committed to miracles, one needs to be committed to some weaker form of the AUN (else, one couldn’t, again, recognize miracles as miracles, as they’d all be random business). The principle might be: as a general rule, nature is uniform so as the past is a reliable guide to the future. One’s being committed to this rule still allows induction, but it also allows us to be able to recognize the miraculous exceptions. That is, if one does not accept the universality of the AUN, one does not thereby accept the view that none

74

Evidentialism and the Will to Believe

of the universe is uniform. Such is Clifford’s error of contraries—he takes the denial of All of nature is uniform to be None of nature is uniform. But this is a fallacy of illicit contrariety. Just because not all, it doesn’t follow none. And again, the Clifford dilemma, what I’ve called the poison pill, works only if you’ve accepted none. So Clifford’s reductio here seems more like heavy breathing than demonstration. But I think there’s still something to be salvaged, which is a more dialectical point than demonstrative. Recall that Clifford is situated between two dialectical opponents: those who argue from miracles and revealed theological traditions, and those who are worried about the skeptical consequences of his evidentialism, particularly the IER as applied to the AUN. Presumably, those who hold to the trustworthiness of miracle testimony and religious traditions are also anti-skeptics about moral and inductive reasoning. But it is here that Clifford’s non-optionality argument has its dialectical bite as the point becomes that if one relies on the AUN, even if as a general rule, one is put in a difficult position in being able to reliably identify the exceptions. How does one sort the exceptions to the AUN from the mistaken reports, distorted recollections, and surprising uniformities discovered? In short, one cannot. The consequence, then, is that insofar as one makes inferences from experiences at all, one is committed to the AUN. And insofar as one is committed (even modestly) to the AUN, one has no recognizable evidence for miracles or to believe on the basis of revelation. That is, to distinguish one case from the other, one must rely on evidence, and to do that, one would nevertheless be relying on the AUN. So even with miraclecases, the AUN needs to be invoked. So Clifford’s dialectical case is that the AUN is a non-optional commitment for those who hold they have evidence. But how does this satisfy the skeptic? He’s shown that one can’t have evidence against the AUN, but remember the IER requires that he have evidence for it. So far, all that’s been shown is that the AUN is very important and that global skepticism follows if we don’t accept it. But that’s not a justification of it, but rather a restatement of the skeptical challenge. It’s at this point that Clifford’s game must change. Instead of proceeding dialectically, he proposes two modifications: a reconception of what the evidence for the AUN could be and a correlate reconception of what the AUN really is. These reconceptions are in terms of the practical consequences of the AUN’s application. And so, Clifford offers what should be considered a practical justification of the AUN. First, Clifford identifies what the positive case for the AUN would comprise. In effect, it is that the AUN is a rule of rational practice:

Reading Clifford’s “The Ethics of Belief ”

75

And practically demonstrative inference—that which gives us a right to believe in the result of it—is clear showing that in no other way than by the truth of its result can be the uniformity of nature be saved. (L&E 203)

The form of practically demonstrative evidence, then, is that: If a hypothesis (h) is held and employed by S, then if its applications to inference are true (or generally true), then S has reason to believe that h.

Notice that once we put the account this way, Clifford’s case for the AUN is clearly a reversion to a question-begging inductive defense, as the fact that the AUN’s having yielded truths in the past is relevant to its performance in the future only on the assumption of the AUN itself. It is here that the second stage of Clifford’s anti-skeptical case must be made clear, as Clifford’s first stage was to specify what kind of evidence counts for the AUN. Clifford’s case specifies what the assumption is and how we hold it as an assumption. Given that the AUN is a non-optional element of having systematic belief, and given the two-levels line, a native commitment that organizes and directs inference, it is a regulative norm of thought. It is how we ought to reason. That’s rationality. But do we, thereby, have reason to hold that the view is true? Clifford demurs: Are we bound to believe that nature is absolutely and universally uniform? Certainly not; we have no reason to believe anything of this kind. (L&E 204)

We have no reason to believe the AUN is true of what it is about. The first question is whether this is also a poison pill thought, given how Clifford plays the game with his opponents. Despite the fact that the thought is native, nonoptional for systematic knowledge, and can have no evidence brought against it, this principle, it seems clear, has no evidence for its truth. Clifford, really, is a kind of mitigated skeptic, then. He restricts what inference can justify by way of the AUN—namely, only empirical hypotheses. Consequently, the AUN is a kind of regulative principle or rule within a range of questions and inquiries. The rule, as Clifford holds, “only tells us that in forming beliefs that go beyond experience, we may make the assumption that nature is uniform as far as we are concerned” (L&E 204 emphasis added). This is to say that the realm of application for the AUN is only to the “range of human action and verification,” the empirical and practical realm. The AUN tells us nothing beyond how we ought to run our cognitive lives, nothing about the world itself. It is, as an assumption, a rule of thought or norm, not a hypothesis about what is true of the world. As a consequence, we test it as a norm of thought in terms of whether its employment

76

Evidentialism and the Will to Believe

outperforms the employment of other norms (or our refusal to employ it). This is a different kind of test from whether it is true, but more a test as to whether the rule is rational, and perhaps, more importantly, whether we can see its use as in the service of our moral lives. Hints of a pragmatic perspective in Clifford’s views have already been noted, and it’s worth highlighting this “practically demonstrative inference” here. Clifford’s case for accepting the AUN is an early instantiation of what’s now called the pragmatic justification of induction. In essence, when Clifford holds that the assumption can be “saved,” he holds that he can provide reasons for the AUN, but just not reasons for the AUN’s truth. And so the practical inference is less evidence for the rule, but a kind of vindication of making use of it. Two questions immediately come to mind at this stage. First is whether the pragmatic case is just a matter of moving the goalposts for what counts as a justification. In essence, what the moving the goalposts objection amounts to is that Clifford needs to justify the AUN, and in looking to see what can be said for its favor, he then finds the threshold for justificatory success. As if to say: I have this much reason to hold that p … and lucky, lucky me… that’s just how much reason I need to hold that p.

The trouble is that the goalposts objection highlights Clifford’s case as one of special pleading for a favored principle. And this trouble leads us to the second objection. Notice that Clifford’s case for the AUN is from its practical payoff, not from the evidence that it is true. This seems an utterly delicious irony. In fact, it seems rank hypocrisy. Peter van Inwagen’s case against Clifford is that the IER proves too much, and as a consequence should render Clifford a skeptic. But Clifford is not a skeptic, as he makes exceptions, or as I’ve argued, special pleaded, for moral principles and the AUN. The case Clifford is making against religious beliefs, either that of beliefs on faith, given by revelation, or about miracles, is that it runs afoul of the IER. But why does Clifford get to criticize those who make exceptions to the IER for their viewpoints and important purposes when he does exactly the same thing himself? Peter van Inwagen then poses a dilemma for the Cliffordian: [D]o you think that other important categories of belief stand convicted of similar epistemic impropriety under Clifford’s Principle…? If you do, are you a skeptic as regards those categories of belief…? If not, why not? If you think that the only important category of belief that stands convicted of impropriety under Clifford’s principle is religious belief,… how do you defend this position? (1996: 151)

Reading Clifford’s “The Ethics of Belief ”

77

The trouble, as van Inwagen sees it, is that Clifford proceeds according to a “double standard” (1996: 150), and we can see it clearly in his special pleading for the AUN but treating religious belief as without the same kind of consideration. If Clifford is willing to move the goalposts for the AUN, then why not for religious belief? There can be three points for the Cliffordian here. Two have some promise, one, not so much. First, note that the AUN is weakened at the end of Clifford’s defense of it. If defenders of religion receive the correlately weakened version of theism, they are likely to be disappointed. This will be the central thesis of my critical commentary on James, but the point is Clifford changes the content of the AUN drastically to save it. It still serves the function it did before, given its status as a regulative principle, instead of a proposition known as true. The question would be: could religion really be saved with this tactic? I am inclined to think not. Further, there are two related topics bearing on the double standard problem. On the one hand, the Cliffordian answer might be to say that the AUN is a native commitment and it is central to the “sacred tradition of humanity.” We have to ensure that it is allowable. Religious belief is not native, and it has an inconsistent track record of contribution to humanity’s dignity. Timothy Madigan takes this line of defense: It is because these are the types of beliefs that tend to have the greatest amount of repercussions to society as a whole…. Clifford was fighting against the hegemony of organized religion. (2010: 146)

Trouble is this strategy is not much of a defense against the charge that Clifford begs the question and moves the goalposts. The answer is exactly the sort of thing someone who’s got the view that religion needs a higher degree of scrutiny and a practical justification of induction would say, but this is not a justification. Instead, it is a restatement of the view, which, given the dialectical state of play, is not sufficient. Moreover, doesn’t the AUN have even a greater amount of repercussions to society than pretty much any other commitment, anyway? On the other hand, the Cliffordian needn’t concede that there is a double standard. Religious commitments are optional, as the Enlightenment project has shown. The AUN isn’t.24 Religious commitments are accepted on the basis of testimony we cannot have grounds to trust and we are committed to facts we cannot come to verify. The AUN and our moral principles are accepted as methodological rules for well-run cognitive culture and are provided with refinement and clarification every day. It is, then, not just the sources and

78

Evidentialism and the Will to Believe

contents of religious life that makes it a legitimate target. Rather it’s that it cannot contribute to the proper functioning of a cognitive culture.

3. Three norms Clifford closes “The Ethics of Belief ” with restatements of his three core cognitive norms. The first, what I’ve called the Assumption of the Uniformity of Nature (AUN): We may believe what goes beyond our experience, only when it is inferred from that experience by the assumption that what we do not know is like what we know. (L&E 204–5)

And so Clifford’s restriction is that one may have justification for belief in contingent general truths or truths about the past, future, or present beyond one’s immediate experience, only on the basis of induction from experience. And thereby, one relies on the AUN. The second norm is that of the reductionist requirement for testimony: We may believe the statement of another person when there is a reasonable ground for supposing that he knows the matter of which he speaks, and that he is speaking the truth so far as he knows it.

Clifford’s requirement for taking attesters’ testimony as evidence is one that makes testimonial evidence dependent on other factors we must establish about the attesters, namely, that they’ve satisfied the knowledge and sincerity conditions. Clifford, of course, sees the testimony norm as dependent on the AUN, as we must be justified in taking facts about attesters in the past to be good rules for present and future (and other past, with historical sources) attesters. Why we would care about either of these rules for evidence, again, is a matter derived from the moral rule that one ought not pointlessly endanger others, and so one must care for one’s cognitive life in a way that is attentive to evidence, both what one currently has and what one should have. And so, Clifford proposes an evidentialist norm that brings together the synchronic and diachronic elements of responsible (if not merely non-negligent) cognitive life, the IER: It is wrong in all cases to believe on insufficient evidence; and where it is presumption to doubt and investigate, there is worse than presumption to believe. (L&E 205)

2

Reading William James’s “The Will to Believe”

William James and “The Will to Believe” William James was a conflicted thinker. Even up to his death in 1910, he struggled to square the Romantic aspirations of finding meaning and transcendence with the scientific worldview and its capacity for explanation. He was born to Henry James (Sr.) in 1842, who had hoped his son would be a scientist and student of life. William, on the other hand, preferred art. But when the time to decide on colleges and training arrived in 1861, William faced a choice: enroll in the Lawrence Scientific School at Harvard or enlist in the army for the Civil War. He chose science at Harvard. At Lawrence Scientific School, James learned of the theory of evolution, On the Origin of Species having been published only two years before he arrived. Before he went to medical school (at Harvard), he participated in a scientific expedition to Brazil in 1865 with the famous biologist Louis Agassiz. The trip was disastrous for James. He did not enjoy the work associated with the scientific study of the plants and animals, and he held Agassiz’s views that biology was a form of revelation to be ridiculous. On top of this all, he fell very ill with smallpox, which physically depleted him to the point where his back constantly hurt. He, afterward, suffered regularly from bouts of depression. When he came back from the trip, James was still suffering from smallpox, and he found Cambridge uncomfortable. He travelled to Germany to find a better environment and to continue his studies. He found little relief from his symptoms there, but in Heidelberg, he availed himself of the famous scientists there. James returned to the States to receive his degree of Doctor of Medicine in 1869, but he had no interest in practicing. From the stress of trying to decide what to do with himself and from despair at the fact that his health was not improving, he suffered a nervous breakdown in 1870. He returned to Harvard,

80

Evidentialism and the Will to Believe

later, to teach physiology in 1872, and then psychology in 1875. It was in his outside activities, however, that James was intellectually stirred. His interests had turned to philosophy, specifically sparked by his engagement with the members of the Metaphysical Club, a group of intellectuals who met regularly to discuss the virtues of the empiricist and rationalist approach to philosophical questions. The group included Oliver Wendell Holmes Jr., who would later become a Supreme Court Justice; Charles S. Peirce, a logician, mathematician, and philosopher; John Fisk, a historian of philosophy; and the scientist and philosopher, Chauncey Wright. James’s interest in philosophy and psychology blossomed. He brought his knowledge of science and psychology to bear on the philosophical issues before the group to momentous results. One early case of James’s broad intellectual skills yielding results in philosophy was his 1881 address as a physiologist to the Unitarian Minister’s Institute at Princeton, “Reflex Action and Theism.” In it, he argues, from his scientific findings of the teleological structure of thought (which he later makes explicit in his Principles of Psychology) that theism is “the most practically rational” option between itself, gnosticism, atheism, and agnosticism (WB 101). “The Will to Believe” is the later inheritor of the line of thought that the demands of a healthy soul and properly functioning mind are ones that incline toward theism, even when the evidence is weak or nonexistent. Philosophy springs from our natural reaction to uncertainty, to the need for meaning, and James’s pragmatism is posited on ensuring that human thought be weighed in terms of how well it responds to these challenges. Religion is central to this aspect of human life, and James’s psychologically informed theory of religion was designed to capture the appeal of the piety that he had seen as the highest aspiration of religion. “The Will to Believe” was an essay devoted to making the case for a more lenient epistemological evaluation of religious belief, specifically on the basis of the fact that such belief is an expression of our deeper psychological tendencies to see meaning and to want to contribute to it. James followed “The Will to Believe” with his Gifford Lectures, The Varieties of Religious Experience, which developed that background account of how and why religious views arise. The broader Jamesian commitment to pragmatism, that the quality of a view, whether it is worthy of our assent or not, depends on the benefits or costs we incur in believing it, what difference it makes in our lives, is a development of these psychological views on religious belief. What James calls healthy religious belief in Varieties of Religious Experience is not just that the world is suffused with goodness, but that one feels impelled to contribute to it. James finds in his later work a way to apply this criterion to many other philosophical views,

Reading William James’s “The Will to Believe”

81

from the commitment to free will, to pluralism, to the rationalist–empiricist debates in epistemology. His 1906 Lowell lectures (published as Pragmatism in 1907) are the culmination of this line of thought, and in them, he also revisits religious belief with the same heuristic. The instigating element of James’s broader philosophical views was his psychological view of and pragmatized case for religious belief. “The Will to Believe” is James’s starkest and most influential statement of this commitment. “The Will to Believe” and all the works James collected in The Will to Believe and Other Essays in Popular Philosophy were public lectures. They were prepared for an educated, but not necessarily scholarly, group. In the Preface, James is careful to define this audience. There is the “miscellaneous popular crowd,” and then there is the “academic” audience. The former, James concedes, do not need to be encouraged toward religious belief. Instead, their faith needs “to be broken up and ventilated” with a dose of science, something to make it such that they understood their own faith as in the midst of a real debate, as something that really can be challenged. The latter is already “fed on science,” and so they have a different need. They need the courage to overcome their “mental weakness,” brought about by the thought that patiently waiting for more scientific evidence will settle the issue of religious belief. He proposes to “preach” the right of the individual to indulge his personal faith at his own personal risk (WB 7–8). So James’s preferred audience is scientifically literate, philosophically inclined, and likely religiously neutral. His objective is to make a case for religious toleration on the basis of the thought that “religious fermentation is always a symptom of the intellectual vigor of a society…. The most interesting and valuable things about a man are his overbeliefs” (WB 9).

Preamble On one side of the ethics of belief debates are the evidentialists, who hold that it is inappropriate to believe without sufficient evidence. On the other side, there are those who hold that there are legitimate cases of belief without evidence. William James argues in “The Will to Believe” that one may believe ahead of one’s evidence, and so the essay is posed as a defense of religious faith. What is unique about James’s defense is that the justification of faith is to be found in its practical consequences. For James, the will-to-believe is part of successful living; it is a praiseworthy practical comportment, and so it is a legitimate case of believing.

82

Evidentialism and the Will to Believe

“The Will to Believe” was originally an address to the Philosophical Clubs of Yale and Brown Universities. James opens by contrasting the “good old orthodox College” views about religious matters at those two universities with the “freethinking … indifference” to such matters at his home institution, Harvard. This contrast is full of portent. Since its founding, Yale had a tradition of religious education, especially emphasized in contrast to the liberalism of Harvard. Brown was on the more liberal side, as it had no religious tests for student attendance. Brown’s charter had been designed to protect the liberty of conscience for all students. Yale, on the other hand, had no explicit faith test, but nevertheless restricted what religious services students could attend.1 This said, the religious openness was explicitly one directed toward other Christian sects. James makes it clear that there is a picture of the religious climate at the two universities with his humorous references to Leslie Stephen’s biography of Fitzjames Stephen. Those at Harvard have an image of how religion is discussed at these universities, and it looks like the kind of instruction Fitzjames underwent as a child: The teacher, a certain Mr. Guest, used to converse with his pupils in this wise: “Gurney, what’s the difference between justification and sanctification?- Stephen, prove the omnipotence of God!” (WB 13)

With this humorous image, James is not just drawing the contrast between Harvard and these colleges, but painting a picture of a kind of presumption held by freethinkers. The presumption is that religious upbringing and education must be authoritarian and dogmatic. Fitzjames Stephen, of course, was one of the outspoken agnostics in the Metaphysical Society. In fact, he was one of the most active speakers in the society, having given seven talks (equaled only by R. H. Hutton, and Clifford having given only three) (Brown 1947: 338). This kind of education pictured by the story of Mr. Guest, not surprisingly, alienated Fitzjames from the church and from the faith professed in it, who reported to his brother that he and the students were well taught and well fed, despite the “excess of evangelical theology” (Stephen 1895: 73). But with the likely laughter the story provoked at Yale and Brown at the image of the dogmatic barkings of a schoolteacher insisting on instant, rote, replies to his queries, James was out to show that he knows better. He says that at Harvard, “we are prone to imagine” that the educational environment and dialogue at religiously affiliated institutions “continues to be somewhat upon this order” (WB 13). Another laugh, surely. The contrast, then, is twofold. First, there is the contrast between believers and the freethinkers. Second, there is the contrast between what the freethinkers think of the believers and what the believers really are like.

Reading William James’s “The Will to Believe”

83

Now the importance of this second contrast is useful, as James follows it up by noting that his lecture will show that those at Harvard “have not lost all interest in these vital subjects” (WB 13). The point is that despite the fact that the picture of Mr. Guest’s demeanor is wrong about the way religion is discussed by his audience, the issues themselves are of genuine importance to the Yale and Brown students and to some of Harvard’s intellectuals, nonetheless. That is, the question of justification and sanctification, God’s attributes, and other such items for theological reflection are still matters of importance and genuine concern for the religious believers and others interested. The interested parties just don’t proceed with Mr. Guest’s dogmatic zeal. Consequently, the orthodox conversation in James’s usage extends to the theological matters bearing on the conditions for salvation (namely, whether one is sanctified in one’s works being in Gods service, or one is justified in receiving God’s grace) and the nature of divine attributes (for example, whether God’s omnipotence is consistent with his omniscience, and so on). Again, returning to the first contrast, the Harvard students, once they’ve been imbued with what James calls “the logical spirit,” reject the “lawfulness” of religious faith and the intellectual matters attendant to it. Hopefully, the reaction will be different with his present audience, as James promises to make a case for faith. I have brought with me to-night something like a sermon on the justification by faith to read to you,—I mean a justification of faith, a defense of our right to adopt a believing attitude in religious matters, in spite of the fact that our merely logical intellect may not have been coerced. (WB: 13)

A case for faith, given James’s contrasts, then, must satisfy two desiderata. First, it must survive the contrast between the freethinking indifferent attitudes of his Harvard students and those of his audience. That is, whatever beliefs and commitments constituting faith in “religious matters” must be substantially different from those held by freethinkers and humanists. They must be about or relevant to the recognizably religious matters James references as worthy of the appellation “these vital subjects,” beliefs that reflect the importance of the “difference between justification and sanctification and the omnipotence of God.” Let us call this, following James’s terminology, the orthodoxy requirement. The second desideratum of the case for faith is that it be a defense of a belief meeting the orthodoxy requirement from the objection that such beliefs are held without sufficient backing, and so are improperly held. That is, the Harvard students, once they’ve been “imbued with the logical spirit,” will hold

84

Evidentialism and the Will to Believe

that religious belief does not pass muster. The objection occasioned by the evidentialist’s cognitive scruples, in Clifford’s vein among others, is that because these beliefs are not supported by evidence, they are contrary to our duties of belief management. One may answer this objection either by (a) constructing a natural theology to demonstrate the sufficiency of evidence for religious beliefs, or (b) making a case that despite the fact that these beliefs do not have sufficient evidence, they are nevertheless lawful. Taking the first route is that of showing that one has justified religious beliefs, not faith. James, as we will see, holds that a defense of faith can concede that there is insufficient evidential backing. So a case for lawful exceptions to the requirement of evidential backing must be made. James makes the case with the doctrine of the will-to-believe later, that one may decide an intellectually undecidable matter by one’s passional nature. Let us call this the lawfulness requirement. Any evaluation of James’s essay must answer whether and how James’s justification of faith meets these two requirements. If he fails either, the justification fails. That is, James is out to make the case for faith using a series of counterexamples to the rule used by those who reject religious belief. The counterexamples lead up from simple cases of making friends to the more significant cases of having religious belief. But the argument requires that the counterexamples lead up to religious beliefs that are both recognizably orthodox and lawful. Ultimately, I hold that James, in fact, fails both requirements. First, James fails the lawfulness requirement. James’s case against the evidentialist is that there are exceptions to the rule that one may believe only when one has sufficient evidence. The trouble is that his cases do not show that evidence backing is unnecessary. Rather they expand the notion of evidence that provides backing. And so James does show lawful cases of belief (e.g., in having moral confidence in one’s actions), but these cases are so because evidence actually backs them. Specifically, the beliefs that are James’s purported exceptions to the rule of evidential backing are ones where having the belief makes the belief ’s truth more probable, but in that case, having the belief is having evidence. Beliefs with this property will be called doxastically efficacious,2 and are not counterexamples to the requirements of evidence, but rather are consistent with them. A further problem emerges if we take James’s case from doxastically efficacious belief to extend to the objects of religious faith. Again, doxastically efficacious beliefs are those beliefs that, when a subject accepts those contents, that subject contributes to their truth. But in what sense can we say that the contents of religious faith, those of belief in the existence of God, the afterlife, and so on, are doxastically

Reading William James’s “The Will to Believe”

85

affected? This occasions a second problem for James’s case—whether the faith justified is still the faith of the faithful. James’s case fails the orthodoxy requirement. In response to the looming trouble of holding faith is doxastically efficacious, James redefines the religious hypotheses as conjoined commitments that “the best things are the more eternal things,” and that we are “better off ” if we believe that affirmation (WB 29–30). This allows faith to be doxastically efficacious, but the beliefs have been changed. The beliefs that survive the tests of the will-to-believe are no longer recognizably and heartily religious. Rather, they are more the weak tea of Harvard freethinkers. In short, the beliefs James champions are not those one would call the objects of “good old orthodox College conversation,” but a radical reconstruction of those beliefs. This occasions a dilemma for James: in short, the commitments constituting faith either (i) satisfy lawfulness by being doxastically efficacious, but therefore are unorthodox, or (ii) the beliefs are orthodox and not doxastically efficacious, but are thereby not, given the argument, lawful. (Making good on how this dilemma is destructive will have to wait to Section X-6.) James’s justification of faith in “The Will to Believe,” then, is a failure, at least by the standards his introduction sets forth. That said, “The Will to Believe” is successful in providing a template for subsequent pragmatist accounts of doxastic efficacy and reconstructions of religious views in line with naturalist and humanist commitments. My view here is that preambles set the interpretive defaults with the way they contextualize the project and set the stage. James’s contrasts, I hold, set his goals such that his ultimate view should be different from the Harvard freethinker and more like the religious believer who holds that the issue of the means of salvation and God’s attributes are “vital matters.” If James really is going to provide a defense of faith, then given those whom he is addressing and given those whom he is holding himself in contrast to (the agnostics and atheists), he should have a kind of view that must be recognizably in the play of the dialectic. If he proposes a view that’s not recognizably in the game or proposes a view as faith that the faithful can’t recognize as their own, then he’s made an error. Or at least he’s guilty of false advertising. I’ll call this the old switcheroo problem. Second, whatever the view is that he’s defended; it’s got to be also recognizably acceptable from the perspective of our cognitive scruples. It’s got to be the kind of view that the evidentialists (perhaps using Clifford’s IER) would not accept, but nevertheless is clearly acceptable. So there are two ways to fail this requirement. On the one hand, the proposed commitment may be a form of overbelief that is rejected by

86

Evidentialism and the Will to Believe

the evidentialist’s rule, but it is not recognizably acceptable. I think that some of James’s cases fail this way. On the other hand, the proposed commitment may be a form of belief that is recognizably acceptable, but it is perfectly consistent with the evidentialist’s rule. I think the rest of James’s cases fail this way. Much of the discussion of these elements will have to wait for Sections IX-5 and X-6, but for now, we can see that James’s objectives are clear and what the desiderata for successfully achieving them are.

Section I—Hypotheses and options 1. Introduction and definitions James’s ultimate case against the view held by those “well-imbued with the logical spirit” that only one’s evidence should determine one’s beliefs is by way of counterexamples, or perhaps better put, acceptable exceptions. In order to properly present them, James must set the stage. This requires the arrangement of “some technical distinctions,” as he puts it, to understand the force of the coming cases and to then deploy them in the justification of faith. James begins the stage-setting here in Section I, but he, by his own admission, will not be finished arranging things until the opening of Section VIII. That’s seven sections of what James calls “introduction” (WB 25). To be fair, the argument has a steep hill to climb, so the arrangement of tools is important. Let me, however, raise one methodological worry to begin, which can be captured with an axiom from rock music: If a singer must tell a story before singing a song for us to appreciate the song, the longer the story, the more likely the song stinks. Less talkin’ more rockin’

Now, the point is not to say that, by analogy, James’s argument will stink. Rather, that if a song’s quality is dependent on a very particular background set of commitments, then it likely doesn’t stand on its own intrinsic merits. The same can go for arguments—if they rely on a very lengthy setup, they are likely only as good as the setup can make them. I will call this the talking-to-rocking problem, and given James’s lengthy and detailed setup for the argument, I will take his setup seriously. So instead of buzzing through James’s introductory material and granting him the distinctions and then turning critical attention to the argument, critical attention must be turned on the introduction too. My view is that it is in Sections I–VII that James does most of the heavy lifting for the argument, and once we get to Sections VIII–X, the board is arranged so that his case can work more smoothly than most evidentialists would allow.

Reading William James’s “The Will to Believe”

87

James’s counterexamples to evidentialism will be restricted to the class of commitments comprising what James calls genuine options. The importance of this restriction will be made clear later in Sections VIII and X, but for now, the case for the lawfulness of beliefs that run ahead of evidence, what I’ve taken to calling overbeliefs, will be restricted to those that are ones arising as responses to options of belief that are genuine. The definition of genuine options proceeds according to the following cascade of definitional schemata: A proposition (p) is a hypothesis iff p is something that can be proposed for belief. Proposition p is a live hypothesis for a subject (S) iff p has some appeal to S. Proposition p is a dead hypothesis for S iff p has no appeal for S (or perhaps: p’s appeal is below some minimal threshold). O is an option for S iff O is a decision for S between two hypotheses (p, q). Schematized: O:(p, q) O:(p, q) is a live option for S iff both p and q are live hypotheses for S O:(p, q) is a forced option for S iff (i) p and q are mutually exclusive for S, and (ii) p and q (and not both) are the only alternatives—there are no third hypotheses (r) for S. O:(p, q) is a momentous option for S iff (i) one of the hypotheses are unique and significant, and (ii) committing to p or q is not reversible. O:(p, q) is a trivial option for S iff the hypotheses are (i) not unique, (ii) not significant, or (iii) not reversible for S. O:(p, q) is a genuine option for S iff O:(p, q) is live, forced, and momentous for S. For most who are familiar with “The Will to Believe,” this definitional cascade is likely the most familiar part of the essay, perhaps the least controversial, too. But remember the talkin’-to-rockin’ rule. There are a few places for critical pushback.

2. Live and dead hypotheses James’s definition of a live option is one between two live hypotheses, and live hypotheses are those that are “real possibilit(ies)” for those who consider them. Importantly, the liveness of hypotheses is indexed to the subject for whom the hypotheses are posed as options. James makes this clear by the contrast between the liveness of some hypotheses for the New England Christians in his addressed audience and that for “an Arab.” His first case is whether one “believe in the

88

Evidentialism and the Will to Believe

Mahdi,” or the Twelfth Imam, destined to redeem the world before the Day of Judgment. His presumption is that his audience members feel no “electric connection” to it. It does not “scintillate with any credibility at all.” As such, for his audiences at Yale and Brown, the hypothesis is dead. For “an Arab,” even if this person is not a follower of the Mahdi, it is “among the mind’s possibilities” (WB 14). Two points are to be taken from this. The first is one James makes explicitly: “deadness and liveness in a hypothesis are not intrinsic properties, but relations to the individual thinker” (WB 14). Liveness and deadness of a hypothesis is always that for a subject. The second point is that liveness is possible even for hypotheses we ultimately reject. James’s hypothesized Arab may not be a follower of the Mahdi, and so reject believing in him, but the hypothesis is nevertheless live for him. We might say that the difference between live-butnot-held hypotheses and dead hypotheses is that the former are actively rejected while the latter are merely ignored. James holds that there are determining factors for the live–dead determinations. He says that the determining factor is “our nature”; that is, whether the hypothesis makes an “electric connection” with it. It may be easy to misread the line of argument here that the Arab’s nature is different in a way that makes believing in the Mahdi appealing, but his audience’s nature is not. Might that nature be race? For sure, we see in James the similar attitude of cultural smugness we’d seen in Clifford, but we needn’t attribute any of Clifford’s overt bigotry to James on this matter. For sure, James’s contrast relies on the alien otherness of the Orient, but it needn’t imply too much beyond the difference. We might say that our natures are conceived in terms of our dispositions and available capacities to act. And so the Arab, perhaps in Cairo, may have an inclination to be curious about the Mahdi in a way that the average undergraduate in New Haven wouldn’t. Their educational backgrounds, perhaps their own native comportment, and what cultural options are available are determinants of what they might act on. Let’s say neither come out to believe in the Mahdi, but only our Arab in Cairo has had to wrestle with the issue, deliberate about it. Our New Haven Protestant undergraduate might be aware of the option of believing in the Mahdi in only passing attention. So liveness is a function of how the hypothesis fits with the subject’s range of actions, which are themselves dependent on the subject’s history, temperament, and cultural milieu. Let’s take that to be the determining factors of what James calls “nature.” James makes good on the indexing to subjects by noting that for the average New Haven undergraduate, it is, by these factors, a dead option between Islam and theosophy (neither being live). On the other hand:

Reading William James’s “The Will to Believe”

89

[I]f I say “Be an agnostic or a Christian,” it is otherwise: trained as you are, each hypothesis makes some appeal, however small, to your belief. (WB 14)

Note, first, that the liveness of the option O:(Christianity, agnosticism) is comprised by the liveness of the two hypotheses being appealing by the fact of the subjects being “trained as they are,” that liveness is a function of both personal history and cultural availability. Note, second, that even though he presumes that his audience is overwhelmingly, if not universally, Christian, he holds that agnosticism is still a live option for them. Here, it’s worth noting that because of their training, agnosticism is a live hypothesis for his audience. Without this training, it is likely not. Recall that James remarks in the Preface to the Will to Believe collection that his preferred audience is an audience “already fed on science,” who have their “native capacity for faith” thereby paralyzed (WB 7).3 Those for whom agnosticism is not a live option are not James’s audience (and, again, James thinks they need their beliefs “broken up and ventilated,” that their “barbarism” should be cleared away by science). Despite claims otherwise, James is not much of a consistent pluralist, at least when it comes to faith.4 This is an issue of no small import, as James restricts the defense of faith in “The Will to Believe” only to overbeliefs in the case of genuine options (WB iv), and the defense of faith here is a defense only of the faithful who feel, if but a hint of, doubt. It is not a defense of the faith that feels no doubt whatsoever. This is not a mere grammatical matter, as one might say a subject’s acceptance of a proposition without evidence is on the basis of faith only when that subject knowingly does so and recognizes that she is overbelieving in that case. For subjects who do not recognize agnosticism or (gasp) atheism as live hypotheses, their acceptance is less a matter of faith, but rather more like a reflex. Perhaps in the same way that it is no great failure of faith that a New England undergraduate might resist the pull of belief in the Mahdi, but a matter of indifference, the naïve believer does not have the kind of faith, or faith proper, that James is out to defend. Again, it’s worth a question as to what faith proper is: either belief without even an inkling of doubt, or belief despite doubt. James has clearly gone to the latter as his preferred model over naïve, simple faith. But surely those committed to the former may object. Parents who send their children to religious schools and colleges do so to ensure that they are not exposed to things that would occasion doubt, to undermine their faith. And eulogies for the pious regularly invoke their unwavering faith, that they never had doubts. Those who express their faith in others even go so far as to say they “never doubted” those in whom they had put so much trust.5

90

Evidentialism and the Will to Believe

A final thing to note about liveness of hypotheses and, by extension, the liveness of options is that it is a graded notion. The liveness of a hypothesis for a subject is “measured by his willingness to act” (WB 14). So what makes or explains the liveness of an option is the subject’s education, inclination, and opportunities, but what expresses the degree of liveness is how inclined the subject is to act on the hypothesis. Importantly, the upper limit of liveness is the “willingness to act irrevocably,” which James holds, practically means “belief ”. A maximally live hypothesis (p) for a subject (S) would be one that S not only believes, but also would irrevocably act on. This means that, as I take it, a maximally live hypothesis is the kind of commitment S would assert even when knowing that it is something that can’t be taken back or revised, it is something that would have S sign a binding contract, devote a full life to. And so even in cases where it was looking like p is false, S must still be committed. Note that James’s criterion here is actional, and it is instructive to contrast the maximal credence in liveness with evidential notions. An evidential maximum of credence (using Clifford’s IER) would be that a hypothesis that (i) has sufficient evidence in its favor, (ii) has no other commitment with more evidence in its favor, and (iii) there is no evidence that it is false or that the evidence is insufficient that has not been rebutted. We might call this maximal evidential credibility certainty, and, again, Clifford’s model has such certainties as definable. But it does not mean that Clifford’s model requires that we may commit only to certainties. Rather, Clifford holds that sufficiency of evidence is the criterion of assent. This difference between evidential sufficiency and evidential certainty will be important later in Sections VII and IX. Returning, again, to James’s graded notion of liveness as the inclination to act, we see the first means by which the pragmatic case for the justification of religious belief will be pursued. We determine belief in terms of liveness of the commitment before the subject. The maximally live hypothesis is quite easily classified as belief. But now James adds something curious, and important: “there is some believing tendency wherever there is a willingness to act at all” (WB 14). James’s first point seems unproblematic, which is that fully confident, unwavering action is a reliable indicator of belief. But in what way is the slight tendency to act a form of belief? Why must all tentative actions bespeak belief? All beliefs can be defined in terms of actions they direct. We can concede here for the Clifford–James debate, as both are committed to this thought (Clifford, for the slope arguments, James, for the Will-to-Believe’s consequences). So, all belief is an inclination to act. But if we concede this, the converse does not follow. Clifford’s strategy was to hold that some hypotheses could direct action,

Reading William James’s “The Will to Believe”

91

even if they are not beliefs. James’s case depends on Clifford’s thought being false. But it is not clear upon what James’s case for the converse of All beliefs guide action to All actions are guided by beliefs can go. I will return to this point in Section X-5, as what I will call the Conversion Fallacy. I think it is rampant both in James’s argument and in further applications and defenses of it. My plan in this commentary is to highlight all the points in James’s argument where a version of the conversion fallacy occurs.

3. Forced options As opposed to the liveness of options, which depends on the liveness of component hypotheses, the forcedness of an option is a function of the option itself. It is not determined by the forcedness of the hypotheses, as hypotheses are not forced on their own. Forcedness is a relational property, like liveness, but forcedness is one that, instead of obtaining between hypotheses and the deliberating subject, it obtains between the two hypotheses. It is, as James puts it, “a complete logical disjunction” (WB 15). James’s first examples are of what are not forced options they are: Go with or without your umbrella Either love me or hate me Call my theory true or call it false In each case, we are presented with a dilemma, but the two options presented do not exhaust all the available options—they admit of at least one tertium quid. And so, one may stay home, remain indifferent, and suspend judgment in the respective cases. We are offered two choices, but we can avail ourselves of a third option that amounts to refusing to play along with the offered game. Notice also that it is possible with some dilemmas to take both horns, and so with “love me or hate me,” we are capable with many in our lives to do both, and as such be ambivalent. And so, forced options, if they offer no third choice, must be (a) exhaustive of the options so there is no third thing to choose, and (b) also be an exclusive disjunction, so that the strategy of taking both is not available, either. James’s sole example of a forced option is the challenge, “Either accept this truth or go without it” (WB 15). The first question is whether this offers no third option. It certainly seems plausible as forced, as the three option choices between acceptance, rejection, and abstention now becomes acceptance and doing

92

Evidentialism and the Will to Believe

without (which encompasses rejection and abstention). And so, it can be put that the matter can be rephrased in a variety of ways, each yielding the same result: Accept this hypothesis, or not. Reject this hypothesis, or not. Abstain on this hypothesis, or not. The form of an option, again, is that of O:(p, q). The form of a nonforced option between p and q, then, would be with r as a third option: O:(p, q), but r and/or both p and q are possible. So really, the form of the option is: O:(p, q, r, p&q) James’s strategy, then, is to reduce any of the third (and fourth) options to one of the two preferred options, so the form of the strategy can be revealed in three developmental stages: Stage 1: O:(p, q), but r is possible. So, O:(p, q, r). Stage 2: q and r are inconsistent with p, let them form a joint class, φ. Given that with Stage 2 we can classify our commitments as doing without another commitment, we can restate the original option as forced, with one target proposition and a doing-without the target proposition class, φ: Stage 3: O:(p,φ), and there are no third options. Given these means of transformation, we can, actually, restate any of James’s original cases of nonforced options as forced, where we simply choose a target hypothesis in Stage 3 as p and classify all other hypotheses in the aggregate class φ, that of doing-without-p. And so: Love me or go without loving me Hate me or go without hating me Go out with your umbrella, or don’t go out with your umbrella Go out without your umbrella, or don’t go out without your umbrella The point is that insofar as we have options and must deliberate, the hypotheses must exclude each other in some fashion (else, we would not be deliberating), and so once we are able to set up the patterns of exclusion with some target proposition, we can transform, just as James does, the unforced options into forced options by renaming all the options as not identical to the target proposition.

Reading William James’s “The Will to Believe”

93

The upshot of my argument is that James’s case for forced options as a unique kind of option is pretty thin. Any option, given my reconstruction, can be transformed into a forced option. So far, I think, we have reason to suspect that the category of forced option may be merely verbal. Now, in James’s defense, it is worth clarifying something important about how we class the alternatives into our catch-all grouping in φ. Here, I think, the Jamesian would want to say that there is another desideratum in classing alternate hypotheses into φ. Let’s return to the nonforced option with a simple tertium quid O:(p, q, r). The next stage, taking p as our target hypothesis would be to put q and r into our class φ, doing without p. The principled Jamesian, I think, would not allow us to do so willy-nilly. We cannot lump, say, hating and being indifferent to and staying home and going out without an umbrella to be classed as the relevantly same type of going-without. Certainly, in the umbrella case, they yield drastically different types of actions, one being on the couch, the other walking in the rain and getting wet. The second, I think, yields different behavior, too, as being indifferent to someone and hating someone do yield different treatment at least usually and in the long term. So what might put the two disparate cases together? One answer to the challenge may come in the form of an example from published correspondence between James critic Dickinson Miller and Curt Ducasse, who defended James’s will-to-believe doctrine.6 Here’s the basic version: Imagine you’re on a trolley that’s lost control. You can believe that it’s best to jump off or believe that it’s best to stay on. Or you can suspend judgment. However, if you suspend judgment, that’s actionally equivalent to believing that staying on is best. So it really is a forced option between staying or jumping.

What allowed us to suspend judgment with the commitment that staying on is best is that they yield the same action, staying on. And so it is appropriate to put them together in order to make the case for forcedness. Not so, however, for lumping abstention and the belief that it’s better to jump. This is because they yield drastically different actions. James proposes a parallel kind of consideration in his earlier “The Sentiment of Ratinality”: [I]t is often practically impossible to distinguish doubt from dogmatic negation. If I refuse to stop a murder because I am in doubt whether it be not justifiable homicide, I am virtually abetting the crime. If I refuse to bail out the boat because I am in doubt whether my efforts will keep her afloat, I am really helping to sink her. (WB 88)

94

Evidentialism and the Will to Believe

Given the tie between belief and action, with forced options “the universe will have no neutrals” (WB 89). And so, we have the tools to reconstruct the conditions for James’s forced options not to be merely verbal reconstruals of the variety of options, but rather, as forced practical options, in light of what kind of cognitive attitudes we take. And so, returning to the formal presentation, the move is from a (seeming) nonforced option O:(p, q, r) to a forced option O:(p, φ), with φ = (q, r) at Stage 2. The aggregation of q and r into φ is not merely on the basis of the fact that they are cases of doing-without-p, but on the basis of their doing-without-p in a relevantly similar fashion. So the streetcar case has abstention and belief that staying is best as doing without the jump in a relevantly similar fashion. The same goes for James’s murder and boating cases: suspension of judgment yields behavior that is identical to one of the positive commitments. So they may be grouped together in φ, which then yields a forced option. With this clarifying rider to the notion of forced options, I think we can save James’s notion of the forced option. But it comes at a cost to James’s overall argument. The cost comes in two forms. First, notice that this way of saving James’s notion of forcedness is inconsistent with the view that all action must make an appeal “however small, to your belief ” (WB 14). In the case of abstention, we act in a way alike to having the belief, but we do not have the belief. This will, later in Section X-6, with what I will call the conversion fallacy, be serious trouble for James’s overall case. But it is worth making it clear here: if abstention from belief yields action equivalent to a form of belief, we have a clear counter-example to James’s view that all action requires belief. The second cost for James is in terms of how this requirement of the aggregation of two hypotheses into one option is deployed later in the essay. As we will see in James’s friendship and romantic conquest cases in Section IX, I do not think the options are sufficiently similar to merit them being placed as equivalent elements of the option. This will then become a serious question of whether religious belief is forced in the same way in Section X-4. In short, for the forcedness criterion to be deployed properly, the cases we treat as functionally equivalent must actually be functionally equivalent. I believe they, with many of James’s cases (friendship, romance, and religion), are not. A final point about forced options must be made, which is that a good deal of the rest of James’s case is already made by the doctrine of forced options. The doctrine of forced options and the broad pragmatic theory of belief we see tying action and belief (namely that all action bespeaks belief) is enough for the case against the evidentialist. If all action bespeaks belief, and there are cases where there is no third alternative to adopting one action or other, then there

Reading William James’s “The Will to Believe”

95

are cases where the suspension of belief in the face of insufficient evidence is not possible.7 Since it is not possible to suspend, any theory demanding suspension is inappropriately posed. Note, again, that the case depends on the converted form of All belief directs action, All action is directed by belief. But, I must note, not only does the latter not follow from the former, it’s pretty obviously false, as, again, even the reconstruction of James’s own doctrine of forced options shows.

4. Momentous options It is likely best to start, in the clarification of the momentousness of an option, with James’s account of how options fail momentousness. This is because James only really provides examples of momentous options, but he discusses how we may have trivial options. James holds: Per contra, the option is trivial when the opportunity is not unique, when the stake is insignificant, or when the decision is reversible if it later prove otherwise. (WB 15)

From this, given James’s “Per contra” we see that an option is momentous when it is unique, significant, and irreversible. Let us, then, see how his example of the Polar Expedition is supposed to be paradigmatic. Imagine you are offered the opportunity to go to the North Pole with the intrepid Norwegian explorer Fridtjof Nausen. You’ll ride on a boat into the ice pack, then you ride sled dogs as far North as you can go. Sounds like a once-in-alifetime opportunity. Once they leave, you can’t change your mind. So it’s yes-orno. Certainly no maybe—that will get you left behind. It’s, then, forced, for sure! It is unique, I will agree. It would be so, even today, with ice-cutters and snow mobiles. It’s significant, as James says it offers “North Pole … immortality,” a kind of achievement that would make you a household name, a somebody. Finally, it seems that it is not revisable, either. Your saying “no” locks you into staying home, and saying “yes” puts you on the boat heading North. Perhaps it’s worth quibbling that one can reverse one’s decision to go. As Nausen leaves the Siberian Islands, you could say: “No more. I’m cold and want to go back to New Haven. Leave me at the port.” But let’s say it’s pretty much irreversible once you’re on the ship’s docket and Nausen’s counting on you for whatever you’re good for. James clarifies the irreversibility requirement with a contrast with a reversible option. A chemist finds a hypothesis appealing enough to perform experiments

96

Evidentialism and the Will to Believe

to verify it, but they yield nothing conclusive. He can junk the theory. “Such trivial options (as this) abound in scientific life” (WB 15). The trouble is that the chemist case doesn’t make it entirely clear where the difference is. In a purely tautological sense, all belief and all actions, once done, cannot be undone. That’s just what it is to have done them. They are in the past, and so one cannot reverse that they occurred. The chemist can’t undo the experiments, can’t unmake the fact that he’d believed the hypothesis, and once the experiments are underway, he, like the Arctic explorers, is practically locked in. Sure, he could knock over all those test tubes and beakers as they bubble and foam in the same way that a person with second thoughts about the Arctic adventure could jump off the ship, but there’s little else that prevents us from saying one’s less or more reversible than the other. A final observation is that, again, James employs a version of the conversion fallacy to make the case that the experimenting chemist believes the hypothesis he’s testing, in that what’s required for him to devote the time and energy to investigating the truth of the hypothesis is that “he believes it to that extent” (WB 15). Consistently, the case for belief is made as being the genesis for action that can contribute to completely satisfying and rich lives. But just because all belief guides action, not all action (in this case experimentation and inquiry) requires antecedent belief in what’s being investigated.

5. Religion as a genuine option Before we proceed with James’s argument, let’s briefly put our attention to James’s endgame: a justification of faith. James will make the case that genuine options may be decided, even without evidence, on the basis of our passional natures. We will be attending to that case over the next number of sections, but we should ask at the forefront: does the description of genuine options really apply to religious belief? That is, James’s defense of faith depends on whether it is a genuine option, so it relies on the following string of conditionals: If S’s faith is to be lawful overbelief, it must be a genuine option. If S’s faith is a genuine option, then it must be live, forced, and momentous. The problem, as I see it, even granting James the pragmatist conversion (fallacy), is that religious belief seems prima facie neither forced nor momentous. I will grant that it is, for many, live. James will make the case for forcedness in Section X, and so I will table that discussion until then. However, I will present the thought that religion is not momentous in the way James requires it is here.

Reading William James’s “The Will to Believe”

97

In short, I think religious belief is more like the chemist’s trivial hypothesis discussed earlier. Consider: how often can we change our minds about what church we go to, whether God exists, whether He’s a good god, and so on? The answer is that our ideas are regularly in flux over the matter. Once we come to believe that God is triune and sent His son to Earth to surprise some country bumpkins with baskets of fish and bread, we can later change our minds about that. It happens all the time. Once we have a religious belief, it’s not like we’re on the boat for the North Pole, or the boat’s off sailing without us. Only if God is a very high-pressure salesman, hawking salvation just for today and saying stocks are running out would it be that way. But it just isn’t that way (and were it that way, God would be pretty jerky). In short, our religious beliefs are reversible. Just as autobiography, I’ve been a Methodist, an agnostic, an atheist, and now I’m what might be called an anti-theist. I’ll likely change my mind again. Who knows? Surely, the same is the case for many who are in James’s preferred audience. Recall that James’s defense of faith is for those who have doubts, not for those who have no doubts. The important thing about those who harbor doubts is that they will change their minds. They are thinking things through. Surely some of James’s audience members have felt the tug of agnosticism or they’d, especially at Brown, heard of the various Christian sects and been tempted by the Baptists or the Calvinists. And this is as it should be. James even, in the Preface to the Will to Believe collection, analogizes religious faith to scientific experiments, contributing to the marketplace of ideas: He should welcome … every species of religious agitation and discussion, so long as he is willing to allow that some religious hypotheses may be true. (WB 9)

The point, of course, is that over the course of the discussions, the open-minded inquirer will change his views. And that’s just like the chemist. The Jamesian can reply that when one believes a religious hypothesis, one must believe it as if it is irreversible. That is, one doesn’t become a Baptist, for example, with the thought that it’s just an experiment that may or may not work out. That’s not religious belief, but something else much weaker.8 So to believe a religious hypothesis as a religious hypothesis, one must be all-in, committed, and have given a kind of final direction to one’s life. Even if later on, one does not remain committed, one must commit as though it is irreversible. This response is reasonable as far as it goes, but, again, there are costs associated with it. First, the all-in model for religious belief certainly undercuts the fallibilist–pluralist model we’d seen in James’s Preface. Moreover, it sets James up for the worry of deepening dogmatism, which I will develop later in Chapter

98

Evidentialism and the Will to Believe

3, when James’s most vocal critics will have a say on Jamesian philosophy of religion. So the balance sheet for the first cost will have to wait for the end of the essay. The second cost is a familiar enough problem—that if the Jamesian wants to say that one can participate in the activities of the religion without the belief in the religion, that seems fine and intuitive. Probably even correct. But notice again that this is more evidence that the pragmatist theory of action is erroneous, since it shows that not all actions are dictated by belief. Now, so far, I’ve argued that religious belief is not momentous, because it does not seem unique or irreversible. We have many, many opportunities to form them, change them, reject, or endorse them. However, I do think James is right that religious beliefs are significant. So there’s that. But notice, really, that significance is, once we see it by itself, reducible to liveness. The appeal and importance of a hypothesis is indexed to believers, their backgrounds, values, and what’s available for them to act on. The same for significance, as wanting to have the immortality of North Pole mythology in the Polar Expeditions is a matter of personal appeal. It sounds great to me, but it’s not a live option for my seven-year-old, even though she likes polar bears very much. My thought, then, is that James’s argument can simply do without momentousness. What’s wrong with momentousness can be jettisoned without harming the overall argument, and what’s right about momentousness, specifically significance, can be captured by liveness. This, of course, does not solve all the problems, as liveness has its own difficulties, and one on analogy with the irreversibility problem with momentousness. Recall that James notes that the “maximum of liveness” of a hypothesis for a subject is that the subject is willing to “act irrevocably”. And we have liveness, again, inherit the problems of irreversibility for momentousness. I take it that so long as James simply holds that the options be live enough, not maximally live, the argument can avoid this difficulty. However, once we allow a graded notion of liveness, specifically for religious belief, we will have a difficulty in articulating how faith is saved. Faith’s not graded. Again, I think this is a matter for Section X, and I will reopen it then.

Section II—Pascal’s Wager 1. Four stages of “The Will to Believe” It is necessary at this point to make explicit the developmental features of James’s essay. Seeing the essay’s argument proceeding in four basic stages is useful in

Reading William James’s “The Will to Believe”

99

evaluating it. The preliminaries, as noted earlier, run through Section VII, but the preliminaries themselves are layered, with developing stages internal to the introductory work. There are four main argumentative stages in James’s essay.9 Logical Stage (Section I): Establishes that there are intellectual questions that pose genuine options. Psychological Stage (Sections II–III): Establishes that nonintellectual passional causes determine human beliefs, especially when there is little evidence. Meta-epistemic stage (Sections V–VII): Establishes that there are distinct normative takes on epistemic evaluation of beliefs and that the determinations of which is psychologically passional. Epistemic Stage (Sections IV, and VIII–X): Establishes that the will-tobelieve is appropriately applicable to a class of genuine moral options, which includes religious belief. So far, James has completed the logical stage of the argument, designating a specific class of issues for subjects to consider, that of genuine options. The next stage is to turn to the psychological truths relevant to how beliefs are actually determined. For the sake of James’s argument, maintaining the difference between these stages is necessary. Two reasons are readily available. First, if James is to make a case against the Cliffordian challenge to faith, especially if the objective is to establish lawfulness in a way his dialectical opponents (or at least those sympathetic to them) can recognize, then he must first begin with the question of how humans actually operate. Recall that Clifford’s evidentialist rule is demanding, as even Clifford himself was well aware. The first case to make against the Cliffordian is to emphasize that the rule of evidence is too demanding, given the kind of minds we have. This is the first objective, and one established by the psychological stage. The logical, meta-epistemic, and epistemic stages are devoted to showing that we acquire a specific class of commitment according to a specific form that, in light of the facts of human psychology established, makes them legitimate. Consequently, even if these commitments run afoul of the evidentialist’s norm, they are nevertheless lawful. This requires first, identifying the class as genuine (Section I), moral (Section IX) options that are rationally undecidable, and then showing that they are, given our cognitive objectives (Section VI), acceptable (Sections IX and X). These will be, first, cases of friendship and social coordination (in Section IX), and then cases of religious belief (in Section X). With the overall plan in mind, let us turn to James’s Psychological Stage.

100

Evidentialism and the Will to Believe

2. Voluntarism and its limits James is out to show that some of our commitments are determined by “our passional and volitional nature” (WB 15). He must, however, square this claim with what seems a more obvious fact about human psychology: with the vast majority of our beliefs, willing them one way or the other has no effect at all. James’s strategy is to concede the general truth, but hold that there are notable exceptions to it. First, the general tendency is that our beliefs are immune to our preferences. James’s examples are that his audience cannot, just by wanting to, believe that Abraham Lincoln was a myth, that the evidence for him and his achievements in history was a fabrication. We cannot, while horribly ill, believe we are well. We cannot look into a near-empty wallet and bring about the belief that it contains 100 dollars. Beliefs are not under our direct control in the way our actions are. Contrast belief with assertion. We do have that kind of volitional control over what we say, so we can assert that “Lincoln is a myth,” “I am just fine,” and “There’s one hundred dollars in my wallet.” But saying isn’t believing. There are a few things to note here. First, simply as a matter of principle, I must highlight the fact that the concession here is further evidence against the pragmatist conversion with belief—all beliefs guide action, but not all action is guided by belief. Second, these cases of no-direct-control over belief function as counterexamples to the view of belief denoted by the term voluntarism, that believing is a function of the will, and so under its control. We see a version of the voluntarist’s view of belief in Descartes’s Meditation IV: [T]he will simply consists in our ability to do or not do something (that is, to affirm or deny, to pursue or avoid); or rather it consists simply in the fact that when the intellect puts something forward for affirmation or denial or for pursuit or avoidance, our inclinations are such that we do not feel we are determined by an external force. (AT VII 57; CSM II.40)

The Cartesian model of commitment, like with Clifford, is on that of assent. And so we may, with the intellect, consider a proposition, but the will provides the movement of assent or denial (or withholding), and it is not moved by the contents. We then regulate our assent by regulating our will. James’s point is that the model for belief is wrong in the broad sense. We not only can’t believe Lincoln was a myth, but we can’t help but believe that he existed. The same goes for the sick bed case—not only is it psychologically impossible for you to believe you are well when ill, it seems unavoidable that you’ll believe you’re sick. You,

Reading William James’s “The Will to Believe”

101

again, can say you’re well, and thereby lie. But you can’t bring yourself to believe it, much as you might like to. And so, for the vast majority of our beliefs, our inclinations and preferences do not have any effect.

3. The wager James’s continued case against voluntarism is made with a brief consideration of Blaise Pascal’s famous Wager. It is not necessary to present a detailed version of the wager or criticisms of it here. A rough version will suit the needs of James commentary. The wager, in short, is that there are two independent variables, each with two values: God exists or not, and one believes he exists or does not believe. This yields four possible outcomes: (i) God exists and you believe; (ii) God exists but you don’t believe he does; (iii) God doesn’t exist and you don’t believe he does; and (iv) God doesn’t exist, but you believe he does. As we’d seen earlier with Clifford’s cases, the reasoning is about error costs, and how to regulate one’s belief (the one variable you control) in order to reduce the costs as much as possible. The errors are (ii) and (iv). The most costly, given the Christian doctrines of Heaven and Hell, is (ii) as God (the preening egomaniac he is) turns out to be pretty touchy about unbelief. He sends you to Hell for eternity for that kind of error. The other error isn’t so bad—you live a life of piety, but you’re wrong about God’s existence. No biggie. The benefits are pretty significant too, for one correct answer, (i), as God (being a self-congratulatory cosmic potentate) prefers to surround himself with yes-men, and provides them with heaven. Being right about there not being a God yields little benefit beyond the smug smirk of the supercilious satisfied atheist. The choice is simple: believe. James poses, first, a normative challenge to Pascal, but then a second psychological challenge. James’s normative challenge is that religious belief should not be formed “in the language of the gaming table” (WB 16). He even imagines that were we to put ourselves “in the place of the Diety,” or perhaps in St. Peter’s shoes before the Pearly Gates, “we would take particular pleasure in cutting off believers of this pattern from their infinite reward” (WB 16). The wager is one thought too many, and were the wagerer to explain how he arrived at his belief to St. Peter, we can imagine Peter nodding knowingly but all the time fingering a hidden lever. Once the soul is finished explaining the wager, Peter replies that the wagerer is very clever and that he richly deserves his reward. Peter pulls the lever, and a trap door opens beneath the wagerer’s feet. The wagerer blinks in

102

Evidentialism and the Will to Believe

disbelief, and then he falls down to the flames and screams below. Peter calls down after, “Say hello to Pascal for us!” The point here about the difference between a bet and a belief yields James’s psychological challenge. James’s challenge is simply that recognizing the belief in God is preferable and the kind of thing one should want to induce in oneself is not equivalent to effectively inducing it oneself. He observes that “a faith … adopted willfully after such a mechanical calculation would lack the inner soul of faith’s reality” (WB 16). That’s St. Peter’s problem with the wagerer—the avowal of the belief is a hair short of a lie, as belief adopted strategically is not the genuine article. If one must use the wager in order to bring about a belief, then one likely already has a hard unbelieving heart, and the wager is merely “a last desperate snatch at a weapon” against it (WB 16). Rather, the option of belief, for the wager to yield real belief, must be live. Consider a contrast, instantiating a version of the swamping many-gods objection to Pascal, posing the wager to someone with no familiarity with the Christian doctrines of everlasting life, and Heaven and Hell. James holds that certainly “no Turk ever took to mass and holy water on (Pascal’s) account.” In further contrast, consider a parallel wager offered to New Haven Protestants by a follower of the Mahdi, with the same variables of existence and belief, but a different version of God (but one nevertheless similarly touchy about unbelief!). Because the God on offer is not a live hypothesis for the New Haven Protestants, the Mahdi’s wager falls flat, it is used vainly on them. And so now, we see the beginnings of an interplay between the logical factors of James’s argument and the psychological features of belief, which is that only live options are ones that have even the slightest psychological purchase on an audience. But, again, even liveness, so far, has its limits. Pascal concedes, after his wager is presented, that he can’t believe on the basis of the wager. He can, however, act, and his actions bring him before the live options as they are for him—he goes to mass, eats communion, gets into the life of the church, stops talking to atheists. That, later, yields belief. So, too, might the follower of the Mahdi’s wager go for a person in Cairo, on the fence about sects of Islam, to bring about actions that may, perhaps down the road, bring about belief. But only, again, under the restriction of liveness.

4. Clifford’s veto Despite the fact that James is presently criticizing Pascal’s wager, and he, at first, brings him in approvingly, it is Clifford’s views that are James’s main target for

Reading William James’s “The Will to Believe”

103

the essay. As I take it (and as we will see in Section III), James’s rhetorical strategy is to at first downplay the role of will and passion in determining belief, and set the limits of voluntarism and the evidentialist’s objection of the use of preference to determine belief. Things, as the rhetorical strategy goes, are looking pretty bleak for faith. But then things turn around, and faith is saved. It must, though, be dark before the dawn, and so enter Clifford and the evidentialists. From the psychological perspective, voluntarism is “silly.” From the evidentialist’s perspective, voluntarism put in the service of directing one’s belief by preference is “vile” (WB 17). The sentimentalist seems, given the objectives of belief and its contributions to the ends of human knowledge, “besotted and contemptible,” James acknowledges. The aspirations of the “incorruptibly truthful intellect” require that we accept truths that are inconvenient, disbelieve tales that are contrary to what we know, and to suspend judgment when we cannot objectively make a determination. Clifford is the “delicious enfant terrible” representing this tradition, and James quotes him liberally, invoking the core principles of the evidentialist’s view, which, again, morally require that evidence and evidence only determines belief. Clifford’s veto to Pascal’s Wager, then, is that even were the wager possible to perform as a belief, it would “desecrate” the faculty of belief; that the pleasures of the belief and its consequences would be “stolen” pleasures; that the strategy of belief determination would be a “defiance” of our moral duty to be caretakers of human cognition; that it would be simply “wrong” to believe on insufficient evidence. The culmination of Section II is that the following seems to be true: Faith is a case of bringing oneself to believe without or contrary to one’s evidence. This seems, first, psychologically impossible. Second, it seems (on the evidentialist’s view) morally unacceptable.

James’s argument in defense of faith must first answer the question of psychological possibility, and then he must turn to the normative evaluation of the beliefs.

Section III—Psychological causes of belief 1. A concession to evidentialism James’s case against Clifford must proceed in two steps. The first is the descriptive case, that we do not follow the evidentialist’s rule in all cases. This is established,

104

Evidentialism and the Will to Believe

again, with the psychology of belief outlined here in Section III. The normative step will be to show that not only do we not follow the evidentialist’s rule, we need not follow it. He thereby establishes a “right to believe,” on the basis of there being appropriate lawful exceptions to the evidentialist’s rule. This is accomplished by the aggregation of the meta-epistemic and epistemic stages of the argument, later. It is necessary to note that though James is now turning to answer the evidentialist’s challenge to faith, he nevertheless concedes that the account of belief and its norms offered “strike one as healthy” (WB 18). That is, as a general rule, James thinks the evidentialist’s norms are about right and that willful believing, especially against one’s evidence, is not a good idea: wishful thinking is only a “fifth wheel to the coach” (WB 18). But there is a difference between the evidentialist’s rule being a good general rule and it being a universal rule, one with no outliers or exceptions. James’s argument is that there are many cases where neither mere preference nor reason makes the determinations. But willing in a broader sense determines beliefs. So a distinction is in order. There are what might be called mere willings, and they are our individual preferences, specific choices, a hope for this or that. These, James has conceded, not only do not influence belief, but where they to, they shouldn’t. So our deliberate particular volitions are ruled out, but that does not eliminate all of our “willing nature.” What is left over is our broad willings, the inclinations behind our intellectual lives that make options live or dead for us. James clarifies the scope of broad willing: I mean all such factors of belief as fear and hope, prejudice and passion, imitation and partisanship, the circumpressure of our caste and set. As a matter of fact we find ourselves believing, we hardly know why. (WB 18)

In this case, it is, as we’d seen with the liveness notion before, a function of our social milieu, the current trends, native human instincts yielding beliefs, not on the basis of reasons, but on the basis of a function of our being properly synced as a society. James gives us some examples he believes are true of his audience: Here in this room, we all of us believe in molecules and in the conservation of energy, democracy and necessary progress, in Protestant Christianity and the duty of fighting for “the doctrine of Monroe,” all for no reasons worthy of the name. (WB 18)

James’s point is that members of his audience, members of a philosophical club, would agree they do not have full-bore reasons to believe in these matters. Yet they do believe, and the fact that they do not have the evidence does not undercut

Reading William James’s “The Will to Believe”

105

their commitments. It’s here that, I think, a genuine Cliffordian may demur, as the Cliffordian doctrine of evidentialism allows testimonial evidence and the sacred tradition of humanity in questioning, testing, and then endorsing views endorses (most of) the commitments on the list—we believe on the basis of the testimony of tradition (L&E 197). This, for the Cliffordian, is weak evidence, but in many cases, that is sufficient. Ironically, James’s case is similar to Clifford’s, but James does not take the commitments of the larger cognitive culture as evidence. With the average believer, James takes it that commitment is bought not by “insight, but the prestige of the opinions.” The fact that we link up correctly with others in some beliefs “makes sparks shoot from them and light up our sleeping magazines of faith” (WB 18–19). We live by ad populum reasoning, we are social animals prior to being rational animals, and it’s often that the sociability is the one that takes the lead over our rationality. And thus, in our cognitively dependent positions, as Clifford also observed, we defer to others given what they believe. But we do not do so, James holds, on the basis of Clifford’s testimonial principle. Rather our psychology is so that for us, “our faith is faith in someone else’s faith” (WB 19). Recall that Clifford’s testimonial principle was that one should attribute quality of evidence from testimony only on the basis of being able to assess the veracity and the knowledgeability of the attester. Failing either, we cannot responsibly accept and believe on the basis of testimony. This means that in a testimonial chain there is a knower (or at least someone with first-hand evidence) at its head, who attests to a second person, who assesses the first’s veracity and knowledge, and thereby can come to know. Contrast this with James’s model. There is transmission, but the requirements of establishing knowledge are suspended—our faith is in someone’s faith, and so on up the chain. This contrast is significant, as it reveals something that is altogether overlooked in commentaries on the Clifford–James debate. It’s a regular observation that Clifford has a restricted notion of evidence, and James’s reply, even if unsuccessful, nevertheless is in the service of broadening the notion Clifford and his positivist counterparts were working from. I’ll even tell a version of that story, later, with doxastically dependent facts (see Section IX-5). However, this current contrast cuts a very different way. It is James that has the thought that the attitudes of a culture or that testimony from those who would know is not evidence. The whole point of Clifford’s project is to show that because a cognitive culture functions as evidence for its participants, we have an obligation to make sure it functions properly. Clifford sees cognitive inheritance as our default ways of making sense of the world, our background

106

Evidentialism and the Will to Believe

assumptions that allow us to interpret discrete pieces of information. Clifford, then, would have no evidential problem with many of James’s examples, then. Molecules and the conservation of energy we believe in on the basis of our deference to chemists and physicists, who have publicly verifiable means of demonstrating their conclusions. Democracy can be justified as the political arrangement that allows the sacred tradition of humanity, that the freedom of inquiry and questioning, to remain unmolested. Necessary progress and the Monroe doctrine, I think, are products of colonialist group-think, and that’s too bad for those who believed them (and worse for those who suffered from them as a consequence), and Protestant Christianity has precisely all the problems all religions have. I don’t see any trouble a good Cliffordian would have here in the face of James’s case, but the line from James works because Clifford’s account of testimony is not part of the evaluation. It won’t be the last time Clifford’s views don’t get a fair hearing.

2. Truth and other useful ideas James’s model for our broader willing nature bearing on belief, then, is not a model for wishful thinking, but a social model of belief formation. Our faith is faith in another’s faith, and we are directed to this faith over that faith in terms of the prestige the options have, what use we feel we can put them to. This isn’t just for our beliefs in molecules and the Monroe doctrine, but with regard to what James calls “the greatest matters” (WB 19). James’s example is “our belief in truth itself.” This surely is one of the greatest matters. But James asks us to consider: what do we have to say to the “pyrrhonistic skeptic,” the person for whom our reasons can always be challenged, the person who unremittingly doubts. We certainly think it would be preferable that truth is such a way and our minds are in a way that we can know it. But that, it seems, is all we have to go on when we ask whether this convenient arrangement of mind and world is true. We cannot answer the skeptic’s challenge, but we are all together in this, we have a “social system that backs (us) up” in remaining committed to truth. It’s a useful commitment, but it is, again, the most we can say for it. Now James turns this point on the Cliffordian evidentialist. He holds that our beliefs are determined by our passional natures, the social opportunities they afford, and the uses to which we put them. When they hold no appeal to us, or offer no social purchase, or we have no use for them, we reject them. “As a rule

Reading William James’s “The Will to Believe”

107

we disbelieve all facts and theories for which we have no use” (WB 19). Consider, now, that on James’s view that it is more a social phenomenon, then, that religion receives so much critical treatment. Those whom James calls “the logicians” have no personal use for religion and the hypothesis of the divine. These are dead options for them, so they reject them. But, now it is clear, they employ a double standard in doing this—they value truth and empirical evidence, and have a use for it, and so they valorize it, even when there is no evidence for doing so. But because they have no use for religion, they think it is not worthy of their consideration. The problem is that they have taken the subject sensitivity of liveness for themselves to be universal. This very law which the logicians would impose on us … is based on nothing but their own natural wish to exclude elements for which they, in their own professional quality of logicians, can find no use. (WB 19)

James has two points, one negative and one positive. The negative point is a kind of internal criticism of the logician evidentialist, which is that it is an error to move from propositions of the schema X is of no use to me to the schema X is useless. The positive point is that the initial motivator for the logician is the logician’s broader willing nature making a determination of what’s salient, and then setting to deliberate. As James concludes, “Evidently, our non-intellectual nature does influence our convictions” (WB 19). The point has been to establish the psychological groundwork for a case against the evidentialist’s rule. With the case of truth, James has tried to do two things. First, he’s proposed an example of psychologically plausible overbelief, even in the face of skeptical challenge—our hopes determine our belief in truth. Second, he holds that the evidentialists themselves perform this move. They value truth, so they cling to it in the face of the pyrrhonistic skeptical challenge, but they then impose a species of that valuing on others. And so, here, James has proposed not just a counterexample to evidentialism, but also a form of argument from hypocrisy. We’d seen two arguments from inconsistency against evidentialism before. The first is the self-refutation problem—that the moral argument for evidentialism is insufficiently evidential (see Section Chapter 1, Section I-5). The second is the double standard objection (see Chapter 1, Section III-2). As I take it, James’s case here is a special version of the double-standard objection, as he does not think this inconsistency shows that the evidentialist’s view crumbles, but that the evidentialist is not circumspect with how the view is applied. But remember, first, that Clifford had the tools to reply to both objections, and specifically to

108

Evidentialism and the Will to Believe

the second. He devoted all of Section III of “The Ethics of Belief,” which took our commitment to truth and the possibility of inquiry through induction to be what I’d called a “regulative principle,” and not a full-bore belief. Again, the Cliffordian requirement is that we accept and treat this commitment (among others) as true to run inquiry, but we are not justified in thinking it is true beyond the scope of affairs to which it is applied. In short, what James takes as a counterexample to the evidentialist’s belief rule is precisely what Clifford had refused to countenance as a belief. It is hypocrisy only if the purported hypocrite actually does what you’ve charged is hypocritical. Clifford did not endorse a belief, but a rule of action and inference. Again, everything for James’s case, then, depends on how one reads the connection between beliefs and action. Clifford is committed to the proposition that All beliefs direct action, but for the Jamesian story to work on Clifford, the converse needs to be true, All action is directed by belief.

3. Pascal is a regular clincher To close the section, James pauses to save Pascal’s Wager (and the faith endorsed by it) from the dustbin to which it seemed to have been tossed. Remember that Section II had the psychological and normative challenges to faith making it look pretty bleak for the faith James had promised to justify. But it was a rhetorical device, one to the effect: Critics say… One then gives the critics their due, you let it look pretty bad for your target proposition, the preferred view. But then there’s the dramatic comeback. Like a boxer pummeled for a few rounds, beaten and knocked down, faith finds renewed strength for this round. James’s proleptic strategy has now come to fruition, as he closes to reconsider Pascal’s Wager. It looked “silly” and “vile” last time we saw it. But once we have these new considerations before us, James holds that “it seems a regular clincher,” as the Pascalian strategy of giving himself motivation to put himself in the right social nexus where the belief will dawn on him seems right. Pascal, as we’d seen before, goes to mass, thinks lovingly and deeply on Christ’s sacrifice, and stops talking to doubters. The social pressure, the satisfactions promised by and consequent of the belief, and the costs he knows he will pay can quash the doubts and strengthen belief. “Pure insight and logic,” James intones, “are not the only things that really do produce our creeds” (WB 20). Faith, then, is psychologically possible, once we see that it is not mere willing that is the source of the belief, but our broader willing natures that fix and maintain our deepest belief.10 And so the psychological case for faith is complete. It not only is psychologically possible,

Reading William James’s “The Will to Believe”

109

but it also seems to be behind even the critics criticisms of faith. It looks, as a fact about our minds, unavoidable.

Section IV—The thesis of the essay 1. A thematic transition Earlier, in Section II-1, I made an interpretive stipulation that James’s case proceeds according to two descriptive stages (the logical and psychological) and then two prescriptive or normative stages (the meta-epistemic and epistemic). Recall that as Section II closed, the two concerns about faith, so presented, were that it was impossible and vile. Sections I and III together identified the liveness of a hypothesis with specific features of a subject’s broader willing nature to show that sentiment does play a role in determining belief. So the psychological challenge to faith has been answered—faith is possible, as one can and regularly does believe a good many propositions on the basis of flimsy or no evidence. James makes it explicit that he’s now turning to the normative question of faith, as it is clear that just because it is possible to do something, it does not follow that one has the right to do it. James then poses the transition as “ask(ing) whether it be simply reprehensible and pathological, or whether on the contrary, we must treat it as a normal element in making up our minds” (WB 20). Of course, here, everything hangs on the “normal,” which has either generality or normativity (or both). And so, one could say, “Normally, if you leave your purse on a park bench, it will be stolen” without endorsing the appropriateness of the theft, but just that one predicts it. Alternately, one could say, “Normally, when people receive presents, they say ‘Thank you’.” And with this, you’re commending the common practice of expressing gratitude. So how normative is James’s “normal”? My hypothesis is that James means it in both the descriptive and normative senses. He’s surely committed to the descriptive normality of the tendency—his previous section bristled with examples of the habit. That James is committed to the normative normality here depends on showing not only that the practice is allowable, but also that it is preferable (and perhaps, that it is obligatory). James’s “The Will to Believe” and normative force of “normal” here are weaker than views James had earlier espoused. In a review of P. G. Tait’s Unseen Universe, James (as related by Ralph Barton Perry) holds that there is a positive duty do believe:

110

Evidentialism and the Will to Believe

[A]nyone to whom it makes a practical difference (whether in motive to action or mental peace) is duty bound to make it: If “scientific” scruples withhold him from it, this proves his intellect to have been sicklied o’er and paralyzed by scientific pursuits. (Perry 1935: I.529)

James had earlier, then, a view that there is a positive obligation to believe in accord with one’s broader will. Richard Gale’s “Promethean Pragmatic Syllogism” takes James’s work to be an extension of the same line of thought: We are always morally obligated to act so as to maximize desire satisfaction over desire dissatisfaction Belief is an action. Therefore, we are always morally obligated to believe in a manner that maximizes desire satisfaction over desire dissatisfaction. (From Gale 1999: 24)

I think Gale is on the right track here, which is that the ethical principle designating desire–satisfaction as of primary import does make the normativity of the case much stronger than that one is merely allowed to believe, or that one has a right to do so. Consequently, the force of “natural” (and “must” later) will be read in normative terms. This said, James’s promise of a justification of faith is not a promise that he will show that faith is obligatory, only that he promises “a defence of our right” to overbelieve (WB 13). James’s own weaker characterizations of his view, for example, in the Preface to the Will to Believe collection, cast the project as “preach(ing) the right of the individual to indulge his personal faith” (WB 8). My thought, then, is that James is using “normal” with normative force alongside its descriptive force. There is a range of normative assessment this force has, and as we’ve seen, it ranges from: Weak: The right to believe, so it’s not wrong to overbelieve. Modest: The right to believe, and belief is preferable to unbelief. So it’s not just not wrong to believe, but good. Strong: The duty to believe, so belief is an obligation and not to believe is wrong. Now, I think James very clearly was committed in his earlier work to a strong normative assessment, so a duty to believe and Gale’s interpretation of the Promethean syllogism and Wernham’s “heretical” reading have James consistently maintain this strong view. James not only makes a case for the right to believe, but also about the problems of the failure to believe. This, I think, is equivocal evidence. It surely inclines us to take James as providing something stronger than merely the weak assessment. But it does not follow

Reading William James’s “The Will to Believe”

111

that he is committed to the strong assessment. My thought is that James will make a modest defense of faith, one that must come on two levels. First, he must establish that it is within our rights to overbelieve some propositions, and second, that it is preferable to believe them than not to. That is, the comparative case he makes against the evidentialist is not that they are wrong to be evidentialists, but that they are making choices that he thinks are less choiceworthy.11

2. The thesis James announces that he will defend the thesis that overbelief under specific circumstances is not only lawful but also necessary. He is good enough to put it all in italics. Our passional nature not only lawfully may, but must, decide an option between propositions, whenever it is a genuine option that cannot by its nature be decided on intellectual grounds; for to say, under such circumstances, “Do not decide, but leave the question open,” is itself a passional decision—just like deciding yes or no — and it is attended with the same risk of losing the truth. (WB 20)

Given our schemata, here is a more formal presentation of the doctrine of the will-to-believe: WTB: If S has option O:(p, q), and the option is (i) genuine, and (ii) intellectually undecidable, then S may and must decide the option on the basis of S’s broader passional nature. Given how things have progressed so far, we do not have too much clarifying to do. However, there are two parts of the WTB in need of attention. The first is the force of “must.” In essence, we have a similar interpretive issue as we had before with “natural”—is it purely descriptive or is it normative? In this case, is “must” an alethic or deontic modality? The crucial phrase is: Our passional nature not only lawfully may, but must, decide… Now, the two interpretations yield the following: Descriptive: Not only are we within our rights to decide on the basis of our passional nature, but it’s psychologically determined that we will. Normative: Not only are we within our rights to decide on the basis of our passional nature, but it is what we ought to do.

112

Evidentialism and the Will to Believe

Again I don’t take the interpretations as exclusive, but it is important to identify the strength of the normative force of James’s commitment here. Again, my defaults are set on taking James to be defending a modest normative commitment in this context, but this is one place where my modest line of interpretation has trouble, because there’s a strong reading of normative “must” which seems very plausible, namely that it is an obligation. Now, my thought is that James’s statements varied widely on the matter. We have a version of weak right to believe in the Preface, we have James give a strong duty version here. The best interpretive option, it seems to me, is to allow James a wide play in the form of the view in his statements of it, and then dial our interpretation of the view to what degree of normativity his arguments can support. If they support a strong version of the view, then we take him as committing himself to strong here, and we treat his weaker assertions as noise. If the arguments support weak, then we’ll take his sober assessments in the Preface and Prologue as his reflective judgment, and the stronger statements as rhetorical overstatements. My thought, though, is that the modest view, one that gets over the rights to believe, but only then establishes the preferability of the option, is what his argument best supports. The problem for my view is that James rarely states the view that way. Ah, but I say back to this: that’s why he needs a commentator. I will try to finally resolve this issue in Section VII, where we can assess the strength of James’s normativity once we see the strength of his meta-epistemology. Finally, we have the question: what does James mean by an option “that cannot by its nature be decided on intellectual grounds”? We are very clear about the first (sufficient) condition for the WTB—that of genuineness—but James doesn’t devote an entire section to the fine definition of what constitutes an issue by nature intellectually undecidable. There are two parts to the notion. Let’s start with the idea of something being rationally undecidable. We start with James’s earlier examples. They are: the existence of God (the Wager presumes it being unresolved), the existence of molecules, the conservation of energy, democracy, necessary progress, the Monroe doctrine, and that truth exists. James’s presumption is that none in the audience could possess the sufficient evidence to demonstrate these hypotheses true. Again, I think there’s a reason for why there’s a lack of evidence for at least two, which is that they are false. That would make them intellectually decidable, but against them. The takeaway is that James proposes with the WTB that we have (at least) a right to overbelieve, and that one of the background requirements of this is that when we form the belief, we do so on the basis of insufficient evidence.

Reading William James’s “The Will to Believe”

113

But it is not just that the issue is not currently intellectually decided for S, but that it is undecidable. This is connected, now, with the thought that the modalities of decision are not determined by the kind of proposition up for consideration. So the second element, by its nature, must be invoked to explain the fact that the issue is not decided at that time for that subject—that is, it is by its nature, undecidable. Notice that this by-its-nature-undecidable is a very strong restriction. In fact, as I see it, the restriction is the strongest of the ones possible as combinations. First, three degrees of an option O:(p, q) being undecided: Skeptical: O:(p, q) is undecidable because there is evidence for neither p nor q Stalemeate: O:(p, q) is undecidable because the evidence for p and q is balanced or equipollent Insufficient: O:(p, q) is undecidable because there is a slight evidential edge for one, but it is insufficient to determine its truth.12 Now, let the option have its status of being undecidable explained by its nature. “By nature undecidable” comes in four degrees: Degree 1: O:(p, q) is undecidable now, but at some time it will be decidable. Degree 2: O:(p, q) is undecidable now, but at some other time (in the past or in the future) it could be decidable. The trouble is that these first two degrees don’t feel like they are cases of an issue undecidable by its nature. The first might be undecidable because the nature of the question requires long-term data collection, and the second might be because the nature of the issue is that it’s about an event in the distant past. They are by nature undecidable in a very weak sense, but not by their essence. Consider the stronger versions: Degree 3: O:(p, q) is undecidable now, and there are no times when it could be decidable. Degree 4: O:(p, q) is undecidable now, and there is no evidence whatsoever that could decide the matter. Now, my thought is that an issue that is by its nature intellectually undecidable is either skeptical or stalemeate and either Degree 3 or 4. I’m willing to allow the Pascalian thought that the question of God’s existence is like that, or perhaps that it’s stalemeate and 3rd or 4th degree. That trouble is, none of James’s other examples seem to be by nature undecidable. Is the existence of molecules or

114

Evidentialism and the Will to Believe

the law of conservation of energy by nature undecidable? How about our commitment to democracy? Even if the people in the lecture hall at the time had no answers, it’s not like they couldn’t walk out of the hall and within 15 minutes be speaking to a chemist, physicist, or political philosopher. They are on campus at Ivy League schools, for Pete’s sake. Surely somebody there could answer these questions. With necessary progress and the Monroe doctrine, I don’t think they are still undecidable. They were both versions of American neocolonialism, and though they made many Americans rich, they were not wholly just doctrines. And so they, too, are not by nature undecidable, but they were decided later. So they were Degree 1 (or 2) undecidable, which is not the nature invoked, but really an accident. Note that the play between degrees of undecidability will be relevant when James turns to the cases that are allowable later (what I call the friendship, courtship, and social coordination cases), which all seem to be forms of insufficiency and Degree 1 with undecidability. The God question, the religious option, given James’s setup and analogy with Pascal, is supposed to be skeptical or stalemeate and Degree 3 or 4. My thought is that if the case is supposed to be made that because the former may be allowed, the latter can be, we need a stronger analogy between the kinds of undecidability the cases manifest.

Sections V and VI—Absolutism and empiricism 1. Two forms of faith From the skeptic’s perspective, we are all overbelievers. Our deepest commitments, even the commitment to the fact that there are truths at all, cannot be defended from the skeptic’s withering gaze. But we nevertheless believe. That’s overbelief, and so far, James has established this as a fact of human psychology—overbelief is natural. Our natural inclination to belief, however, comes in two strengths. There is high-powered and more modestly posed belief. The high-powered purported knower believes not only that there is knowledge, but also that we can know when we know. We can, as a result, not only know, but also be certain. The modest believers are not quite as sure about the beliefs they maintain. They think it is possible to know, but they are not committed to the thought that one can know when one knows. James calls the first strong form of natural commitment absolutism, and the second modest form empiricism. The distinguishing element of the two views is how they come down on the question Can we know when we

Reading William James’s “The Will to Believe”

115

know?, and in light of how the two camps are divided on knowing when one knows, they will be divided on the kind of grounds one needs in order to know at all. This will lead James to a substantive discussion of the nature of evidence, which yields a parallel distinction between two views on the nature of evidence. James distinguishes between what he calls “objective evidence” and “subjective opinions” about one’s evidence. My plan here is to survey these distinctions and what hangs on them. I’ve pushed these two sections of “The Will to Believe” together, because there is substantial connection between the two (and, in fact, they read as though they were likely written as one long section, but broken into two for the sake of presentation). The distinctive claims of the two sections can be presented as follows: Absolutists hold that one can know when one knows, which requires that they are committed to the existence of objective evidence. Empiricists hold that one does not infallibly know when one knows and are committed to a modest subjective notion of evidence. The genealogical story James presents is in the form of an explanatory dialectic. It runs in two directions. Absolutists, committed to the value of knowing when one knows, must hypothesize objective evidence as what can provide not just knowledge, but second-order knowledge about that knowledge, too. When they turn to articulate this objective evidence, however, it seems to slip away into interminable disagreements and reinterpretations. The empiricist, seeing this pattern, starts with the observation that we seem to have nothing but our interpretations and tentative hypotheses about our evidence, and then she concludes that one may know, but one does not, given the difficulty with the inputs, have the resources to know when one knows.13 Rather, one muddles along with one’s hypotheses, hoping they do well in the long run, and changing them out when they don’t work in the short run. Let us begin with the absolutist’s aspiration: to know when one knows. Surely even the most empiricist Jamesian or fallibilistic pragmatist can acknowledge why this is a goal worthy of at least trying to attain. The skeptic challenges us, asking how we know what we think we know, and that’s just annoying. Socrates was accused of corrupting the youth, because of the way his critical methods ate through pretensions. Now natural dogmatism is one strategy, and one might ignore the skeptics or laugh at them, but these don’t provide the intellectual satisfaction of answering them. That’s the absolutist’s objective: to answer those who express doubts or have critical questions. Those who know can address challenges until there aren’t any more doubts. Again, isn’t this a perfectly

116

Evidentialism and the Will to Believe

reasonable and even admirable model for knowledge? The model of possessing one’s reasons, being held responsible for them before critical questions of another. It is an error of dishonesty to fail to pay respect to the aspirations of the absolutist’s model for knowledge. The primary requirement, if we are to possess these reasons in ways that can satisfy our felt duty to answer the critical questions, is that we are able to access, and assess our reasons. We must not only be able to see how we arrived at our conclusions, but that these were good reasons for doing so. And so critical questioning turns to critical reflection—we turn inward and test our knowledge in order to endorse it from what its grounds are. The value of knowing when one knows, then, is of prime importance, given this objective. Now, notice a few things about James’s formulation: The absolutists … say that we not only can attain knowing truth, but that we can know when we have attained knowing it. (WB 20)

Note, first, that absolutism isn’t yet a full-on dogmatism of any sort—we can know truth, not that we do know the truth or that it’s truth of one kind or other. So the absolutist is a kind of cognitive optimist, but one that’s not necessarily committed to dogmatizing. That comes later, on James’s story. Note, second, that the absolutist just holds that knowing when one knows is not a requirement for first-order knowledge, just that it is a possibility of it. It’s worth making clear the force of this thought. Take Samantha (S). She knows the cat’s on the mat (m). So she knows that m. She knows that m on the basis of some evidence set: She sees Fluffy, and Fluffy’s on the mat; she has a good track record of Fluffy identification; and she generally does not confuse the mat with the rugs, pillows, and other things around the house. Imagine, perhaps, S needs to be really sure that she knows that m, so she surveys the evidence again, sees it supports m, maybe makes sure Fluffy hasn’t moved, and now she knows that she knows m. Now, I don’t think that’s all that hard, but knowing that one knows causes some pretty heavy breathing when you throw skeptics into the critical discussion. Well, says the skeptic, how does S know that this isn’t that one time she’s mixed up her coat with the mat? It’s easy to do with a cat on it, you know. Or: the fat possum that lives in the alley—Sam once mistook it for Fluffy and gave it some treats. Did Sam make sure that instead of the cat on the mat, she’s not got a case of fat possum on the mat? Oh, and what if Sam’s dreaming, in a world controlled by an evil demon, enslaved by a deceiving computer, or abducted by aliens who’ve taken out her brain and put it in a vat?

Reading William James’s “The Will to Believe”

117

With answering the skeptic, knowing that one knows turns out to be really hard, and not just any old evidence will do with that project. You need objective evidence. Now, we’ll get to that part of the story in the next subsection. For now, let’s focus on knowing that one knows and James’s formulation. James takes it that the absolutist holds that (a) one can know truths, and (b) when one knows, one can know that one’s knows. Note, again, that it is not the requirement that one must know that one knows in order to know. It is only that if one knows, then one can know that one knows. That’s still a pretty substantial view about knowledge, but not as strong as holding that knowing that one’s evidence provides knowledge is a prerequisite for knowledge. Let’s put it this way: Strong: It is necessary that if one knows, then one must know that one knows. Weak: If one knows, then it is possible to know that one knows. The strong version of the view, called the KK thesis, is associated with rabid epistemic internalisms, like those of Descartes, Spinoza, and Augustine. James’s stated version for the absolutist is the weaker, and I don’t see any reason why the absolutist view here must be associated with strong internalism, as surely even epistemic externalists can acknowledge that it’s possible to know that one knows. Externalism is just posited on the thought that one needn’t be aware of the reasons that provide one’s grounds, not that it’s impossible to achieve that.14 The upshot of James’s distinction, then, is that we recognize the situation for even the weaker version of the knowing when one knows view is that if you can’t even possibly answer the skeptic, then you don’t know. The empiricist, given that James portrays the view as nevertheless being committed to knowledge, rejects even this weak version of the knowing-when-one-knows view. And so, for James, empiricism is the commitment to knowledge, but to the fact that one is not only never, but cannot possibly ever be, sure that one knows, even when one knows. Before we turn to the treatment of evidence, I must pause to ask whether empiricism, so construed, is much different at all from skepticism. That is, let us make use of the old Jamesian forced option model here. We have an option between skepticism (s), empiricism (e), and absolutism (a). So the option is O:(s, e, a). Can’t the absolutist say, in Jamesian fashion: Accept the absolutist program or go without it, and then turn this into a forced option? Perhaps, and here’s why. Empiricism and skepticism share the attitude that one shouldn’t even try to show or determine that one knows that one knows, one lacks certainty, and one just muddles through life without any objective grounding. Where is the practical

118

Evidentialism and the Will to Believe

difference between knowing but not knowing it and not knowing at all? From the inside, it’s all the same. Jamesian empiricism, so described, is just skepticismlite with an apology for being dogmatic here and there.15 The Jamesian tools for reply, I think, are twofold. One strategy could be to bite the bullet and concede the family resemblance with the skeptic here, but then note that once one sees that (as we will see in the next subsection), absolutism can’t work, one must go without it. Once one takes on going without absolutism in such a forced move, empiricism isn’t the forced move, it’s just preferable to (and more psychologically plausible than) skepticism. This, in fact, could be the Jamesian dialectical genealogy recast. Alternately, the Jamesian can try turning the tables on the forcedness move. We could run forcedness with empiricism. So, again, the option starts out O:(s, a, e), but because skepticism and absolutism are about the aspiration of knowing when one knows (one’s optimistic, the other’s pessimistic), they go into the grouping φ of doing without empiricism, and so the forced option is now: be an empiricist or go without. I think there are two lessons. The first is that it’s all too easy to play the forced move game. And, again, I think this is more evidence that it’s a merely verbal category. The second lesson is that everything about the appeal of empiricism hangs on whether the absolutist can make good on the promise of objective evidence. Empiricism, so cast, is appealing only because absolutism doesn’t work.

2. Objective evidence and its discontents Everything for the absolutist’s program hangs on the evidence that can be proffered, evidence not only sufficient to warrant a subject in believing a proposition and thereby knowing, but also evidence that would allow the subject to know when she knows. This is what James calls objective evidence, and it is characterized in two ways: positively and negatively. James’s positive characterization of objective evidence is that it consists of “things (which) illumine my intellect irresistibly,” a “final ground” that ensures that our intellects are adequate to their objects (WB 20). A Cartesian would call it clear and distinct perception—we are aware of a truth and cannot confuse it with any other truth. Other traditions might characterize it by acquaintance, or by recollection. Regardless, it, positively, is a reason that provides its own guarantee for acceptability—a self-justifying proposition. Now, when we have such propositions, by hypothesis, we don’t have room to question them. We naturally assent to them. Consequently, the negative

Reading William James’s “The Will to Believe”

119

presentation of objective evidence is that it “leaves no possibility of doubt” (WB 20). And thereby, we have certainty—we have clearly adequate evidence that provides us not only reason to believe but also reason to have no reason to doubt. Here’s an irony. James holds that he’s an empiricist, and so does not believe that he knows that he knows things. And when he provides us with a list of propositions that are self-certifying to illustrate what objective evidence is, it seems he’s trapped himself. That is, James motivates the notion of objective evidence with some examples of certainties, the kind of things absolutists latch onto for their cases of objective evidence: I know I exist before you Two is less than three If all men are mortal, then I am mortal, too Considering these propositions, it seems they are about as evidentially well grounded as they get. But doesn’t that put James in a bind? In giving examples of objective evidence that illustrates what the absolutists are looking for, James makes his empiricist view look pretty silly. Empiricism is the view that there is no objective evidence, but in order to explain what the absolutist commitment to objective evidence is, James has provided some objective evidence. That’s a case for absolutism, not empiricism. “Thanks,” says the absolutist. “With critics like this, who needs yes-men?” But James’s strategy is wilier than that. He qualifies his examples with an “If…” “If ” you can’t doubt that you exist, “if ” you are illumined by the natural light of the intellect when you consider two is less than three, “if ” you are certain that if all men are mortal, you are, too. If. And it’s here that James’s first case against the notion of objective begins. Consider the basic truths in the exemplary list for objectively evident propositions. Notice that even though they seem pretty much obvious, you can work some doubts behind them. James gives some examples of obvious claims that have serious challenges: The fifth Axiom of Euclid can be rejected, even though it seems obviously true. We call that non-Euclidean geometry, now. All of the logic of the syllogism in Aristotle’s logic has been rejected by Hegelians. (WB 22)

James concludes from these findings, “no proposition ever regarded by anyone as evidently certain… has not ever been called a falsehood, or at least had its truth sincerely questioned by someone else” (WB 22). So his initial list, to illustrate absolutism, was another rhetorical hook. As if to say:

120

Evidentialism and the Will to Believe

Consider these indubitable thoughts … ah, but see how they can be doubted? Not so indubitable are they? Now, I think this is pretty nice rhetoric, and it does show our natural tendency to assent to the thought that there’s objective evidence. The trouble is that James’s rebuttal doesn’t hook up with his initial examples. Where’s the sincere and informed rebuttal to “two is less than three”? Who’s denied that he exists? James’s one example of the Hegelians rejecting Aristotelian logic had better not make his enthymeme that since all men are mortal, he’s mortal invalid, else that’s not much of a challenge. But James’s point can be made more broadly. The challenge to the absolutist can be posed in terms of asking what the criterion of objectivity of evidence is. The history of philosophy is littered with failed answers. James briefly surveys them: the moment of perception, revelation, cultured agreement, instincts of the heart, clarity and distinctness, common sense, and synthetic judgments a priori. Such a wide variety of standards: inconsistent, wild, fractious. Consequently, there should be doubt about the criterion, and if there is doubt about the criterion, then we should doubt its results. And if we are led to doubt the results, then there is, alas, no objective evidence. The much-lauded objective evidence is never triumphantly there; it is a mere aspiration, a Grenzbegriff, marking the infinitely remote ideal of thinking life. (WB 23)

The best we get, as James sees it, is that “practically” all we have is the “subjective opinion” that our evidence is objective. And that’s about it. And so we all start off as consciousness. But of what? And then the disagreement starts. We are all different: different histories, inclinations, objectives, and so on. And so we interpret it all differently. The empiricist, then, can even say: sure, there may be objective evidence, but all we have access to isn’t evidence naked and objective, but rather what we make of what we get, clothed as we see fit and subjective. An example may help to clarify the subjective evidential view. To put it baldly: there’s evidence and then there’s what we make of it. Recall our friends from Chapter 1—Karla and Ben. They are at the Cowpokes game. Karla, being a football fan, knows all the plays. Ben is there for the sausage. Now, they are both looking at the same thing, so they have the same visual evidence, and they may even have the same facts before them. The Cowpokes have the ball, it’s third down, they have seven yards to go, and the game is close. Karla, being knowledgeable about football, infers that the Cowpokes will be passing—third and long is a passing down, especially in a close game. Ben just knows that there will be some

Reading William James’s “The Will to Believe”

121

football-related mayhem. They both have the same evidence, but given their interests and background, they do different things with it. For something to be evidence for someone to make the right inferences and be justified in them, it needs to be interpreted, put to work, and set into cognitive motion. For Karla, it’s all evidence. For Ben, it’s an excuse to eat another sausage.16 The consequence, as James sees it, is that empiricism’s model for inquiry, one where we have no ground for certainty, only our best interpretations and tests, is our only option. That one sees certainty as impossible does not mean that one can’t still pursue truth. “We still pin our faith on its existence,” James reminds us (WB 23). As opposed to the closed absolutist’s system, the empiricist’s system is open-ended, one that is never completed, but in-progress and always up for revision. Empiricists are interested neither in the certainty of the hypotheses nor in the pedigree of the source, but in how well they serve the purpose before them. How the evidence can mean something, whether “the total drift of thinking continues to confirm” the hypothesis. And so, we have a rough pragmatist model for evidence—one that makes our tools for thinking features of our subjective and practical lives, but keeps the hope for truth alive.

3. Truth for empiricism Now the empiricist epistemological alternative would have been sufficient for the case against the absolutist, but James follows it with a final semantic turn on the debate. James reconstructs truth in the form of what his reconstructed evidence supports. The criteria for endorsement the empiricists take to be subjectiveinterpretive is not just a reconstruction of the notion of what justification is, in the wake of the failure of the model of objective evidence. Rather, James proposes this as a new view of truth: that the “total drift of thought … confirm(s)” a view is, so far, just evidence that it is true. But James’s closing independent clause of Section VI is that his drift of confirmation is what the empiricist “means by its being true” (WB 24).17 James’s program, up to this point, has been mostly either commentary on the psychology of beliefs or the classification of a variety of patterns of how people respond to evidence and choices. What we see at the end of Section VI is a new move from James—that of conceptual reconstruction. The natural distinction between truth and criteria for truth, or truth explainers and truth justifiers is collapsed with this move. It is consistent with the overall tendency of James’s program to reduce a feature of objective judgment to some element of

122

Evidentialism and the Will to Believe

psychology. Liveness of options seemed before reflection a matter of objective possibilities, but they were better seen as subjective inclinations to act or be moved. The same goes for evidentialism and absolutism—they are natural expressions of a certain kind of temperament, but that’s about it, so goes the Jamesian line. Of course, if truth just means that experience stays on a total drift of confirmation, then it doesn’t even require faith to answer the skeptic. It just requires the pragmatic reconstruction. Truth just became something relatively unorthodox, something no absolutist (or skeptic!) would recognize as the sort of thing that was originally being wrangled over. This is a bit of foreshadowing, as religion, too will be turned into something different so as to be saved. I’ll call this move the old switcheroo. James performs it here with truth, and he’ll do it later with religion. Whether the pragmatist reconstruction is appropriate should be weighed in different cases, but reconstruction regularly has similar problems, no matter what concept is under scrutiny. So it’s worthwhile to note a few things about the pragmatist reconstruction of truth here that might be considered objectionable by those who want truth saved, but see this salvation as perhaps not much better than not being saved. Objections to the pragmatist reconstruction of truth come along two lines. The first is the it’s not all about you objection. In short, truth takes care of itself, reality is what it is, and we can be right about it or not. The pragmatist denies all this—truth is all about us, it’s about getting what we want, feeling fulfillment. But that’s not truth, goes the objection, that’s narcissism. The second is the inconsistency objection. In short, truth is consistent, and reality doesn’t tolerate contradictions; there may be lots of good answers, but only one right one. But the pragmatist line is: because there are many subjective drifts of confirmation, there must be multiple (but inconsistent) truths. Pragmatists celebrate calling this pluralism,18 but, as the objection runs, it’s more a conflicted mess.19 Now, the point of this brief overview is that the pragmatist reconstructions of deep notions require that the notion be radically retooled from what it originally was taken as. This is sometimes a surprise to traditionalists, especially when pragmatists say they’ve saved these notions. That’s the switcheroo. Truth isn’t the kind of thing we’d held out for in the face of the skeptic, and once the Jamesian is finished with the reconstruction, we ask whether it was worth all the trouble. Many philosophers (me included) think the pragmatist reconstruction of truth makes truth look like the kind of thing we should positively beg for the skeptic to criticize and cause us to doubt and reject. The question is whether this problem of narcissism and inconsistency occurs with the target concept for reconstruction

Reading William James’s “The Will to Believe”

123

in “The Will to Believe,” namely religion. If James is guilty of the switcheroo with religion, then he’s not saved religious faith as religious faith, but as something else. We will return to this in Section X-6 and in Chapter 3.

Section VII—Two different sorts of risks in believing 1. The two commandments The second part of James’s meta-epistemology is that of articulating how the empiricist’s program is implemented. The best way to articulate the program is in terms of how it functions in relation to two deep norms of cognitive life that, as James notes, have not been as explicitly developed in the tradition as the norms of meta-knowledge seen in Sections V and VI. James identifies these two norms “our first and great commandments as would-be knowers.” They are: We must know the truth; and We must avoid error. (WB 24)

He restates the norms later as direct commands: Believe truth! Shun error! First, notice the difference between the two statements of the first law, the Truth Norm. The first is that we should know truths, the second version is that belief of truth is our objective. One question is whether this is a difference that makes a difference. Well, certainly if one thinks that knowledge is not mere belief, but a unique subclass of belief, those with sufficient justification and appropriate connection to the truth believed. A simple example from the history of philosophy can show the importance of this difference. Socrates, in Plato’s Meno, asks why knowledge is more valuable than true belief. One of his examples is called “the road to Larissa” case (Meno 97e). We can restate it here in the following format: You need to get from Athens to Larissa. There are two guides, A and B. A knows the way to Larissa, and B only believes the way, but B’s belief is true. So both guides can get you to Larissa, but it seems we’d prefer A’s guiding. Why? A true belief is as good as knowledge when it comes to getting the truth and getting what we want. So what distinguishes knowledge form true belief, and what about that makes knowledge seem to be more valuable than true belief? Socrates’ begins his answer by making an analogy. The great inventor Daedalus made some moving statues, and they run all over the place. They are valuable, but one you have on a tether is more valuable, because it’s with you

124

Evidentialism and the Will to Believe

instead of running around willy-nilly. Beliefs are like that, too. Knowledge is a species of true belief, so it is valuable, but it has explanatory reasons in its favor. And these reasons work like the tether, which keeps us reliably striking upon the truth. The tether of reasons keeps truth nearby. It is something we can depend on, instead of mere true belief, which though valuable, isn’t tethered.20 This distinction between knowing truths and merely believing them is important, at least for the traditional take on the matter. So why does James switch the Truth Norm from know truths to believe them? From the traditionalist’s perspective, such a move would be retrograde. The answer is that James’s case, ultimately, is that it is being right that matters in the end. Consequently, true belief, whether mere true belief or with the explanatory account working as a tether that makes it knowledge, is what matters. Knowing the truth is one way to come to it, but there can be other means, too. Those who think that knowing is the only way to come to truth, those who, like Socrates, put the emphasis on the certainty that one will have it, are wrong. Truth doesn’t always have a tether. In fact, with deep, important truths, they are never tethered by reason, or they must be grasped first to be understood. So we have, given the overall arc of James’s argument, a reason to take the belief form of the Truth Norm as canonical for James. The knowledge form is one closer to the heart of the tradition of epistemology, but James is out to answer and correct that tradition. The two norms, then, are Believe truth and Shun error—what I’ll call the Truth Norm and the Error Norm. On their face, they sound identical, but they, when put into practice, are not. James makes the case that following the Error Norm is not equivalent to following the Truth Norm. He gives two arguments for this nonidentity. The principle of distinction for both of James’s nonidentity arguments is a version of the nonidenticality of discernibles. Roughly: if, in following one norm, one runs afoul of another, the two norms are not identical. So if James can show that in following the Error Norm, one fails to follow the Truth Norm (or vice versa), he’s shown the two are not identical. James proposes two examples of conflict to show the two “materially different.” James’s first case is that we imagine a circumstance wherein we are deliberating about B. Let B be some take on the situation that is an error, and perhaps even an obvious one. Let A be the truth of the matter. Here, we, given that B is erroneous, would in following the Error Norm, not believe B. Now, at this point, James rightly notes two things. First, he notes “it hardly ever happens that by disbelieving B we necessarily believe A.” This is right, at least as far as

Reading William James’s “The Will to Believe”

125

avoiding error goes, as it is clear that if one is out to avoid error, one does not move from the negative case against one option to take it as a positive case for another. James’s second point is that after avoiding error with B, we may “fall into believing other falsehoods, C or D, just as bad as B” (WB 24). Now, to be sure, one could, in avoiding this error (in B), fall into that error (in C or D). But this is not what we would take the injunction of the Error Norm to be (generally). That is, an example of avoiding one error but in doing so committing another is not a case of avoiding error. It is a case of avoiding a specific error, but not error generally. So I’m not convinced that James’s first case actually shows the two norms are different, as, recall, the cases must show that in living up to one norm, one fails at the other. For sure, James has shown that in avoiding error B to thereby commit error C or D, the subject in question has missed out on the truth in A. But the subject hasn’t avoided error. So the first argument is pretty much garbage. James’s second case is considerably better. James presents it with a single sentence, but this is the model for the rest of the argument of the essay. The case is simply that the agent, in seeing that B is an error, but has no evidence for A (or C or D), suspends judgment with regard to all the options. “[W]e may escape B by not believing anything at all, not even A” (WB 24). This is a legitimate case for material difference, as our agent in this circumstance does avoid error by not committing to B, C, or D. But our subject does not live up to the Truth Norm, because A is true, but the subject does not believe it. And so we have a dilemma, of sorts. Both of the two norms seem right about our cognitive lives—we desire the true and despise the false. And James, despite the shaky start, has shown these two inclinations, which seemed to be different ways to express the same orientation toward our cognitive lives, actually yield materially different outcomes. They, then, are not only not identical, but they also conflict. What’s to be done in the case of a conflict of two norms?

2. The case for the Truth Norm Despite the fact that the Truth Norm and the Error Norm seem similar, they are different and conflict. How do we arbitrate the conflict? There, it seems, is no higher norm to appeal to, as James holds that these are the first commandments. Were they dependent norms, ones determined by an ordinally prior principle, we could determine which more effectively manifests that norm. Instead, we must look to other means to decide which norm has precedence for us.

126

Evidentialism and the Will to Believe

James’s argument will be that the matter of which norm we take to live by in these cases of conflict is determined by our passional nature. Just as we are (both descriptively and prescriptively) natural overbelievers with regard to first-order cases of undecidable intellectual commitment, we are also naturally committed to the second-order cases of undecidable epistemic norms. The second move is to show that if it is a matter of acceptable choosing, then there are preferable choices to make over others. And so even if one has the right to choose the Error Norm over the Truth Norm, one would do better to choose the Truth Norm. James presents the consideration of how one’s intellectual character is displayed with either of the two norms. He notes that “by choosing between them we may end by colouring differently our whole intellectual life” (WB 24). How we choose is a function of what kind of thinkers we want to be. James turns to Clifford and his evidentialist scruples driving him to agnosticism, suspension of judgment. He diagnoses this tendency of valorizing the Error Norm over the Truth Norm as a response to a kind of fear: Believe nothing, he tells us, keep your mind in suspense forever, rather than by closing it on insufficient evidence incur the awful risk of believing lies. (WB 24)

The inclination to forego belief forever, rather than to believe a lie, to be in error, is the product of what James identifies as a fear: the “preponderant horror of becoming a dupe” (WB 25). The horror of dupery, the indignity of error, the shame of should-have-knownbetter—these are the motivating factors driving the evidentialist’s project. That is what James sees the Cliffordian program posited on, the thought that error is in itself costly and horrible. This leads James to an analogy between marshaling one’s beliefs and a general preparing his soldiers for battle. It is like a general informing his soldiers it is better to keep out of battle forever than to risk a single wound. Not so are victories over enemies or nature gained. (WB 25)

The force of this analogy is that those who are risk-averse will be cowards. Not only will they fail to get the good things—victories over enemies and the truth— but doing without these things yields a serious defect in character: “Excessive nervousness” (WB 25) and a kind of “sickliness” (WB 7). The contrast is with the character expressed by committing to the priority of the Truth Norm. In committing to the priority of the Truth Norm, one is, on the analogy with the general, expressing courage, vitality, bravery in the face of risk. James, later in “The Will to Believe”, endorses faith as an expression of “courage” and action over reticence and indecision (WB 32).

Reading William James’s “The Will to Believe”

127

The calculation must be that the losses one can incur are worth the risk, especially given what one gains when one is right and what one loses when one is wrong. There are costs to playing it safe. Comparatively, the judgment is that “the risk of being in error is a very small matter when compared with the blessings of real knowledge” (WB 24, emphasis mine). Notice that knowledge has returned, and it is not clear why. Knowledge has classically required justification, and by hypothesis, the cases James is considering don’t offer that. Just truth. He follows the first contrast with a follow-up that one, then, should be “ready to be duped many times in your investigation rather than postpone indefinitely the chance of guessing true” (WB 24–25, emphasis mine). Again, James is moving back and forth between knowledge and true belief, and certainly those defending or inclined toward evidentialism would require more care. My interpretive hypothesis, again, is that James is ultimately out to make the case that there is a class of true beliefs that when held properly, can yield knowledge. But, ultimately, what matters in all the cases is truth. One takes a risk, and when one has correctly guessed, one has the truth. Once one has the truth, one may have the means to understand and justify it, but only after one’s gotten it. And in many of these cases, the belief will not only be a way of grasping that truth, but also contributes to making it. The discussion of doxastic dependence will have to be tabled until Section IX, but it is important that matters of faith, as I’d noted earlier, are not truths that would be properly tethered by reason beforehand. Again, the case for the preferability of the Truth Norm must be made comparatively, and the first half of the job is done with portraying the Error Norm as an expression of nervousness, cowardice, and sickliness. The second half of the job is to show that following the Truth Norm does better. This begins with James holding that errors aren’t really that costly. A kind of lightheartedness with falsity is necessary. It’s OK to make mistakes—don’t let your fear of saying something false prevent you from saying anything at all. And so James holds that error is a “very small matter,” we must approach risks with our chins up and actively, not with paralyzing fear. We must see that: Our errors are surely not such awful things…. [A] certain lightness of heart seems healthier than this excessive nervousness on their behalf. (WB 25)

And the comparison is complete: the Error Norm is an expression of excessive nervousness and the horror at being a dupe. And this is a “Fear (the evidentialist) slavishly obeys.” This yields suspension of judgment and inaction. The Truth Norm, by contrast, expresses light heartedness and healthy mindedness. This

128

Evidentialism and the Will to Believe

yields a willingness to take risks, a kind of intellectual courage.21 It is a matter of where our passions and nature take us, James notes. [T]hese feelings of our duty about either truth or error are in any case only expressions of our passional life. (WB 25)

The case, as James has made it, is not that following the Truth Norm is obligatory, but that it is permissible and preferable. Those whose natures are inclined toward cowardly sulking or are deeply afraid of error to the point of never believing anything—they can follow the Error Norm. And that would be fine for them. But it being fine for them does not make it fine or obligatory for everyone else. Others, as James expects his audience to agree, are less risk-averse. And so the courageous and active path calls to them. To the empiricist, James holds, the Truth Norm is more “fitting” (WB 25).

3. Two critical points There are two evaluative questions to ask about this culmination of James’s metaepistemological stage. The first is a dialectical question: how salient are these criticisms of Clifford? If James is making a comparative case, the comparisons must be accurate, so it’s worth asking how well he’s done in that regard. The second question is a purely first-order philosophically hermeneutic question, which is what kind of meta-epistemology is on offer here? Let us first consider the dialectical question. Surely if you, the reader, have read Clifford’s “The Ethics of Belief ” at all, and especially if you’ve not just skipped to my commentary on James and left my treatment of Clifford to the dust, you can easily see that Clifford isn’t being treated very fairly. That’s, actually, an understatement. James has constructed a straw man. A pretty shameless one, to boot. To show this, it doesn’t take much arguing, and I’ll limit myself to three places of pushback. First, there is the point about fearing error. Clifford’s case for fearing error isn’t that he’s got, as James calls it, a “preponderant horror of becoming a dupe.” It’s because errors are practically costly. Recall the ship owner case. The error cost wasn’t dupery. It was dead emigrants. Recall the island case. The error cost wasn’t a false belief. It was that people had baseless accusations of kidnapping lodged against them. Recall the slope arguments: the error costs weren’t just false beliefs, but that those beliefs magnified errors and consequent costs in group settings. For sure, it’s horror at error, but it’s a moral horror at what consequences error

Reading William James’s “The Will to Believe”

129

brings with it. James, in portraying it as a matter of “private horror” has gotten Clifford’s reasons for the view utterly (but conveniently) wrong. Not a good start. Second, Clifford does allow for action when there is insufficient evidence. James’s portrayal of the Cliffordian as one who just waits around and is frozen in place by skepticism is simply incorrect. Remember, sometimes there are default actions when one does not know, but nevertheless must act. So the ship’s owner should requisition an inspection. The island agitators should do some investigation. They do not “keep (their) mind(s) in suspense forever,” or “stay out of battle forever,” as James has portrayed them. Rather, Clifford’s case is that one can act in tentative ways, often to collect evidence, to behave responsibly, to be safe. Suspension of belief is not inaction for the Cliffordian. Again, I believe the contributing cause of this feature of the straw man here is the pragmatist dogma of converting All belief directs action to All action is directed by belief, which I will criticize later as the conversion fallacy in Section X-7. Third, and finally, it seems that James’s case proceeds on the evaluation of certain belief-regarding characteristics of people. One expresses a virtue, of sorts, that of courage and health, the other a vice, that of slavish fear and nervousness. But if the matter really is one of evaluating the features of character here, perhaps it’s worth hearing from a Cliffordian on the matter. The following is what I call the sauce for the goose answer to James. In effect, he wishes to paint a picture of the evidentialist’s character and use it as an argument. Perhaps we should see what an evidentialist’s picture of his character looks like. Maybe the argument will be a bit different after that. What does a light-hearted attitude toward error yield? Cavalier and irresponsible assertion, reckless and self-indulgent speculation. Noxious bullshit. What does care to avoid error yield? Careful and studious assessment of the evidence, prudent decisions, and safe paths. Responsible belief management. Who do you want as your general, head of state, or teacher of your children?

What I think the sauce for the goose reply to James yields here is that the problem of deliberating between the two norms of belief management is not as easy as we’d seen. My thought is that James is right that we often must choose between the two, but that it’s not a wholly simple case that the Truth Norm always wins out, or that when it does, it’s not so easy a win. The reality is that we must often satisfice between these two norms, sometimes tentatively hypothesizing and thereby risking error, but with the hopes that we gain truth. We do sometimes make errors, but we must be aware of what comes with those errors in order to responsibly take them. I may have the hypothesis that my daughters often

130

Evidentialism and the Will to Believe

need spankings. This hypothesis often occurs to me and looks live and electric around 7 PM when they are misbehaving, screaming, I’m tired, and I just burned dinner. But what if I’m wrong and corporal punishment messes kids up? What if it messes me up, too?22 The error costs are between brutalizing your kids for no good reason and thereby scarring yourself and them or raising brats. I’ll opt for the latter. Moreover, I think I can avail myself of other means to reduce their brattiness. Surely no Jamesian would be light hearted about the errors of spanking, but we do desperately need the truth, not just error-avoidance. The closing critical question of this subsection is: what is the nature of James’s meta-epistemology? My rough answer, following Jeff Kasser’s recent work (with Shah in 2006 and forthcoming), is that James is an expressivist about epistemic norms. This line of thought runs that James’s argument is that our feelings of duty are only expressions of our passions, they are manifestations of our broader willing natures. Consequently, the Jamesian case against evidentialism is not that it is self-defeating (as we’d seen the case run before regarding the Cliffordian argument in Chapter 1, Section X-5), but that it is held on the basis of sentiments that themselves are reasons in a natural-explanatory sense. The case against Clifford is not so much one of inconsistency, but one of identifying the passional grounds for the view and evaluating them (Kasser and Shah 2006: 5). That is, James’s case for the Truth Norm over the Error Norm is not one from reductio, but that the correctness of epistemic rules is determined by what passions we have, what desires motivate us. Once we identify the desire that makes the respective belief policies appealing, we can ask ourselves which desire is more appealing, appropriate, and fitting. The case, then, given this expressivist meta-epistemic reading, allows us to be able to make sense of the thought that James’s ad hominem abusive case against the character of those for whom evidentialism appeals can show that one should reject evidentialism. The case being, again, that evidentialism is the view of cognitive cowards. It does not show that evidentialism is false. Ad hominem arguments don’t do that. What they do show is that if the connection is right, we have reason not to accept a norm because of the kind of person it makes you. We have the right to be timid and sickly, but it’s just better not to be. And so the healthier option is to follow James in developing the doctrine of the will-to-believe. The benefit of the expressivist reading here is that we are able to find logical space between two unacceptable extremes. On the one hand, the thought of the right-to-believe is appealing as an interpretation, as it is a defense of faith as a legitimate form of believing, but it does not have any of the valences of endorsing it, which we certainly see, even in James’s double-use of the term “natural.” On

Reading William James’s “The Will to Believe”

131

the other hand, we have the duty-to-believe reading, which seems to get us the endorsement of faith the right reading lacked, but it is far too strong, as James concedes that the evidentialist’s view with regard to many matters is healthy and he regularly invokes the thought that those opposed to his view have a “right” to their lifestyles (WB 23). This strategy, then, avails us of the double category of being (a) allowed, and so one has a right to believe according to the two belief policies, but also of being (b) preferable, as one policy expresses characteristics we desire, the other does not. If we are to take the line that James has shown that evidentialism is false, then we are driven into the more demanding waters of the duty to believe. With the expressivist reading, then, we find the right modest valence of justification for James’s case for faith and we further integrate this with his view that the decision is still a passional one.

Section VIII—Some risk unavoidable 1. Applying the meta-epistemology James opens Section VIII by noting that all seven previous sections (of the ten for this chapter) have been introduction. The layout has been twofold. First is the psychological point that “as matters of fact, we find our passional nature influencing us in our opinions” (WB 25). As noted earlier, we are naturally inclined to believe, and this inclination is still present even in cases wherein our evidence is insufficient. The second part of the introductory layout is that, further, “there are some opinions between options of which this (psychological) influence must be regarded as both inevitable and as a lawful determinant of our choice” (WB 25). This is to say, we see these choices as natural in both the descriptive and normative senses of the word. The evidence, so far, for the normative thesis, that we have lawful determinations of will on belief, has been in our choice between the Truth and Error Norms. We not only lawfully may, but also must choose between them, and the one we choose is a function of our will, broadly construed. This is the lesson of the meta-epistemology in Section VII. Our policies of belief management are expressions of our deeper temperamental character. The question is whether the extent of the cases wherein will lawfully determines belief is restricted entirely to the level of meta-epistemology. That is, James’s argument has only shown that at the very deepest orientations toward belief policies do our passions rule. But it needn’t follow from the fact that meta-

132

Evidentialism and the Will to Believe

epistemology is an expression of our passions that our first-order epistemological norms are passional, too. Once our passions determine our orientation toward the priority of Truth Norm over Error Norm (or vice versa), passions may play no further role. As James pauses to consider this objection, he frames his objector as conceding the meta-epistemology, but holding that the best means to either of the two choices is to refrain from will to make determinations: “but the surest path to those ideal consummations, you will probably consider, is from now onwards to take no further passional step” (WB 25). The objection, as James sees it, is that the norms of meta-epistemology may be determined by our wills, but the norms themselves may not yet endorse further passional judgment. So the evidentialist, say, can concede that, yes, defaulting to the Error Norm is an expression of her passion, but once you endorse the Error Norm, it’s better to decide things by evidence. Alternately, one can imagine someone taken by the Truth Norm and who concedes that this is the expression of his natural temperament, but he holds that the pursuit of truth is most effective when one does not pursue it on the basis of one’s preferences or will. Call this the Levels Objection, which can be stated simply as: the fact that meta-epistemic matters are decided on a passional basis does not imply that firstorder epistemic questions are. James concedes that, in fact, the Levels Objection is broadly correct. He says, “I agree as far as the facts allow.” He considers two exemplary cases that show that the levels are, in fact, distinct. They are cases that are regarding either nonmomentous, non-living, or non-forced options. So long as the cases are ones where the matter is one bearing on what James calls “scientific questions,” we can suspend judgment and “wait till objective evidence has come” (WB 26). This is to say, so long as the matter is not up to us, and not a matter of our preferences, it is worth suspending judgment. [I]n our dealings with objective nature we obviously are recorders, not makers, of the truth; and decisions for the sake of acting promptly and getting to the next business would be wholly out of place. (WB 26)

With regard to objective nature, the facts are the facts, and they remain so regardless of our attitudes toward them or our beliefs about them. James then holds that as a consequence, these matters are neither live nor momentous, nor forced: The questions here are always trivial, the hypotheses hardly living (at any rate not living for us as spectators), the choice between believing truth or falsehood is seldom forced. (WB 26)

Reading William James’s “The Will to Believe”

133

James poses a series of examples to illustrate these features. Consider the matters of Röntgen rays, the belief in mind-stuff, or theories of mental causation. James holds that “it makes no difference to us,” at least to his audience, how we come down on any of these matters. In all cases, it is worth continuing to weigh reasons “pro et contra with an indifferent hand” (WB 26). James here has conceded that with the logical class of issues that are nonmomentous, nonliving, and nonforced, the form of evidentialism endorsed by Clifford is appropriate. Regardless of how one comes down either the Truth or Error Norm, one should nevertheless be an indifferent inquirer until all the evidence is in. The inquirer with partiality, the thinker with a preference and an inclination to be convicted on the basis of his will on this order of deliberation is in error. And so, again, James’s initial concession is that Cliffordian evidentialism and the requirements of evidence are healthy is repeated. And James, thereby, has conceded that with a large majority of cases, the levels distinction is appropriate. But he has a special class of exceptions.

2. Interested inquiry Recall that James’s essay proceeds in four stages: logical, psychological, metaepistemic, and epistemic stages. In the case for interested inquiry, we see the first application of the first three stages to yield the fourth. James has shown that psychologically, we do decide many options on the basis of our will, and that the meta-epistemological options of Believe Truth! and Shun Error! are lawfully decided by temperament, but he has conceded many first-order options are not so decided, specifically what he calls “scientific” matters. However, James now turns to argue that in the remainder of the cases, the interests of the inquirer are relevant. At first, James makes the case by contrast with the impartiality of the ideal Cliffordian inquirer: [I]f you want an absolute duffer in an investigation, you must, after all, take the man who has no interest in its results: he is warranted incapable, the positive fool. The most useful investigator, because the most sensitive observer, is always he whose eager interest in one side of the question is balanced by an equally keen nervousness lest he become deceived. (WB 26)

James’s contrast is that between the disinterested inquirer and the interested inquirer. The former is a dull and inattentive “duffer,” the latter is a keen and “sensitive observer.” James’s pushback against the Cliffordian prohibition on

134

Evidentialism and the Will to Believe

interested inquiry, inquiry begun when the inquirer has an antecedent belief or preference for outcome, is that Clifford is wrong about the way an interested party can be attentive to evidence. There is something to the contrast. If you are not interested in the issue, then the details of the matter will not have the same force or vivacity as they would have for someone embroiled in it. Interest directs attention, and attention is a condition for directing and detecting the truths of the case. However, I am concerned about James’s case here. For two reasons. First, interest in outcome regularly has deleterious effects on how people interpret their evidence. The fact of confirmation bias is widely known, and those who believe in a solution prior to investigating a problem overwhelmingly are sensitive to supporting evidence for their pet theories, but are notoriously bad at finding or even acknowledging counterevidence. So James is right about one part of interested inquiry yielding “sensitive observers” for the preferred view, but he’s wrong that this interest is “balanced” easily with counterevidence. If one lives by the sword of facts of human psychology, one dies by that sword. My second worry is that James’s case for interested inquiry proceeds from an equivocation on “interest.” There are two ways one may be “disinterested” or show “indifference” with regard to an issue. On the one hand, one may have no interest at all in the matter, and so prefer not to even attend to it. Call such indifference general disinterest. On the other hand, one may not have preference which party to a dispute or hypothesis comprising an option wins out, but nevertheless one wishes to know which in fact does. Call this specific disinterest. Notice how James’s case involves an equivocation on disinterest between these two. To start, James’s observation is that disinterested inquirers are “duffers,” but this would be correct only for those with general disinterest, as they won’t care for the issue überhaupt and won’t be sensitive to the evidence. But note that this is not the case for specific interest, as, again, specific interest has its own distortions, and we prefer our judges not to have prior conflicts of interest. However, when James turns to the performance of interested inquiry, his cases are those of players with “pet hypotheses,” and that in those cases, “human passions are stronger than technical rules” (WB 27). It is the one with “eager interest in one side of the question” that is the “keen observer” (WB 26). This, note, is not a univocal case of general interest and specific interest, one with interest in one side of the issue, not the issue generally. So to make the case that interested inquiry is better than disinterested inquiry, James equivocates on two kinds of interest. If he remains univocal on general

Reading William James’s “The Will to Believe”

135

interest, the view is uncontroversial and breaks no ground against the Cliffordian evidentialist. If he remains univocal on specific interest, he must commit to the false proposition that prejudice makes one a more reliable (instead of merely more decisive) judge on an issue.

3. Two analogies To close the section, James makes the same concession with which he opens it: for most cases, particularly ones wherein the issue is neither live nor forced, nor momentous, it is best to be an indifferent inquirer. But in making the concession, James carves out a particular ground for the exceptions to come. The concession comes in the form of two analogies. The first is with baseball: [H]owever indifferent to all but the bare rules of the game the umpire, the abstract intellect, may be, the concrete players who furnish him the materials to judge are usually… in love with some pet “live hypothesis” of their own. (WB 27)

The ideal, he concedes, is that of “the dispassionately judicial intellect”, so long as there is no forced option. There are two things to note about the baseball analogy. First, issues are no longer strictly “scientific,” in that the facts of the mater in these cases are up to humans. Baseball games are comprised of the interested actions of players. This is a different class of issue than the question of Röntgen rays that James had proposed before. Second, the call of umpires has (at least in the days before instant replay) in fact made the facts. There is an old joke about umpires taking pride in their objectivity, and in the process the notion of their objectivity seems in flux. #1: I calls ‘em as I sees ‘em. #2: I calls ‘em as they is. #3: Until I calls ‘em, they ain’t. The point here is that James is clearly making a move from matters not up to us as inquirers, interested or otherwise, to ones that are. The specifically disinterested umpire (whom we presume is generally interested enough to make the calls when appropriate) is the ideal judge for the games. The calls and plays aren’t, until he makes them (as the joke goes), but the game is an ebb and flow of interest and action. James’s second analogy is that of the boarding house table, set and prepared for us. Expecting evidence to come ready for the well-positioned judge is a

136

Evidentialism and the Will to Believe

kind of idealization, borne of dream, not reality. Just as it would be to expect in boarding houses that the food be prepared palatably: It seems a priori improbable that the truth should be so nicely adjusted to our needs and powers as that. In the great boarding house of nature, the cakes and butter and the syrup seldom come out so even and leave the plates so clean. (WB 27)

We, when we eat the cakes, want the right amount of butter and syrup for them. Each one. But sometimes, there are more cakes than butter and syrup, sometimes more syrup or butter, but all too often, it’s too few cakes. And when it’s time for cleanup, who would not be suspicious when there’s not a sticky mess of plates? The point is that reality, even when it’s up to us (we make the cakes and procure the butter and syrup) is a mixed bag, and we must act to accommodate those facts and unhappy proportions of cake to syrup. That the cakes, syrup, and butter ought to come in proportion is not a reason to expect them to actually come out that way. It is, on the analogy, up to us to make do in and make the best of the regular times when they don’t. Consequently, even when being served the cakes, one must be interested in outcomes and thereby be active. And so one must be not only interested but also active as an inquirer, one that arranges what is given so as to construct the best (or at least preferred) outcomes.

Section IX—Faith may bring forth its own verification 1. Moral and scientific questions James has made a concession in the previous section—that with regard to “scientific questions,” one’s will should not function in determining belief. However, with this concession, James’s plan is to make gains on the questions that are not in that domain. The specific class of issues James is concerned about is what he calls moral questions. James has three points of contrast between the two types of questions. They are: ●

●

●

[I]n our dealings with objective nature we are obviously recorders, not makers, of the truth… (WB26) Moral questions immediately present themselves as questions whose solution cannot wait for sensible proof. (WB 27) A moral question is not a question of what sensibly exists, but of what is good, or would be good if it did exist. (WB 27)

Reading William James’s “The Will to Believe”

137

These three points of contrast yield the following schemata: A question (Q) is a scientific question when those who ask Q are recorders of what sensibly exists, and in cases where there is insufficient evidence, one may suspend judgment (one may wait for sensible proof). Q is a moral question when those who ask Q are makers of what is good, and those who ask Q may not suspend judgment (one may not wait for sensible proof). Scientific questions, given the account are of what “sensibly” exists, which, it seems, is a form of positivism with regard to the notion of existence. But the is answers of science do not focus the ought questions approached by moral reflection. In those cases, as James says (invoking Pascal), we must consult “our heart.” A second matter of the contrast between the two kinds of questions is that of the role the subject asking the question has with regard to the answers. With scientific questions, James emphasizes the role being a recorder of truths. With moral questions, the invested subject not only has an interest in the issue generally, but also interest in specific outcomes (as James has argued before), but that interest may play a role in making the truths. This element of the contrast between the two classes of questions has not yet come out in any of the previous introductory material. But it will be of real significance going forward. The first element of the contrast between scientific and moral questions can be accounted for by the logical distinctions from Section I of the essay—scientific questions are not live, not momentous, and not forced; moral questions are live, momentous, and forced. But note that the makers-of-truth element of the difference between the two classes is not captured by the initial logical elements of Section I, and it is this element that James will make into a case for religious faith. Just as James’s case from the previous section sets up an argument for invested inquiry, this added component of moral questions will be a part of not only invested but also productive inquiry.

2. Moral skepticism James’s first example of moral questions are what he holds are “wide questions of good.” They are challenges posed by the moral skeptic, who asks whether our moral judgments are true or whether they are merely “odd biological phenomena.” Moral skepticism “can no more be refuted or proved by logic than intellectual

138

Evidentialism and the Will to Believe

skepticism can” (WB 28). The first question is whether the challenge posed by the moral skeptic is a moral question in the sense that James distinguishes them from scientific questions. It surely is live, as, at least once one understands the question, the issue becomes important. In fact, the challenges of intellectual and moral skepticisms are tried-and-true means for instigating philosophical reflection. The early Platonic Dialogues (for moral questions) and Descartes’s Meditations (for intellectual) are powerful ways to easily motivate the work of philosophy. The momentousness of the matter is clear, at least as questions that bear on the direction of one’s life. And, finally, there is the question of forcedness. James never directly addresses the forcedness of the matters, but here might be a rough way to capture the thought, now restricted to moral skepticism: With the moral skeptical challenge, the options are between: (a) hold that what we take as moral reality is an illusion, (b) hold that we are correct about moral reality, or (c) suspend judgment. So the option is O: (a, b, c). However, because suspension yields the same behavior as belief that it is all illusion, these two options collapse into a functional equivalent, φ. So, the skeptical challenge, really, poses the following forced option O: (a, φ).

I am presuming that James has pressed suspension and the negative commitment together, as his discussion ranges from speaking of the skeptic having “knowingness” to explaining why some subjects do not take the “moralistic hypothesis” on account of their “cool-hearted” natures. Given that the option is forced, then, one must decide. James’s case is that the “questions of having moral beliefs at all or not having them is decided by our will” (WB 28). On the one hand, again, those with “cool-hearted” natures will feel no pull for taking a stand for moral reality, and so because “your heart does not want a world of moral reality, your head will assuredly never make you believe in one” (WB 28). And, on the other hand, even in the midst of having such a skeptic’s critical eye upon him, the “hot young moralist” will cling to the view that there is a realm of moral facts. The same, he holds, goes for the intellectual skeptic, whom he terms “Mephistophelian,” whose “head’s play instincts” yearn for satisfaction, instead of other inclinations. As we’d seen before, in Section III-3, these are matters that psychologically are decided by our wills, and, given James’s case so far, are lawfully so, too. The question remains whether James actually holds these cases to be also ones where one is, instead of being a mere recorder of reality, a maker. On first blush, this would be absurd. One does not make a world, or construct moral truths in so-believing. That’s what the skeptical challenges were actually sayingbelieving does not make things so. But the overall cause of the collection of essays

Reading William James’s “The Will to Believe”

139

comprising The Will to Believe bears this reading out. Particularly with moral beliefs. As we see, later, with “Moral Philosopher and the Moral Life” (delivered five years later to the Yale Philosophical Club), James is a phenomenalist with regard to values; and so, to maintain the commitment to them, even when the issue has no determining evidence, he must reconstruct the concepts of moral evaluation. James begins from a pragmatic analysis of fundamental normative concepts, contending that “Goodness, badness, and obligation must be realized somewhere in order really to exist”; he then asserts that the “only habitat” of these moral properties is “a mind which feels them” (WB, 145). Hence moral terms apply to certain psychological states. “[T]he words ‘good’, ‘bad’, and ‘obligation’ ” refer not to “absolute natures” of acts but instead are “objects of feeling and desire” that “have no foothold or anchorage in Being, apart from the existence of actually living minds” (WB 150). Consequently, James holds that “nothing can be good or right except so far as some consciousness feels it to be good or thinks it to be right” (WB 150). To say that x is good is to report that someone in fact thinks it good, or actually desires it. The same goes for obligation. A moral obligation arises, James says, only when “some concrete person” actually makes a claim or a demand that something or other should be done (WB 149). Indeed, “The only possible reason there can be why any phenomenon ought to exist is that such a phenomenon actually is desired” (WB 148). James concludes that “the essence of good is simply to satisfy demand” (WB 153). Hence, he identifies the meaning of moral claims such as, One ought never to lie, as expressing a demand that one might make always to be told the truth; one makes such a demand on the basis of one’s desire to not be lied to. The good is the satisfaction of desire. Despite appearances, James’s view is not a variety of ethical egoism. The egoist ties moral value to one’s own interests, but James recognizes the moral status of any demand, even those issued by others. Whereas the egoist identifies the right action in any given context with that action which best furthers the interests of the agent, James acknowledges that “every… claim creates in so far forth an obligation” (WB 148). Consequently, on James’s view, the interests of others can be the source of our obligations, regardless of whether it furthers our interest to satisfy them. In this way, James’s ethics involves an egalitarian component: every demand and every interest, no matter whose demand or interest it may be, morally counts. James bids us to satisfy “as many demands as we can” (WB, 158). Call this injunction to satisfy as many demands as possible James’s meliorism. The takeaway, as I see it, is that James ultimately does think that the skeptical issues are moral questions wherein the options posed are ones appropriately

140

Evidentialism and the Will to Believe

determined by will and are occasions for interested subjects to be more than recorders, but makers of truths. This account requires not only a demonstration that the options live up to the logical conditions outlined in Section I (and thereby amounting to genuine options), but also matters wherein our believing, willing, and acting is something that contributes to the truths we inquire about. These commitments I will call doxastically efficacious, and James’s further examples will clarify this notion.

3. The argument from friendship From the heavy philosophical questions, James turns to “a certain class of questions of fact, questions concerning personal relations” (WB 28). That is, questions regarding friendship and its benefits. James asks us to consider the question, Do you like me or not? We will assume that we are considering this question when we do not have sufficient evidence, and we will also assume, further, that the matter is of some import (else it would not be live). So let the question be posed regarding a person whose estimation of you matters (as opposed to one about whom one does not care). James’s overall argument with the friendship case is that there are momentous options where we are not merely recorders but makers of truths. And, consequently, the kind of faith that creates the fact of friendship is not allowable by the evidentialist requirements, but is itself lawful and even possibly indispensable. James’s central view is that antecedent faith is requisite for friendship, and his argument proceeds as follows (I have numbered the sentences here for ease of reference in discussing the argument): (1) Whether you do or not [like me] depends … on whether I (a) meet you half-way, (b) am willing to assume you must like me, and (c) show you trust and expectation. (2) The previous faith on my part in your liking’s existence is in such cases what makes your liking come. (3) But if (a) I stand aloof, and (b) refuse to budge an inch until I have objective evidence, (c) until you have done something apt … (d) ten to one your liking never comes. (4) The desire for a certain kind of truth here brings about that truth’s existence…. (WB 28)

For the sake of discussion, let us say person A has asked the question about person B—does B like A? A is asking with regard to this proposition (L) whether it is true or false: (L): B likes A

Reading William James’s “The Will to Believe”

141

James is making the case that L’s truth depends on A’s antecedent belief that L. Let us call such propositions doxastically dependent—where either part of the truth maker for the proposition is a believer’s attitude toward it or that a believer makes the proposition more likely with her attitude. I will work to clarify this concept a bit more in subsection 5, but for now, let this be the rough-and-ready notion. Sentence 1 is a commonsensical observation—people tend to like others that treat them nicely. So, with 1a, A’s meeting B halfway is that of conversational charity, openness to compromise, perhaps even a willingness to share a space. Similarly, trust and expectation in 1c amounts to the friendly gestures of A making plans with B, A allowing B to be aware of some vulnerability, or A promoting cooperative ventures with B. These are conducive of sentiments of goodwill. The key is A’s assumption of L in 1b. James follows sentence 1 with 2 such that because of A’s faith, L is more likely—so 1b explains why 1a and 1c come to be at all. That is, it is because A believes L that he enacts the conversational charity and cooperative virtues. The belief is why A acts the way he does. What follows is that if A believes P, A has reason to believe that he will act in a way that is productive of P’s truth. So there are two principles running the argument: The Principle of Niceness (PN): If A is nice to B, L is more likely. The Principle of Assumption (PA): If A assumes L, then A is more likely to be nice to B than if L did not assume L. The second half of the argument is that A’s not believing that L in 3b and 3c yields actions that undercut friend-making. A will, on 3a, stand aloof. James continues the thought as follows: How many women’s hearts are vanquished by the mere sanguine insistence of some man that they must love him! he will not consent to the hypothesis that they cannot. (WB 28; emphasis in the original)

There are two elements to this thought. The first is that the argument from friendship has a more specific instantiation in cases of romantic conquest. James extends the romantic conquest case in Section X as follows: It is as if a man would hesitate indefinitely to ask a certain woman to marry him because he was not perfectly sure she would prove an angel after he brought her home. (WB 30)

I will address the case for forcedness later, but at this point, it is worth highlighting the specific form that the friendship argument shares with courtship cases. James’s argument is that the antecedent assumption that the romance is

142

Evidentialism and the Will to Believe

worthwhile and is budding not only is a necessary condition for such buddings, but also contributes to them directly. Moral questions such as these, then, are comprised not only by the fact that they are live, momentous, and forced, and have preferred outcomes (friendship and marriage) to others (loneliness), but our attitudes contribute to which of those two outcomes is resultant. The second feature of the romantic conquest cases is the epistemic element. Here the “must” of James’s first mention and the “perfectly sure” are epistemic assessments of the interested parties in what they take the situation between themselves and their target is—the belief does not have sufficient level of evidence. I, here, worry that James has set the standards for evidentialists very high, since the “must” and “perfectly sure” here imply not only having evidence, but also a very strong kind of overall evidence. This stringent requirement James attributes to the evidentialist as an epistemic absolutism, as we saw from Sections V and VII, but it does not seem intrinsic to the evidentialist view, nor is it actually part of Clifford’s stated views, as we’d noted earlier that he was willing to count very weak sources of epistemic justification as evidential. Regardless, the level of epistemic justification for the belief is not the central question here, though it is relevant to the success of James’s overall argument—if he equates evidentialism with absolutism, then he’s not engaging those he purports to target for criticism.23 Returning to the core of the argument from friendship (and, as I take it, the correlate argument from romantic conquest), James argues that A withholding belief, whatever the evidential situation, renders friendship with B highly improbable, because such states of disbelief yield unfriendly actions. Hence, (3d)—ten to one, the liking will never come. This inference requires two correlate commitments: The Axiom of Aloofness (AA): If A is aloof with B, L is less likely. The Axiom of Suspension (AS): If A does not believe L, A will be aloof with B. Given PN and PA, A’s belief that L should be interpreted as evidence that it is likely that B will like A. The issue has to do with feasibility. So for PN to hold, certain conditions must obtain, and we (and A) must see that they do—B must be capable of interpreting A’s gestures, B must be aware of A’s existence, B must not already be actively ignoring A, B must be within a communicative distance from A, and so on. So if we knew that all the ceteris paribus conditions obtained between A and B, then A’s beliefs about B’s attitudes will be a reason to believe that B will (or perhaps already does) like A. If that is the case, then A’s belief that B likes him is, in fact, evidence for the proposition that B likes A. The total evidence will not just be B’s indications as to how he feels about A, but will also

Reading William James’s “The Will to Believe”

143

include PN, PA, AA, AS, and whether A holds that L. If all that is part of the total evidence, then knowing all that gives A evidence for L. There are a few places to push back on the argument from friendship, as I’ve outlined it here. The first is that the Axiom of Suspension (AS) seems downright false. This has to do with two modalities of “not believing” L. One may suspend on L or one may disbelieve L. Surely disbelieving L, perhaps A believing that B dislikes him, will yield suspension. But how does not believing that L likes him make A aloof? There are many, many people around to whom I can be friendly without my belief that they like me. I may be friendly to them as a matter of habit. I may be friendly to them because I like them, and that’s enough. I may be friendly to them simply to be pleasant. It does not follow from the fact that I do not believe we are (yet) friends that I must fail to act in a friendly way. My suspicion here is that James’s best case for AS is from the principle that all action has a measure of belief as its explanation, which is an instantiation of the conversion fallacy. Again, I will address the fallacy later, but it’s best to highlight how consistently James’s case relies on the conversion of the plausible thought that All belief directs action to the implausible All action is directed by belief. Surely, one can see the case for the AS following from the second of these commitments (so that to act in a friendly manner, one must have a measure of belief in the friendship), but, again, such a demonstration would be fallacious. A further problem with AS, as I take it, is due to a persistent ambiguity of scope with negations and belief operators. Consider A’s three possible attitudes toward L: (i) A can not believe L or (ii) A can believe that L is false in the sense that B does not positively like A, or (iii) A can believe that L is false in the sense that B dislikes A. It seems that it is appropriate for A to be aloof only with those who positively dislike him. So the AS is true only for case (iii), but not, as I see it, for (i) or (ii). (How would A taking B to have no opinion about him cause him to be aloof as opposed to anything else?) But James’s argument takes the suspension of judgment regarding B liking A to be equivalent to holding that B dislikes A. Again, we can concede that if it is true that B dislikes A, then only A’s believing that will attain any good results. If it is true that B has no opinion about A, then it seems only the belief that this is true will get the goods—though perhaps A’s suspension of judgment will be a close second place. As I take it, James’s case here seems to be proceeding from the thought that the question of L is a kind of forced option for A. A can believe L or go without L. The trouble is, as it was made clear earlier, the case for placing the varieties of “doing without” into one composite class requires more than just calling it “doing without.” Otherwise, it’s merely verbal, and not a real difference. There must be

144

Evidentialism and the Will to Believe

a kind of functional identity of the class shared between the members. So, like with the runaway trolley case, suspending belief about whether to jump and deciding not to jump are functionally equivalent, because they yield the same behavior: staying on the trolley. The challenge for the argument from friendship, if it’s to be salvaged by the forcedness model, is to make the case that all cases of “doing without L” are functionally equivalent. The trouble, as I’ve shown above, is that they aren’t, and the only form of “doing without L” that yields what James takes to be the telling functional element of the option, being aloof, is positively believing that the other dislikes you. The problem is that neither suspending judgment on the matter nor just thinking that the other person has no opinion about you yields aloofness. It’s not a forced option, then. The third piece of pushback on the WTB-doctrine arising from the argument from friendship is that if James’s argument is correct, he has not shown that one may believe without evidence. Instead, he’s shown that in cases where one has a belief that if one has the belief then the belief is likely true, then one’s actually got evidence when one has the belief. James’s argument from friendship, if sound, is not a counterexample to the evidentialist doctrine, but rather a case for a unique form of evidence, what I’ll call doxastically dependent evidence. I will have a more complete discussion of this later in subsection 5. The fourth piece of pushback on the argument from friendship is that the antecedent faith can actually impede the actions yielding truth. Consider: if A thinks that B already likes him, then he may not need to initiate all the friendmaking gestures toward B. If A believes that friendship with B is a fait accompli, then why invest the effort to bring it about? That is, it’s when you believe that you’re not yet friends that one is most demonstrative of one’s appreciation for the other person and most attentive to those gestures from others. Once we’ve been clear that the friendship is established and we believe that we are friends, the urgency of that clarity is reduced. We simply are friends. But if one skips to the simply friends stage, then others will not be clear about what’s happened. Some friends, like some romantic interests, require wooing, and the belief that they are sure things not only does not help but positively retards the objective. Finally, in the spirit of “The Will to Believe” as intellectual advice, let me take a stab at a little advice to counter James’s friendship argument. Again, I’ll run a sauce for the goose argument. The argument from friendship is, on the one hand, an argument for belief management based on an analogy with making friends. On the other hand, the argument from friendship itself is a piece of advice about how to make friends, pursue romantic interests, and win boons. I think James’s advice is downright terrible. There is a much better (both effective

Reading William James’s “The Will to Believe”

145

and appropriate) way to win friends. It is, in short: Make yourself worthy of friendship and honor the virtues that others embody and the aspirations they have. Instead of taking others to like me so that they may come to like me, perhaps I should focus on being someone they should like. If they do not like me in the end, well, that’s their fault, and the loss is theirs. Equally for my friendship with others—I want to be friends with those who are worthy of friendship, and as such, I should seek those who have the suitable virtues. In the case of romantic conquest, specifically, James’s advice seems utterly ridiculous. As Brian Zamulinski notes, James’s dating advice is “unconvincing”: The one in which a delusional man harasses a woman until she gives in and presumably marries him is repugnant as well as implausible. (2002: 448)24

James’s advice seems to be that one should be confident. And this seems appropriate. But the plausibility that this confidence is without question the best option or highly appealing does not seem right. Consider the man who thinks he’s “god’s gift to women.” He’s performing the act of assumption, but it yields contempt, not conviviality. Zamulinski, above, sees a darker trend to the advice, and I, too, worry that James’s advice is the tenacious belief of too many women’s stalkers. Sometimes the reason why you don’t have evidence she likes you is because she’s just not that into you, and your belief otherwise isn’t going to change any of that. Sometimes you should just leave her alone. These points apply equally for the “promotions and boons” that James later mentions—instead of focusing one’s mind on the thought that they will come and that one will attain them, one should focus one’s mind on making oneself deserving of them and perhaps on the deliberations as to whether they are worthy of such trouble. Surely that not only makes the chances of having friends and attaining status better, but it also makes it such that one is entitled to those goods in the case that they truly are goods. That is, I not only have desires that I have good things, but I also desire that I deserve them when I get them and that they be truly good things. The latter two are the more important commitments. The will-to-believe is one thing, but the commitment to entitlement (whether it be to goods or to beliefs) is another.

4. The argument from social coordination James’s second line of argument clarifying doxastic dependence and the fact that there are lawful exceptions to the evidentialist’s prohibition against overbelief

146

Evidentialism and the Will to Believe

is that social coordination would be impossible without it. He holds that such a faith is a necessary condition for such organized action. In this case, however, James doesn’t present much of an argument, but three restatements of the thought. The first is: A social organism of any sort, large or small, is what it is because each member proceeds to his own duty with a trust that the other members will simultaneously do theirs. (WB 29)

Here, the thought is that as we aggregate and specialize our work, we rely on each other to fulfill our respective duties. Rise in specialization yields and requires a rise in dependence. James then turns to the attitudes one takes when performing this move: Whenever a desired result is achieved by the co-operation of many independent persons, its existence as a fact is a pure consequence of the precursive faith in one another of those immediately concerned. (WB 29)

So James holds, again, that in relying on or trusting in one another, one must have the positive overbelief that those others in whom one relies are reliable. He then makes it explicit that this antecedent commitment of the individuals is a necessary condition for the functioning of the corporate entity: A government, an army, a commercial system, a ship, a college, an athletic team, all exist on this condition, without which nothing is achieved, but nothing is even attempted. (WB 29)

The condition, again, is the positive “precursive faith” that each member of the team, representative of the government, officer of the army, and agent of the commercial system will do what they are supposed to do. As far as I can see, James offers no argument for the view, but only three iterations of the thought that belief in the reliability of the component members of a corporate body is a necessary condition for relying on the corporate body. So to be a good teammate, one must rely on the team, and so rely on the fact that others are good teammates. The same goes for the other corporate bodies. James makes his commitment here clearer with his example of the train robbers and passengers. He notes that when “a few highway men” come to rob a train, they can do so only because the train passengers do not trust each other to rush the robbers all at once. James then considers the alternative: If we believed that the whole car-full would rise at once with us, we should each severally rush, and train-robbing would never be attempted. (WB 29)

Reading William James’s “The Will to Believe”

147

So James, with the train robbers case, shows that when there is no antecedent belief there is no individual action, and when there is no individual action, there is no collective action. The basic form of James’s case, then, proceeds in parallel fashion to the argument from friendship. Let us break up the thought into its components. First are two related points about positive dependence: Principle of Corporate Action (PCA): A corporate body may function only if its individual members perform their respective duties. Principle of Reliance (PR): Individual members of the corporate body perform their respective duties only if they believe their counterparts in the corporate body are reliably performing theirs. PCA and PR together yield James’s thrice-repeated conditional that corporate function has belief in reliable corporate functions as a perquisite. Now, given that corporate function is preferable to dysfunction, it’s better to believe in it than not. This is the lesson, as I take it, of the train robbers case—it’s not just to show the conditionality of corporate action in belief, but its preferability. James, again, doesn’t present much of an argument here, but I’ve done my best to construct a plausibly attributed one on his behalf. The PCA seems reasonable, but the PR seems just as objectionable as the Principle of Suspension (PS) was for the argument from friendship—one does not need the positive belief that others are competently doing their respective jobs to do one’s own. Rather, that’s what it is to mind one’s business and take care of one’s own work. For sure, James is right that belief that one can’t trust others to do their jobs will yield a collapse in doing one’s own. This is what the train robbers case yields—it’s because we positively expect others to do nothing that makes us do nothing in the face of the highway men. But with performing one’s duties as a teacher at a university or in being an honest broker in an economy or in enacting one’s service as an army, that others are not reliable does not make it a reason to not do one’s job. In fact, one should do one’s job despite the fact that there are many unreliable others in the system. Further, that one has a role to play in these corporate entities should be the only relevantly motivating reason, and in this case, suspension of judgment (or even suspicion) regarding others’ performance should not yield opting out of doing one’s duty, but no change at all in doing the duty. For sure, it is a psychological fact that when we suspect, say, others are not reliably following rules, it increases the likelihood of our not following the rules. A simple example is that if students believe that cheating is rampant and unpunished, they are more likely to cheat in their own cases.25 Again, my suspicion is that James’s best case for the view that positive belief in

148

Evidentialism and the Will to Believe

the reliability of corporate others is a requirement that our actions that rely on and integrate with others reliably doing their duties is that All such action is directed by belief. For James to hold that suspension of judgment is equivalent to disbelief in reliability, he must hold that we cannot act in a way that relies on these others without holding them as reliable. Now, I think the better case might be that to maximize, say, a team’s cohesion, one must instill the belief in each member that the others in the group are reliable. Team spirit itself requires a belief that one’s teammates are capable and regularly successful, and regularly, coaches will ask their players to “buy in” to the team. This positive cognitive view, for sure, is an element to excellent team spirit. So James is right that a condition for excellent teams is that the players must have positive views about each other, but that it is not the case for teams whatsoever. A second critical thing to note about the social coordination argument is that James’s case leaves a good bit of the evidential concerns of the participants on the side. He only mentions that there must be a “precursive faith,” which, as I take it, is a matter of overbelief before cooperative action. But here, I believe, we have a problem. In order to learn one’s duties or responsibilities in a corporate entity, one must already be participating in and gathering evidence about those others who take part in that body. Team sports have regular practices, and in the midst of learning your own plays, you also take note of how those around you are performing. The point is that participants in these social coordinations are not, like the friendship cases, starting from evidential zero. One does not simply will oneself to trust one’s teammates, but that trust is earned. That is why it often takes teams a time to “gel.” The same, as I see it, goes for all the other institutions and corporate bodies. Furthermore, one way to even identify what one’s responsibilities are is to see where the others with whom one must coordinate one’s efforts need help, supplement, and so on. Consequently, I am inclined to agree with James that the endpoint of ideal corporate activity has positive belief in the reliability of others, but that is not a result of mere willing, but the result of evidential assessment.

5. Doxastic efficacy and the Will to Believe James closes Section IX by noting that he has shown that there are “cases where a fact cannot come at all unless a preliminary faith exists in its coming” (WB 29). These are cases of what I’ve been calling doxastic efficacy. James’s presentations have doxastic efficacy coming in two degrees, and it’s worth our while to

Reading William James’s “The Will to Believe”

149

distinguish them. Once we’ve seen the range of doxastic efficacy, we can see it as central to James’s program to establish the WTB. This, however, will yield a problem for James, as once we’ve understood how some beliefs are doxastically efficacious, we have no reason to hold them to be contrary to the evidentialist’s rule. Doxastic efficacy and the fact that James holds that it is central to his case can be captured by James’s own way of summing up his case against the evidentialist at the end of Section IX: [W]here faith in a fact can help create the fact, that would be an insane logic which would say that faith running ahead of scientific evidence is the “lowest kind of immorality” into which a thinking being can fall. (WB 29; emphasis in original)

Again, James invokes the Cliffordian moralist tone to contrast his case, and as he provides his examples, he marks them as examples of faith in a fact helping create the fact. This theme of the efficacy of beliefs in bringing about their truth is a regular one of James’s early work, and it is a standard move in commenting on this element of James’s thought to reference the earlier essay, “The Sentiment of Rationality.” In that essay, too, James invokes Clifford’s evidentialism and promises to make a case against it. James poses a counterexample, the famous Alpine Climber case: Suppose, for example, that I am climbing the Alps, and have the ill-luck to work myself into a position from which the only escape is by a terrible leap. (WB 80)

If the Alpine Climber (AC) believes he can make the jump, he increases his chances of making it, as he will jump with confidence. If he does not believe he can make the jump, “I shall hesitate so long that at last, exhausted and trembling, and launching myself in a moment of despair, I miss my foothold and roll into the abyss” (WB 80). It is easy to see that Alpine Climber case is structurally similar to the friendship, romantic conquest, and social coordination cases in “The Will to Believe.” In essence, the setup is that an antecedent belief that p makes p more likely, and no belief that p makes it so that p is either less likely or not possible. Given that p is preferred to not-p, one should therefore believe that p. With the AC, the principles running the example are roughly: Confident Jump (CJ): AC will make a successful jump only if AC jumps confidently Belief Jump (BJ): AC will jump confidently only if AC believes he will make the jump.

150

Evidentialism and the Will to Believe

And the correlate axiom running the suspension alternative is: Hesitant Jump (HJ): If AC does not believe he will make the jump, then AC will (i) hesitate until exhausted, and then (ii) jump, but despairingly and exhaustedly. I will pause momentarily to note that neither BJ nor HJ seems plausible to me. They inherit the same problems as the AS in the friendship case and the PR in the social coordination cases—namely, that James has presumed that belief is the prerequisite for action. But one can perform an action without the belief that it will be successful, and one can throw oneself into it completely. Surely, if one, in the Alpine Climber case, realized that hesitating until exhaustion yielded death, one would summon a nondespairing or exhausted jump, and one could do that without the belief that it would be successful. (One, surely, would have the belief that it would be more likely successful than the despairing or exhausted jump, but this is not the target belief James has in mind.) Again, I take it that the best line James can take here is to be committed to what I’ve been calling the conversion fallacy—that because All belief guides action, All action is guided by belief. Regardless, it is also worthwhile to note, again, the similarities between this case and the paradigm example of doxastic dependence for “The Will to Believe,” the friendship case. In the friendship case, A is interested in whether B likes him. In both cases—B’s liking A and AC making the jump—some outcome is desirable. So having some belief not only makes the belief ’s truth more likely, but whether or not that belief should be true is itself determined by our desires. That is, James presumably takes it as unreasonable to be committed to selffulfilling prophecies of defeatism in both cases—where the belief that one will fail ensures the failure. The difference is that positive thinking yields results that are desirable and negative thinking yields results that are not desirable. The desires, then, determine what beliefs are acceptable in terms of the likely consequences of holding them. It’s important to make these two elements of the view fully clear. The two parts of the doxastic efficacy view favoring the WTB are: first, that a subject having the belief contributes to its truth, and, second, that the truth of that belief is preferable to its falsity. Remember AC. If AC holds he can’t make the jump, then on James’s reasoning, he’ll end up with a true belief. But it’s not something he would want to be true. The same goes for the friendship case. A is asking the question whether the proposition L (B likes A) is true. If A believes that L is false, then A will act in a way that makes L false. But that’s not something that A wants to be true.

Reading William James’s “The Will to Believe”

151

So our will, then, determines not only our belief policies, that of the Truth Norm or the Error Norm, but even once we’ve adopted the Truth Norm, what truths we should believe. Since we prefer to survive our trips in the Alps and to make friends, and our beliefs in these matters are efficacious, we will believe truths we prefer by our beliefs making them true. There are, then, two orders that will determine belief for James. There is the meta-epistemic order, wherein the will determines which belief policy one will follow. Then there is the first-order level wherein the will determines which truths are preferable, and thereby directs belief in doxastically efficacious cases. James captures this function of preference to close the Alpine Climber case: There are then cases where faith creates its own verification. Believe, and you shall be right, for you shall save yourself; doubt, and you shall again be right, for you shall perish. The only difference is that to believe is greatly to your advantage. (WB 80)

And so the preferability of friendship, social coordination, and making the Alpine jump, being matters of our inclinations, determine how we should believe in cases where our beliefs can influence the outcomes. Returning to the strict issue of doxastic efficacy, given James’s examples, efficacy of beliefs comes in two strengths. On the one hand, there are preferable facts that have doxastic preconditions. When James says that for social coordination cases, “without (belief) not only is nothing achieved, but nothing is even attempted,” he is holding that a necessary condition for the coordinating actions is that the component agents of the corporate body have the positive belief in each other’s reliability. On the other hand, when James says that the friendship cases, without the belief, “ten to one the liking never comes.” These are facts that have doxastic contributions, having the antecedent belief makes its truth more likely than if one did not have the belief. Note that the friendship cases are ones that must be contributions cases, as if James had it as the stronger preconditions requirement, that would make it so that B could like A only if A antecedently thought B actually did like A. That could never work. However, we can see why he has the much stronger preconditions requirement for social coordination cases, as distrust yields discord and cooperative failure. Given the way James presents the case with the Alpine Climber, I am inclined to interpret it as one of preconditions, as James even goes so far as to contend: In this case (and it is one of an immense class) the part of wisdom is clearly to believe what one desires; for the belief is one of the indispensable preliminary conditions of the realization of its object. (WB 80)

152

Evidentialism and the Will to Believe

A “preliminary condition of the realization” is, as far as I can see, a necessary condition. Now, the point of this distinction is to recognize that some cases have it so that the only way the target truth can come to be is with the belief, and there are other cases where the target truth can come to be, but is just made more likely with the belief. This is a useful distinction, at least as far as we are clear as to just how doxastically dependent the facts in question are. Moreover, it will determine whether there are alternate methods available for subjects to determine the truths in question. In the case of social coordination and Alpine jumps, it seems the only way for the subjects to come to make the determinations are to come to have the beliefs, given that the beliefs are prerequisites for the facts. However, for the friendship cases, the subject can come to find out that the other person likes him independently of his acts of friendship (perhaps the other person may initiate friendly behavior). Given this distinction, it will be important to classify James’s religious faith in terms of how strongly he holds it to be doxastically efficacious. James presents doxastically efficacious beliefs as counterexamples to the evidentialist constraints on belief. The question is whether these really are counterexamples. Let us restrict our discussion here to the friendship case and the Alpine climber case. Now, let’s concede that both cases have no antecedent evidence regarding the matters up for consideration. So with the friendship case, the question of whether B likes A is intellectually undecidable for A—A may have equipollent evidence or no evidence, or some but insufficient evidence. Note that the matter is not, as James had noted in Section IV, “by its nature” undecidable, but it is only undecidable at the time. The same, let us concede, can be for the Alpine Climber. AC does not have evidence that he can make the jump, but, it seems, must either jump or not. However, both cases are constrained by evidential considerations. With AC, surely the deliberations are superfluous if the jump is physically impossible. Invoking the power of positive thinking is ridiculous with a 100-foot jump. It works only with cases we can antecedently assess as feasible. So our evidence about whether AC can make the jump would be first how far the jump is, what kinds of jumps AC has made in the past, AC’s experience, and so on. Only when the distance to jump is on the high end of AC’s jumping track record is AC’s attitude about the jump relevant. Only when it’s a hard jump does AC’s confidence matter. And this is where doxastic efficacy is important. If we know AC believes he can make the jump and if we (with James) take AC’s confidence to increase the likelihood of a successful jump, then do we not have evidence about whether AC can make the jump? If we can have that evidence, then it

Reading William James’s “The Will to Believe”

153

seems that AC is in the best position of all of us to know whether he believes he can make the jump. If that is the case, then the Alpine Climber case is one where the jumper (AC) also has evidence. James’s argument from friendship is supposed to be a counterexample to the evidentialist requirement that one believe in accord with one’s evidence in that the case is one both lawful but not allowable by the evidentialist requirement. But, given doxastic efficacy and James’s connecting principles, if A holds that L is true and holding that L is true makes L more likely true than false, then A has evidence for L. The case I have made here is that what makes the cases lawful is that they are in accord with the evidentialist requirement. We properly employ the power of positive thinking only in the right circumstances, and they are delimited by evidence. AC is right to use the will-to-believe only in cases where the jump is evidently feasible, and A is right to use it when B is evidently the right kind of recipient for nice treatment. Agents must judge those cases appropriately, so A and AC do have evidence for the truth of their beliefs (the hypothesis is that they are lawful). It certainly sounds strange to say that a subject’s belief that p is evidence that p is true, but given the connections that run James’s examples, these subjects have epistemically justifying reasons for holding their beliefs, and as such, have evidence for them. The consequence is that James’s argument from friendship and other doxastically efficacious cases, far from providing counterexamples to evidentialism, in fact sharpen our notion of evidence. That is, if there are doxastically dependent facts, then when a subject has a doxastically efficacious commitment, we have reason to hold that the fact will come to be. That’s evidence.26 Recall that James’s promise for the essay was to make a case for lawful religious overbelief. That’s what a justification of faith is supposed to be. So far, he’s been looking at lawful overbelief, and we’ll turn to religious overbelief in Section X. The point is that what needs to be shown is that the cases are recognizably overbeliefs, and nevertheless recognizably lawful. As I see it, James’s case for the beliefs here has been that they are lawful, because they are doxastically efficacious cases—the interested subject makes a determination as to what to believe, and once the subject has made the determination, the belief is made more likely true by that determination. The point here is that if the subject in question has taken James’s advice and is now aware of the connection between these beliefs and their truth, then the subject now has evidence about the contents believed. So James has made these overbeliefs lawful, but by making

154

Evidentialism and the Will to Believe

them no longer recognizably overbeliefs. That’s not what was promised. Now, I think the better spin to put on the matter is to hold that James’s observations are that because of doxastic efficacy, there is a wider range of lawful beliefs, ones for which we can have evidence. And religious faith, as it turns out, is one such form of belief. But in doing so, he needn’t hold that the skeptical situation with regard to these beliefs must obtain by their nature. Overall, we can see the case from Section IX to be for clarification of the conditions for the WTB. We’d seen in Section IV that the conditions for passion to decide an option on the WTB are genuineness and intellectual undecidability. The work we see from Section IX and the wide cases of moral skepticism, and applied cases of friendship, romantic conquest, and social coordination is to make the case for moral options wherein there are two further conditions: preferable options and doxastic efficacy. This, as I take it, yields an updated and composite doctrine for the WTB, combining the stated view from Section IV with the conditions of efficacy outlined in Section IX. WTB*: If S has option O:(p, q) and O:(p, q) is (a) genuine, (b) rationally undecidable, (c) has a preferable option of the two, and (d) the preferable option is doxastically dependent, then S may and must decide the option on the basis of S’s passional nature.

The addition of conditions (c) and (d) to the WTB doctrine from Section IV is that James has conceded in Section VIII that “scientific” questions are not decided on the basis of our passional nature and this is not merely because they fail the genuineness condition, but because they are not about what would be better to be and what we can do about it. In short, the composite form of the doctrine of the will-to-believe, WTB*, restricts the will to believe to moral questions wherein we are makers, not mere recorders of truth. The question, now, is whether this composite version of the will-to-believe applies to religious belief.

Section X—Logical conditions of religious belief 1. The overall form of James’s argument James’s justification of faith depends on showing that the assessments of those “well imbued with the logical spirit” that faith is philosophically unlawful are incorrect. Again, the justification will not be that the contents held by the faithful meet the requirements of lawfulness deployed by their evidentialist

Reading William James’s “The Will to Believe”

155

opponents, but rather that they are lawful exceptions to those requirements. As such, James’s argument takes the form of a series of acceptable exceptions to the evidentialist’s criterion for lawful belief. These exceptions are arranged to ascend from seemingly insignificant occasions of social confidence to cases of religious belief. This interpretive point is worth highlighting. James’s case for faith is built as one that begins with small exceptions to a rule and builds to bigger and more significant exceptions (faith being at the height of the ascent). From the initial exceptions, those of friendship, he comes to more substantive cases of romantic conquest and social coordination. What James takes as making all these exceptions acceptable is that they share the feature that in believing that one will be successful in some endeavor (making friends, having some coordinated activity come to fruition, pursuing a love interest), the better one’s chances are in being successful. Let us call the overall thrust of these observations James’s ascending case—that is, the argument for faith as a legitimate exception to a rule for belief is made on the basis of it being an extension of a pattern of other less substantive but similar cases of legitimate exceptions to the evidentialist’s rule. The analysis of the ascending case must be broken into two parts. First, we must articulate the rule being criticized, and second, we must identify the exceptions that ascend in the direction of faith. The rule for lawful belief James associates with the logical spirit is the evidentialist norm that one should proportion one’s belief to one’s evidence. W.K. Clifford’s articulation is James’s target for critique, what I called earlier Clifford’s Evidentialist Norm (CEN): [I]t is wrong, always, everywhere, and for anyone, to believe anything upon insufficient evidence. (L&E 175)

James quotes a good deal of Clifford, again, so as to capture the “robustious pathos in the voice,” and so highlight Clifford’s own sermonizing tone when he says that “belief is desecrated when given to unproved statements,” that such beliefs are “sinful because (they) are stolen in defiance of our duty to mankind.” Importantly, James acknowledges that CEN is “healthy” (WB 18) and a worthy “ideal” (WB 27), but he holds that it is too demanding, as there are acceptable outliers. It is important to note that James here is endorsing the rule generally in calling it a “healthy” attitude, so the case he will make against it is not to overturn it, but to show that there are legitimate exceptions. James’s argument against the universal application of CEN is, again, that of an ascending case, as the initial counterexamples seem inconsequential, but

156

Evidentialism and the Will to Believe

slowly build to those of religious belief. The exceptions for the case to ascend in significance successfully must share a handful of features. The cases of belief James takes to be counterexamples to evidentialism are marked, first, by the fact that they are all what he calls “genuine options.” A genuine option is one that is living, forced, and momentous. Living options are ones that have “some appeal, however small, to your belief.” The living options are ones that would occur to you when surveying your choices; ones that would never occur to you are the ones that are not living. Forced options are ones wherein the usual distinction between acceptance, rejection, and abstention is erased, so that abstaining from judging or choosing is equivalent to either accepting or rejecting an option. Finally, momentous options must be unique, significant, and irreversible, and so must pose important once-in-a-lifetime opportunities. This brings us, then, to what James identifies in Sections VIII and IX as moral questions, that is, decisions wherein it certainly is best to pursue one option over another. Moral questions are those not of what exists (which is what James calls a scientific question). Rather, moral questions are of what is good and what would be better to exist. Further, with a subset of moral questions, how one believes with regard to the question actually is a determining factor in whether the fact in question comes to exist. That is, with some issues, which option you believe is true is the one more likely to be true. Such cases are doxastically efficacious. James surveys a variety of such specific moral questions: ●

●

●

Friendship cases: Whether or not you like me is a moral question, and it is a doxastically efficacious matter. If I believe you don’t, or I am not inclined to believe you do, I will “stand aloof ” and “ten to one, your liking will never come.” Whereas, my antecedent faith in your liking me “is in such cases what makes your liking come.” Social coordination cases: How we do our jobs, whether in an army, on an athletic team, or in a commercial society, is a moral question and is doxastically efficacious. Our performances in our roles depend not only on the cooperation amongst many individuals in doing their own jobs, but also on the shared antecedent trust that we all will continue to do so. Trust keeps it all running. Romantic conquest cases: Whether to pursue a romantic interest is a momentous issue, one that, often, does not yield second chances. The same for marriage proposals. It is infrequent that one knows for sure what the answer will be, but the good things come only to those who make the uncertain proposals. James notes, “How many women’s hearts are

Reading William James’s “The Will to Believe”

157

vanquished by the mere sanguine insistence of some man that they must love him!” And so, James substantiates these moral cases with the thought that, like with the Truth Norm, the goods to be had in being right are worth the risks of being wrong. In fact, in having the faith of friendship and coordination, we actually reduce the risks of being wrong. To substantiate this point, James gives an example of a train of passengers, who, without social trust in each other can be robbed by a few highwaymen, but with the social trust, they “would rise at once … and train-robbing would never be attempted” (WB 29). Notice that each of the cases that James surveys has the pattern of what I’ve taken to calling doxastic efficacy, wherein believing that p makes p more likely and not believing that p (or believing that not-p) makes p less likely (and not-p more). James captures the point by noting that the “desire for a special sort of truth brings about that special truth’s existence” (WB 29). So the moral cases James takes to be worth posing are those with preferred outcomes that are doxastically efficacious. Given that these outcomes are good and that the antecedent beliefs contribute to their coming to be, James makes it clear what the stakes are: There are, then, cases where a fact cannot come at all unless a preliminary faith exists in its coming. And where faith in a fact can help create a fact, that would be an insane logic would say that faith running ahead of scientific evidence is the “lowest kind of immorality” into which a thinking being can fall. (WB 29)

With doxastically efficacious propositions, evidentialism is inappropriate. They are lawful, and to prohibit them would be an “insane logic.” With these cases, then, James has made a case for what I call the composite version of the willto-believe, the WTB*. Notice, importantly, that the reasons offered here only support the composite version, and not the bare version WTB from Section IV. James’s essay depends on the cases being both composed of at least one morally preferable option and that option being doxastically dependent. It is worth pausing, again, at this point for a few observations. First, James’s case for faith is an ascending case, which is that there are exceptions to the evidentialist rule, that those exceptions form a kind of family, and that cases of religious faith belong to that family. So far, we have identified the exceptions as friendship, social coordination, and romantic conquest cases. They form a family on the basis of their being genuine, morally momentous, and doxastically efficacious examples of beliefs. They are lawful beliefs because they, in their

158

Evidentialism and the Will to Believe

circumstances, cannot be withheld and they are effective in bringing about their truths, which are preferred. Second, for an ascending argument to be effective, the cases the argument ascends to must have the properties of the case family it ascends from. And so, the ascending case is, at bottom, an argument by analogy, in that it runs: P1. Prima facie, Clifford’s Evidentialist Norm (CEN) is acceptable. P2. Religious faith seems to be ruled out by CEN. P3. Friendship, social coordination, and romance cases are lawful exceptions to CEN. P4. Friendship, coordination, and romantic conquest chases are momentous, moral, and doxastically efficacious cases, and those properties explain their lawfulness. (An application of WTB*) P5. Religious faith is momentous, moral, and doxastically efficacious. C: Therefore, religious faith (despite appearances in P2) is an acceptable exception to CEN. Thus far, James’s case has made its way through P4, and the question is whether P5 is true for religious faith. James must turn to making the case that, first, religious faith is such a moral and doxastically efficacious matter, and so, second, is lawful. However, before we can turn to the case for religious faith, let us briefly return to James’s argument for lawfulness in the friendship and coordination cases. Is P4 true? Are the moral cases James surveys exceptions to the evidentialist rule? The question is whether James’s case from doxastic efficacy is sufficient to be a counterexample to Cliffordian evidentialism. Consider the simple fact that positive thinking is conducive of better performances—confident players tend to make their shots, those who have an antecedent sense of belonging tend to do better socially. And it is also true that there are self-fulfilling prophecies of doom—those who believe they won’t make the putt or that others don’t like them will generally fail at those endeavors of putting and friend making. First, note, however, that the efficacy of positive thinking is itself evidentially bounded. Recall the example James’s earlier essay, “The Sentiment of Rationality”, the Alpine Climber case. The climber must make a jump across a crevasse. The confidence that he can make the jump makes his successful jump more likely, pessimism less. The climber, James holds, not only lawfully may but also must have the confident belief. But remember that this is appropriate only when the jump is feasible. If it is a 10-foot jump, this seems right, but not for a 100-foot or 1,000-foot jump. Appropriately using the power of positive thinking itself must

Reading William James’s “The Will to Believe”

159

be evidentially bounded—you must have evidence that the case is a feasible one in the first place. Second, note that if a subject has identified a feasible case and has determined that she has confidence, then she has a reason to believe the outcome is likely. She has evidence. Doxastic efficacy is an evidential principle, not a principle of nonevidential reasoning. Consequently, James’s case against Cliffordian evidentialism, even were its core components apt, is less a refutation of, counterexample to, or lawful exception of evidentialist principles, but rather a case that there are surprising ways in which evidence can be had. Under this description, James’s argument is actually in the service of an expanded notion of evidence than in the service of nonevidential reasons for faith. In turn, James may satisfy the Lawfulness Requirement I’d noted in the Preamble by holding that there are cases of antecedent belief that are lawful, but he has not shown that they are cases of lawful belief without evidence. As a consequence, I think James fails to meet the Lawfulness Requirement as he’s presented it. Whether or not this is a devastating criticism is a matter of no small significance, as James’s case is that faith not have the evidential backing; and yet doxastically efficacious examples have them. Of course, the case may be changed so as to say that faith is a specie of evidence-conducive belief, a mode of inquiry, but this is no longer the view James identifies as what is to be defended. For this point to be fully appreciated, a distinction is in order. On the one hand, there may be a specie of doxastically efficacious belief that is conducive of evidence for its truth, and there may be a specie of doxastically efficacious belief that is conducive of its truth, simpliciter. That is, some beliefs may be so that if you have them, you put yourself in a better position to see that they are true. This notion is often associated with Augustinian and Anselmian cases for faith, that one must believe in order to understand.27 On the other hand, there are beliefs that when you have them, you make them more likely true. The former have efficacy in making the evidence for a standing truth available, the latter bring the truth to standing. Call the former evidentially productive doxastic efficacy, call the latter alethically productive doxastic efficacy. It is important to note that James’s ascending case is one driven by the alethic productivity of doxastically efficacious beliefs, not their evidentially productive elements. James, in making the case, notes that “[t]he desire for a certain kind of truth brings about that special truth’s existence” (WB 28, emphasis mine). He holds that he has identified cases where “faith in a fact can create a fact” (WB 29, emphasis mine). And as he turns to the issue of religious faith, he re-articulates the observation of alethically productive cases of doxastic efficacy:

160

Evidentialism and the Will to Believe

In truths dependent on our personal action, faith based on desire is certainly lawful and possibly an indispensible thing. (WB 29, emphasis mine)

For sure, if doxastically efficacious commitments are truth-productive, then they will likely be evidentially productive, too. And so, for example, with a friendship case, if Bill believes that Sam likes him, he will make it more likely that Sam likes him. And thereby, he makes it more likely that he’ll have evidence that Sam likes him, as he’s made it true that Sam likes him and if Sam likes him, Sam will act accordingly toward Bill. This is how James takes it that such a belief can “create its own verification” (WB 29). Consequently, James’s case is not the mere modest Augustinian-Anselmian view that belief is a catalyst for accessing evidence of a standing fact, but a case that commitments in question are appropriate because holding them contributes to their truth.

2. Religion’s dual essence As James turns to religious faith in the closing section of “The Will to Believe,” he acknowledges that there is a question looming about whether religious faith is of the same class as friendship and coordination cases. But now, it will be said, that these are all childish human cases, and have nothing to do with great cosmical matters, like the question of religious faith. (WB 29)

On this challenge, James demurs. Religious commitments take many forms and have a variety of objects. They “differ so much in their accidents,” as he puts it. Consequently, for a discussion of religious commitment to apply to all these cases, we must find what binds them all together as religious commitments. Instead of looking at their accidents, we must look to their essence. Thus, James identifies the core of religion as “essentially two things”: R1: The best things are the more eternal things, and R2: We are better off believing R1 (from WB 29–30) The question here is whether the faith that James has saved as a doxastically efficacious moral question is recognizable as the faith of those in the Brown and Yale Philosophy Clubs. Would the beliefs salvaged by this strategy be within the realm of the faithful discussion of divine omnipotence or the difference between justification and sanctification? James’s answer is yes, because he, again, holds that R1 and R2 are the essence of the religious hypothesis.

Reading William James’s “The Will to Believe”

161

But are they? It certainly seems right that R1 and R2 are parts of the standard religious commitment, but are they essential to it? James notes that R1 amounts to the claim that “Perfection is eternal,” and that “the things in the universe that throw the last stone … (or) say the final word” are the best things (WB 29). He concedes, given the modality of the time regarding the claim that it cannot be scientifically verified, and so it is beyond our evidence to hold it. But R2 runs that we should believe it, and so there is the essence of the faith. As I take it, then, the view is that the good is what wins out in the long run. There are a number of critical things to be said of James’s claim that R1 and R2 are the essence of religion. The first is that it simply doesn’t sound much like any religion I’ve heard of or would dream up. And I’m not alone in this. Santayana said of the philosophy of religion on offer that it has “no sense of severity, no joy” (1958: 47). Arnold Johansen notes that James’s two elements together as a worldview “does not strike one as being particularly religious” (1975: 121). Richard Rorty says the commitments strike him as “wimpy” (1997: 11). David Hollinger calls the strategy “philosophically obscurantist” (1997: 77). Allen Wood says they are “empty and insipid” as accounts of religion (2008: 21). Michael Slater notes that James’s two elements are “awfully thin” (2009: 28). We might concede that religion may have such values at its core, but the religion’s accounts are more deeply about where those values come from, how they are maintained, and so on. The hopes that religion is supposed to provide succor to are not the essence of religion—the desire that what’s valuable will last. No, religion is what comes in to tell us that it will and provides an account as to how. It’s not as though it’s an accident that religions have a tendency to think their respective gods made the world, made up the rules, are very powerful and live for long, long times. It’s because they are the grounds for and enforcers of those values. But James’s account makes no reference to them or to any of the myths keeping religion going. This is not a good start. One response to this concern is that James’s strategy was to identify what David Hollinger calls “basic theism,” which affirms the essentials of Christianity but do not become entangled with the detailed overlays of the various views (Hollinger 1997: 74). The trouble is that taking James’s project in this way fails twofold. First, it is too strong. If it’s out to capture the core Christian commitments, then it’s not going to be capturing the elements of other religions. Or if it does, it will do so on the basis of their being alike to Christianity, not on their religiosity. Second, neither R1 nor R2 seem particularly Christian or theistic at all. So it’s too weak.

162

Evidentialism and the Will to Believe

Consider the following argument by reductio that R1 and R2 cannot be the essence of religion. Let person S believe R1 and R2, and this, by the hypothesis that they are doxastically efficacious commitments, is conducive of their truth. Now they are true. They, then, were not true before S believed them. So they are not eternal truths. Consequently, they must be neither the most eternal things nor the best things. S’s religious belief, then, cannot be the “deepest service,” because it is not among the best things. Or consider another way into this thought: if the essential truths of religion are doxastically efficacious, they are not eternal. By their own content, they are not best. But religion should be about what is eternal and best. So religion and its truths, by its own lights in R1 and R2, must not be doxastically dependent. The reductio argument may not be convincing, so another challenge to James’s view that R1 and R2 are the essence of religion is necessary. Consider that R1 is the view that the more eternal things are better. This view has two problems. First, it is not clear how one thing can be more eternal than another. Eternality does not seem to be a term of comparison—if x is eternal and y is eternal, then they must both be equally eternal. The only way for one thing to be more eternal than another is for one thing to be eternal and the other not. So R1, if it is to make any sense at all, must be not that the more eternal things are better, but simply that eternal things are better than noneternal things. And, perhaps, by extension, the things that last longer are better than the things that don’t. Second, R1, so revised, makes eternality a marker (and perhaps maker) of the betterness of things, comparatively. But is this acceptable? No. Consider Hell. It is, by hypothesis, eternal, and one’s being there would be so, too. Now consider an ice cream cone and your eating it. Those things are transitory, noneternal. By R1, Hell and being there would be better than ice cream cones and their eating. That’s not right. Eternality can’t be a marker or maker of good things, since the eternality of a bad thing clearly makes it worse, not better. And R2, then, must be false, too, since given R1, R2 would have us choosing an eternity in Hell over eating an ice cream cone. It is clear that it would be worse to believe R1, not better. The reason we get this problem is that R1, as stated, is ambiguous. That is, when James says the best things are the more eternal things, he may be expressing either a biconditional: R1a: A thing is best iff it is eternal, or he is expressing a simple conditional: R1b: If a thing is best, it is eternal.

Reading William James’s “The Will to Believe”

163

Of course, the Hell case above is a counterexample to R1a, and so let us take that as a reason not to interpret James as making the stronger claim. R1b amounts to the claim that All things that are best are eternal things, and so it is not open to the Hell case’s counter example. We might put the point more effectively if we take it that for James, eternality is not a marker or maker of goodness, but one that explains comparative goodness. This is to say that the case for R1b should be that it is not a view that demarcates what is good from what is not, but one that explains how some goods are called best—namely, that they must be eternal if they are to be best. And so R1b could be rendered: If there is a good thing, it would be better if it were eternal. This is true, presumably on the basis of the thought that good things are better when they last longer, so eternal things (that are good in the first place) will be best. And so the Jamesian story is not about the nature of goods, but that stability is one thing that improves their value. Religion, given this insight, is what provides that stability (or at least a way of conceiving the values as stable)—it is a way of viewing the things that are valuable from a perspective that preserves them. Notice that with James’s identification of the essence of religion with this view about value, there is, again, no mention of God, the divine, or the role of either in the world. Only “the more eternal.” James’s point with R1 is, I believe, diagnostic, in that he is identifying the essence of religious impulses as a move from recognizing that goods that last are preferable to fleeting goods to an ontology that offers the promise of that permanence. The core of religion is the articulation of how what one values most can last. The trouble with this diagnostic take on religion, certainly from the perspective of believers, is that it identifies religion as a coping mechanism in the face of contingency and uncertainty. The objects of religious attention and reverence play merely instrumental roles. Consequently, the identifying components of religions, and specifically the doctrines of the nature of the divine and the proper ways for humans to relate to it, are no longer essential. These matters are, then, not saved by the doctrine of the will-to-believe. This, I think, would be a surprise for those who’d thought when told that a defense of faith was on offer, these things were not to be salvaged.

3. Religion as live and momentous For James’s ascending case to work with what I’ve been calling the composite doctrine of the will-to-believe (the WTB*), he must establish three things: first,

164

Evidentialism and the Will to Believe

that religious belief is a genuine option; second, it is the morally preferable hypothesis of the option; and third, that the religious hypothesis is doxastically efficacious. These three elements will be analyzed independently. I will address the matter of genuineness broken down into its component elements of liveness, forcedness, and momentousness. The moral element of the question will be answered by the matter of liveness, and then I will turn to doxastic dependence and its varieties. At no point in Section X of “The Will to Believe” does James claim that the question of religious belief is a genuine option. However, he does argue that it satisfies the three conditions for genuine options: liveness, momentousness, and forcedness. In this section, I will assess James’s case for momentousness and liveness. The question of forcedness deserves its own discussion, as it occasions the most significant critical response. James begins his case that religion, so conceived (as R1 and R2), is a lawful exception to the evidentialist’s rule by returning to the “logical elements of the situation,” the components of the genuine option. James’s case that the religious question is momentous is surprisingly short. He claims: [R]eligion offers itself as a momentous option we are supposed to gain, even now, by our belief, and lose by non-belief, a certain vital good. (WB39)

That’s all James has to say for momentousness, and this is disappointing. Recall that momentousness, as defined in Section I, requires: uniqueness, significance, and irreversibility (WB 15). One can concede, even without knowing what precisely the vital good of religion is (which will be made clearer with the issue of liveness), that determinations of religious belief are significant. That fact that religious faith occasions so much heat in discussion is evidence enough for that. The question, then, is whether the matter is unique and irreversible. James is clear elsewhere that he holds religious belief to be unique in its ability to center a life and prepare it for vigorous pursuit of the good. Earlier, in “The Sentiment of Rationality,” he argues that there is one element of active natures of humans that Christian religion has recognized as central to human flourishing: I mean the element of faith. Faith means belief in something concerning which doubt is still theoretically possible; and as the test of belief is the willingness to act, one may say that faith is readiness to act in a cause the prosperous issue of which is not certified to us in advance. (WB 76)

And later, in “The Moral Philosopher and the Moral Life,” James holds that: Every sort of energy and endurance, of courage and capacity for handling life’s evils, is set free in those who have religious faith. (WB 161)

Reading William James’s “The Will to Believe”

165

James holds consistently that religious faith offers a unique form of what Michael Slater has noted is “moral strenuousness” (2009: 7). With faith, one sustains a high level of confidence and commitment to moral ideals, one can be indifferent to present ills. We are capable of the greatest achievements when we have faith. Slater emphasizes: James saw religion as practically necessary to supplement morality…. [M]orality or “the moral life” is not and cannot be complete without some form of religious faith. (2009: 13)

This seems correct as a point of what James’s ultimate view is. But it seems disastrously wrong about the matter. Consider the following thought: if the atheist holds that the good things are not eternal things, but transitory and fragile, then it is this attitude that will provide powerful motivation to preserve those good things and hold their destruction off for another day. The belief that the good will out, that the right will throw the last stone—that belief isn’t motivation, but comfort. Belief that the good hangs by a thread and light can be snuffed out—that belief provides strenuousness. To the contrary to James’s view here, theism can motivate grand gestures, but it is not because of the core elements of R1 and R2 that James identifies. Rather the motivation provided by religion are those of propitiation of a cosmic despot or the desperate acts of worship for a beloved but aloof entity. That’s what gets churches built, chapel ceilings painted so beautifully. Not the belief that the good things “throw the last stone.” As a consequence, James is right that the religious question is unique, but he is so for entirely wrong reasons. We must, again, reopen the question of whether religious belief is irreversible. Recall that this was a requirement of momentousness from Section I-5, but James makes no mention of it here. For good reason. Because religious belief is reversible. Conversions happen, people lose their faith, people change their faiths and degrees of it. Now, for sure, when one becomes, say, a Catholic, one does not do so with the plan that it be reversible. But that is a part of the commitment itself, not a truth about the commitment. Consider: when one rationalizes, one believes for what one convinces oneself are good reasons, but are not. That is a truth about the commitment that isn’t part of the commitment. Irreversibility is a truth of the commitment, not one comprising it. One thing to concede is that the religious hypothesis is irreversible, so we think, after death. This is for two reasons, I suspect. The first is trivial—after you die, you can’t change your mind about anything. You’re dead. The second, I suspect is James’s real reason, and it was Pascal’s, too—it’s however your belief is set upon the moment of your death that determines your fate afterward. If it is set

166

Evidentialism and the Will to Believe

on belief in God, you are saved. If not, you are damned. And both are eternal— there’s no going back on that. Note that if the Jamesian line for irreversibility hangs on this commitment, then the case for momentousness in other areas comprising significance, specifically in the living elements of religious belief, is sacrificed. James gets significance in life, but irreversibility in death. I presume that if one faces an option, it is at a time, but if the times at which the option is significant and irreversible must be different, then there is no time when it can be both. And if there is no time when it can be both, then there is no time when the option can be momentous. Recall that I’d argued earlier that momentousness was likely, except for irreversibility, reducible to liveness. So let’s consider James’s case for the liveness of the religious hypothesis. Here, James’s case is much more developed than it was for momentousness. The liveness of the religious hypothesis is something James takes to be on analogue with the other doxastically efficacious cases, friendship, and social coordination. That is, what makes the religious hypothesis most live for us is the fact that it takes the form of theism, that what maintains the good and ensures its eternality is something personal with which we can have a substantive connection and personal relationship: The more perfect and more eternal aspect of the universe is represented in our religions as having personal form. The universe is no longer a mere It to us, but a Thou, if we are religious; and any relation that may be possible from person to person might be possible here. (WB 31; emphasis in original)

From this analogy, James makes the case for doxastic dependence on analogy with the argument from friendship, and I will address that argument later. For now, the matter is identifying the live element in the commitment. My hypothesis is that James is fleshing out his dual core of religion with a third essential element, one that is a feature of the beliefs of those who are actually believers. I believe, following James Wernham (1987: 48), it is something of the form: R3: The entity ensuring the eternality of goods in R1 is Personal, a Thou. And so R2, likely, should be revised to: R2*: It is better to believe R1 and R3. In holding R3, we can see how R2 (and R2*) is true, as we in holding R1 and R3 come to have a relationship with the personal ground for goodness and stability. That guarantee, as an It in R1, becomes a Thou in R3. We, as James puts it, “make god’s acquaintance” (WB 31).

Reading William James’s “The Will to Believe”

167

Now, as I’d noted earlier, it’s not the eternality elements of the religious hypothesis that provide the maintaining power for believers. Holding that the good is eternal is like knowing you’ll win no matter how hard you play. Instead, I’d argued that it’s when you not only might but likely will lose if you play badly that will encourage good play. But with the personal god in the mix, Jamesian motivation returns—we want to participate in communion with this Thou who provides the good with assurance.28 Liveness is measured by willingness to act, and this personal element to religion is what drives the religions of love and passion. And what must be behind the strenuous life, too.

4. Religion as forced For James’s case to work, religious options must be genuine. We’ve seen that, so far, religion can be considered live insofar as we allow James to supplement his initial dual essence of religion composed of R1 and R2 with R3, the commitment to the Thou-personality thesis. Religion is not momentous as James initially defines it, but I take it as no matter, as all the important elements of momentousness can be rendered as liveness. Now we have the question: is the religious option forced? James’s argument, like that from Pascal, is to focus on the costs and benefits associated with belief and the wider class of unbelief. If one believes, one achieves a great good, and if one goes without belief (either with abstention or rejection) one loses out on that great good. In Pascal’s case, the great good is that of Heaven, and the loss of that great good, in both forms of unbelief, is Hell. In James’s case, the great good is that of “being on the winning side,” if the religious hypothesis is true, and the loss is that of having missed out on that chance. James frames the issue, to open the discussion, as entirely a matter of the Truth and Error Norms: [R]eligion is a forced option so far as that [vital] good goes. We cannot escape the issue by remaining sceptical and waiting for more light, because, although we do avoid error in that way if religion be untrue, we lose the good, if it is true, just as certainly as if we positively chose to disbelieve. (WB 30)

To clarify the matter, James analogizes the circumstance to that of the romantic conquest case from before. We might imagine a man questioning whether the woman to whom he wishes to propose marriage is an angel. If, in doubt, he hesitates to ask her until he has certain evidence of her angelhood, he will never ask her. And consequently, James holds, he will lose out on the truth and the goods that come with it:

168

Evidentialism and the Will to Believe

Would he not cut himself off from that particular angel-possibility as decisively as if he married someone else? (WB 30)

Notice first that James’s argument must proceed on analogy with the social cases from before, and further note that James’s view is that in order to participate in the social exchanges, whether they be friend-making, courtship, or social coordination, one must have a positive belief in the prospects of success of those exchanges. One acts, and to act, one must believe. Again, this is another instantiation of the pragmatist’s fallacy of conversion, and I will have a full consideration and criticism of it in the next section. Second, note that what comprises the forcedness of the option here is that of losing out on a truth, if it is true. This is a significant element of James’s case, and it is also a particularly weak part. Recall that James’s initial account of forcedness was that, with forced options, one is presented with “a complete logical disjunction, with no possibility of not choosing” (WB 15). Now, we had to fill this notion out, so that it was not merely verbal when one says, “either go with this hypothesis or go without it,” as one could do that for any option. So, for any prima facie three-termed option O:(p, q, r) to have two hypotheses reduce to one practically identical option φ, to yield O:(p, φ) is that q and r be practically similar enough to be classed as such. The example is that of the runaway trolley car—one either leaps or not, as suspending judgment is equivalent to not. Now, note what is not part of identifying forcedness in this case: invoking what the options yield of one hypothesis is true or not. Doing that makes all options forced. In order to see this clearly, let’s return to an option James identified in Section I as not forced. Consider the first: Choose between going out with your umbrella or not. One can go with the umbrella or go without it. Or one could just stay in. So it’s not forced. But now take James’s new twist. The option is O:(umbrella, no umbrella, stay home). If the proposition that one should go with the umbrella be true, then no umbrella and staying home loses the good that comes with the truth—that of having an umbrella and taking a walk in the rain. But if the proposition that one should take the umbrella be false, we would avoid the error if we stayed home. See? We can make the umbrella case a forced option, if we invoke the truth or falsity of one of the options. This is because once we make a determination of the truth of one of the hypotheses, the others are identical in a relevant sense—they both are cases of “doing without the truth” we stipulated. Note that I’ve been following the less talkin’ more rockin’ rule: if James is going

Reading William James’s “The Will to Believe”

169

to have a careful and lengthy setup for the introduction to arrange things so the argument goes through just right, then I’m holding him to it. If James plays this way with religion, he makes the forcedness of the option an empty concept, one no different from any of the other ones he’d originally posed as not forced. It reverts to being merely verbal. Another way to consider my criticism of James’s case for forcedness is to consider it from the staunch atheist perspective. Take these thoughts seriously for a moment: God doesn’t exist. Moreover, it’s good to believe he doesn’t. And on top of that, it’s good he doesn’t exist. Plenty of atheists believe these things, me included.29 Let’s call this little triad the atheist hypothesis (AH). Now, I think the evidence is overwhelming for AH but I’ll argue for now that it may be a point of, ahem, faith. Here’s how James’s game can be played for AH. If AH be untrue, then by remaining skeptical, we avoid error. But, if AH be true, then we would lose the good in AH by being skeptical just as well as we would by being Baptists. So, from the perspective of AH, they’re all in the same boat.30 This is another form of the sauce for the goose argument—all of James’s tactics can be used against him by his opponents just as effectively as he uses them on them. If only he’d pause for a moment to take his dialectical opponents seriously. When he intones that the choice, again, is based on the courage of the Truth Norm or the cowardice of the Error Norm, he’s made it clear how he views those opposed: Dupery for dupery, what proof is there that dupery through hope is so much worse than dupery through fear? (WB 30)

He says he, for one, “can see no proof ” that one’s better or worse. But again, what of the moral arguments for evidentialism? Those were something like proof, and he’s got no answer. Some errors of belief commission yield morally egregious errors. That’s what the ship owner and island cases were supposed to show. Moreover, given James’s argument it is not a choice of dupery for dupery, as those who abstain never are wrong—there is no dupery for the evidentialist, by hypothesis. Only those who believe without evidence are the dupes. And here we see that the Jamesian line is that because abstention is practically identical to disbelief they may be placed together, as with both, one forfeits one’s “sole chance in life of getting upon the winning side” (WB 31). The question is why must abstention on and rejection of the religious hypothesis be treated as equivalent, beyond the mere stipulative verbal point? Ultimately, the case rests on the thought that one cannot act and live the religious life and reap its goods without belief—doubt and rejection, because without belief, yield the same

170

Evidentialism and the Will to Believe

practical outcome. But this requires that we be committed to the converse of All belief guides action, that All action is guided by belief. It is time to challenge James on his reliance on this move.

5. The conversion fallacy What is supposed to make the religious hypothesis a case of forced option is that if one God exists and we believe, our lives gain meaning because we are properly related to a divine entity and we are properly related to our own lives. If God exists and we don’t believe (either by suspending judgment or in rejecting his existence) we lose that possibility of meaning. Ellen Kappy Suckiel’s work on this section of “The Will to Believe” addresses precisely the trouble with seeing the religious question as forced. She concedes that while agnosticism is theoretically distinct from atheism, “from a pragmatic perspective, in terms of the concrete role religious belief plays in an individual’s life, [it is] no different from atheism” (1996: 29). This is, again, because if God exists, both atheists and agnostics lose out on the benefits of religious belief. As we’ve seen, the Jamesian case for that good in religious life is that of the vigorous life, a life of purpose. The question is, then, whether this is a justification less of belief than that of action. Take all of James’s examples for the composite version of the will-to-believe, WTB*. With the friendship case, the justification for the belief that the potential friend likes you is that it makes it so you will act in ways that makes it more likely he will. The same for courtship cases—the suitor believes the woman likes him so that he may confidently ask for her hand. The Alpine Climber believes he can make the jump so that he can make the jump. In social coordination, precursive faith in others is a precondition for joint endeavors. And now, belief in God is a prerequisite for participation in the life of vigor and meaning. But in each of these cases, it seems that we can produce the action without the belief. We can cultivate the habits of friendliness, skills of confident courting, abilities of capable and confident action in the midst of danger, and natural inclinations of doing one’s duty despite concerns about others doing theirs. And Pascal, himself, went to church and participated in masses and partook in the life of faith without the faith, at least at first. One needn’t have the belief to perform the relevant actions. If the justification for the belief is that it is necessary for the actions—that the good results of the belief justify it as their unique cause—the justification cannot work.

Reading William James’s “The Will to Believe”

171

Suckiel’s defense of James against this concern is telling, as I take it as both the clearest and most honest attempt to defend James on this matter. It fails, but the failure is instructive. Suckiel’s reply is that there is “a close, if not necessary, connection between belief and action” (1982: 80). My objection here has been that one can have the requisite actions without the beliefs. I can be friendly without the belief that we’re friends, and I can express the natural pieties and live vigorously without believing in God. The beliefs can be sufficient for these actions, but they are not necessary. The Jamesian reply, Suckiel holds, is that this distinction is invidious: James holds either that an individual’s thinking is invariably accompanied by action, or at least that thinking generates in the individual a kind of disposition to act in certain ways. Since believing is a kind of thinking, for James the reference to action is essential in any discussion of belief. It should be clear, then, that James would have rejected any criticism which was based on the separability of actions and beliefs. (1982: 80–81)

Suckiel concedes that the Jamesian case here presumes that the pragmatist view tying belief to action, and so the criticism is “quite appropriate” (1982: 81). But conceding that it begs the question isn’t the right concession here. Let us give James the pragmatic theory of belief and believing, so All belief is a guide for action. Note that Clifford, too, is committed to this view, and so given the dialectical demands of the argument, we can allow this premise. Also, observe that one still may hold that action and belief are separable, as there may be actions without the requisite beliefs. What James’s case needs is not the standard pragmatist view of belief—that all belief is a plan for action—but rather its converse, that all action is manifestation of belief. Only on the basis of that later proposition can the Jamesian case for the necessity of belief for action in the cases of friendship, courtship, social coordination, and vigorous living work. Suckiel’s argument establishes only that All belief guides action, not All action is guided by belief. Her concession that the argument begs the question is, then, off base, as the error with the Jamesian defense is not that a controversial (but plausible) premise has been presumed and is inaccessible to critics. Rather, an uncontroversial premise has been illicitly converted when deployed in further argument.31 I think the Jamesian has a reply. But it requires retooling the thesis. It goes like this: there can be action without belief, but that action will not be full, unapologetic, confident, all-in action that only belief can provide. So one can ask the pretty girl out without believing that she likes you, but it’ll be a pretty lame proposition. You can attempt that Alpine jump without the belief you’ll be

172

Evidentialism and the Will to Believe

successful, but it’ll be tentative and you’ll plummet. You can live a life expressing the natural pieties without religious belief, but it will be a life of weak and tepid tea. So it’s not that All action expresses belief that James must have in mind. It’s all full action expresses belief. There are no fully committed actions that are not expressions of coordinate beliefs. This saves James’s argument, but it comes at a price. First, it’s not how James’s examples actually run. So, really, it’s not James’s argument that gets saved, but a Jamesian argument that gets given. That’s not really that big a deal, because the argument still goes through, just not the original version. The second price to pay is of higher cost. The matter, now, is identifying what precisely real or full action is. The concept has the No True Scotsman concern written all over it. Here’s how the line of reasoning runs. A: No true Scotsman drinks English gin. B: Oh yeah? Well what about Wallace? He drinks stuff from London. A: Well, in that case, Wallace is not a true Scotsman.

The same, then, may go for the Jamesian version. And so it would run: Jamesian: No full action is performed without belief. Critic: Well, what about Bill. He successfully made friends with his neighbor without believing that they were already friends. Jamesian: In that case, Bill’s actions weren’t fully committed friend-making actions. Even if successful.

But if that’s how the game’s played with the concept, then why care for it? Especially for pragmatists, it’s a difference that makes none.

6. Religion as doxastically efficacious James’s case for the acceptability of religious faith depends, given the ascending case on the WTB*, on an analogy between the acceptable exceptions to the evidentialist’s rule (the friendship, romantic conquest, and social coordination cases) and religious commitment. For the analogy to hold, all cases must be genuine, intellectually undecided, and doxastically efficacious moral commitments. So far, James’s case for the genuineness of the religious hypothesis is pretty weak—the only one of the three conditions for genuineness that religion clearly passes is liveness. The question now is whether James can successfully make the case that religious truth is a moral issue that is doxastically dependent.

Reading William James’s “The Will to Believe”

173

James’s argument for doxastic efficacy of religious belief is on analogy with the friendship argument. Recall that James has added a rider to clarify the essence of religion, designated here by R1 and R2, with the commitment R3, that the entity guaranteeing eternality of the good is personal, a Thou. What religious belief comes to, then, is establishing a relationship with this personality. And so James proposes that, “any relation that may be possible from person to person might be possible here” (WB 31). And so, to clarify, James follows that analogy to friend-making through. He proposes that were a man in the company of good gentlemen to never advance conversation, never took anyone’s word and refused to ever budge, that person would be an outcast, incapable of being befriended. [He] would cut himself off by such churlishness from all the social rewards that a more trusting spirit would earn—so here, on who would shut himself up in snarling logicality and try to make the gods extort his recognition willy-nilly… might cut himself off forever from his only opportunity of making the gods’ acquaintance. (WB 31)

The Jamesian line is that without the belief, one is shut up in “snarling logicality,” which is a result of the conversion fallacy. The view is that one cannot be friendly without the belief that we are friends, and correlately, that one will be unfriendly if one does not believe. At least James is consistent. But notice that the line James is taking here is one that extends the friend-making model—the goods of religious truth, religion’s truth itself, must be doxastically dependent. That the beliefs are themselves doxastically efficacious, too, James takes as part of the religious belief: This feeling … that by obstinately believing that there are gods, … we are doing the universe the deepest service we can, seems part of the living essence of the religious hypothesis. (WB 31)

James has, then, further refined the religious hypothesis, adding the commitment of doxastic efficacy to the essence of the content of religious belief: R4: Believing R1, R2 (R2*), and R3 contributes to their truth, and thereby, believing them is a service to the universe. This, of course, requires that we further revise R2’s meta-component to: R2**: It is better to believe R1, R3, and R4 than not. And so, it seems, James has reconstructed religious belief into what his argument can support.32

174

Evidentialism and the Will to Believe

Earlier, I’d highlighted the two problems regularly attendant to pragmatic reconstructions, what I’d called the narcissism worry and the inconsistency worry. Recall that pragmatist reconstructions of truth subjectify truth to a point where it’s all personal and entirely plural. For anyone who’d thought truth worth saving from the skeptic or nihilist, this is terrible news. Truth was supposed to have a normativity to it, and once it’s reconstructed in pragmatic fashion (at least in Jamesian pragmatic fashion), it looks pretty empty. In fact, it looks a hair short of the nihilism it was supposed to answer. The question, now, is whether James’s reconstruction of the content of religious belief yields the same problems. For sure, traditional religious believers may have a trouble with James’s reconstructions. On the narcissism front, for sure, if religious life is about what it can do for you, if it is about the “social rewards a more trusting sprit would earn” (WB 31), it’s not genuine religious belief. Religion is not all about you. The same should go for any religious commitment on the inconsistency side. If you believe in God and are inspired by this relationship, then you are committed to His uniqueness, His special authority. That’s why the first commandments to Moses were up front—God is pretty jealous of those other gods. Consequently, pluralism is a degradation of true religion. In fact, pluralism may not even be a religion. A committed Muslim must think that Catholics are wrong about God being triune; a Buddhist must hold Baptists incorrect about life after death; a Calvinist must think the Jews are in error to deny Christ’s divinity. That’s what it is to hold those commitments—if we hold them true, we must think that others who hold views inconsistent with is believe false things. Any theory of religious belief that does serious damage to those commitments (especially if they are held as inessential) may do so in the name of pluralism and tolerance, but it does so to the detriment of the commitments it was out to save. Recall that in the Prologue, James had announced that his essay was out to provide a “justification of faith,” and one that allows believers to retain their beliefs in the “vital subjects” of salvation and God’s properties (WB 13). I’d called this content objective of the justification of faith the orthodoxy requirement, and it’s roughly that if you’re going to save religious faith from its critics, the faith you save must be the faith that’s criticized. If the faith that comes out of the salvation is transformed to a point wherein the religious believers who’d looked for it to be saved can’t recognize it, then we have a problem. It’s the old switcheroo. Sometimes this argumentative strategy is called the sharpshooter’s fallacy. Here’s how it goes. You can hit bullseyes every time. Shoot your gun at a blank target then paint bullseyes around your holes. In James’s case, he promises to

Reading William James’s “The Will to Believe”

175

save faith, and he proceeds to give an argument. He then calls whatever is saved by the argument the essence of faith. That’s painting bullseyes. The old switcheroo works the same way. Let’s say I want the nicest car money can buy. But I don’t have much money. I go to the dealer and he promises me the nicest car money can buy. I give him my money, and he gives me this stinky old jalopy. He then says this is the nicest car my money can buy. Ah, switcheroo! I believe the switcheroo problem occasions a dilemma for the Jamesian joint reconstruction of religion and the justification of faith. Either (i) faith is doxastically efficacious or (ii) it is not. If (i) it is, it may pass the lawfulness requirement by being a special case of evidential backing (not as an exception to the evidence-backing requirement), but in making religious belief so efficacious, it no longer can pass the orthodoxy requirement. Faith that is lawful is no longer the faith of the faithful. This, again, is the trajectory of the later pragmatist reconstructive tradition, but this is not what James himself promises to open the essay. If (ii) faith is not doxastically efficacious, then James may pass the orthodoxy requirement, as the commitments to trinitarianism, virgin births, and so on may yet count as faith (and not be doxastically efficacious), but the case for the lawfulness of these commitments falls. That is, if James allows faith in miracles in the distant past (n.b., precisely the sort of beliefs Cliffordian evidentialism was designed to prohibit), he may keep the faith as recognizable to the faithful, but all the apparatus arranged to demonstrate its acceptability becomes otiose. The past is fixed, eternal truths are, too—our belief in them or not will have no effect on them, and so they are not doxastically dependent matters.

7. Evidentialism as irrational If the religious hypothesis is true, but evidence for it is insufficient, then the evidentialist would hold that we must suspend judgment on the matter. Even if it is true, if we do not have sufficient evidence, then we are not believing responsibly if we believe it. James holds that if a view is true, then if there is a rule that prevents us from holding it, the rule “would be an absurdity.” Especially in case wherein “some participation of our sympathetic nature [is] logically required,” the prohibition against belief puts a “veto on our making willing advances” (WB 31). As it goes, the Jamesian line is that with morally significant doxastically efficacious commitments, evidentialism is a practical impediment. We will

176

Evidentialism and the Will to Believe

assume here that James is right that suspension of judgment yields apraxia. Consequently, I take it that James’s argument is as follows: P1: Evidentialism requires we suspend judgment on any hypothesis for which we have insufficient evidence. P2: Some hypotheses for which we have insufficient evidence are morally significant and doxastically efficacious. P3: If we suspend judgment in the cases identified in P2, we prevent the morally preferable option in the morally significant case to come to pass. P4: It is rational to take steps to ensure morally preferable options come to pass, and it is irrational to prevent them. C: Therefore, following the evidentialist norm is irrational. We can allow, here, P1 and P2. I’ve argued earlier that P3 is false, as it depends on the conversion fallacy, but I will allow it here. The question is whether P4 is true. In some cases, it seems, in following the Truth Norm, the morally preferable option is having the truth, in other cases, following the rule of wider desiring natures, having friends, winning boons, and being properly coordinated with others is the morally preferable outcome. James then concludes his case: [A] rule of thinking which would absolutely prevent me from acknowledging certain kinds of truth if those kinds of truth were really true, would be an irrational rule. (WB 31–32)

Rules of logic are rendered absurd if they stand in the way of true belief. For sure, the aim of completeness is a worthy one for systems of reasoning— if there are relevant truths your rules for discovery cannot uncover, that’s a problem for your set of rules. But there are other objectives with our systems of reasoning. One is soundness—that all the commitments arrived at by the system are true. Another might be systematicity, that we can see how elements of the system bear on each other, work together, and clarify and buttress each other. Further objectives might be simplicity or fecundity, or beauty. It’s often a trade-off with systems of reasoning which parts we want at all costs and which ones we might make some sacrifices with. I presume that when James identifies the religious hypothesis as a certain kind of truth, he’s not just picking out any old truth. Evidentialism’s rule prevents us from believing willy-nilly whether Cleopatra sneezed on her birthday, because we have no evidence for that proposition. But James is not interested in that truth, but in the truth of religious belief (and personal connection and social coordination). We not only prevent our believing a preferred truth, but we also prevent the preferred truth.

Reading William James’s “The Will to Believe”

177

Consequently, it’s not clear that P4 is true, or at least it’s not clear that it’s always irrational to forego any means to achieve one’s preferred ends. I want to believe many significant truths, so one means of doing that might be to believe many, many significant propositions. In doing so, I may come to believe many false propositions. No matter, as I have a light heart about error. Consequently, I get what I want: belief in many significant truths. Here we can see why the Error Norm is important, and it’s not merely for the moral reasons given by Clifford. Even if I hold that error is not intrinsically horrible, it has an aggregating effect such that if I add errors together with the true propositions I believe, the true propositions are drown out. The trouble with error is that with too much of it, the truth has a noise-to-signal ratio such that it cannot play the role in our lives we would want it to play. If I regularly overbelieve on the basis of my preferences and allow my willing nature to carry me to these many commitments, I will have a wide and plural range of views, to be sure. But I am a conflicted person. The liveness of options seems to fluctuate for me, and in many cases, I have equally live but inconsistent hypotheses appealing to me. If I want the truth, why not believe both? My point is that the Error Norm plays its role for a number of reasons. First, error is bad in itself. Belief aims at truth, so false beliefs are busted. Second, errors are practically costly. That’s Clifford’s argument from before. Third, errors aggregate and drown out truths. It is, then, for the sake of truths we can have that the Error Norm is important and worth living in accord with. And so for the sake of truths for which we have evidence, we must forego truths for which we have none. A final question of evidentialism would be that if religion is true, is evidentialism really the “snarling logicality” James portrays it to be? Is the evidentialist haughty against God? The first thing to note is that the problem of divine hiddenness has at least put many of us in a position wherein we have no access to a truth that would, by hypothesis, be a great good to us. And not having it, again, by hypothesis, is a great evil. For one, I think that if God exists and He’s not given us evidence of His existence, then He’s done us a wrong. This is the problem of cognitive evil. If God’s presence in our lives is a great good and His absence is a great evil, it’s unjust for Him to hide from us.33 No parent would ever do that. This view may be “snarling logicality,” but some hypotheses deserve to be snarled at. But, now, consider the alternative. If evidentialism is so bad, how does the WTB* really fare? I think it would be worse—if it’s a right to believe what we prefer, to fix God’s nature by what is live to us, then it is tantamount to idolatry.34

178

Evidentialism and the Will to Believe

8. Religious tolerance The final stage of Section X of “The Will to Believe” is James’s case that his willto-believe doctrine makes room for a kind of religious tolerance. This tolerance is based on the recognition of a right to believe at our own risk, to try for the truth on our own. Our failures will be ours, and our truths, too, will be ours. Notice that we are not free to believe things we know aren’t so, as the right is restricted to undecidable genuine options with the relevant factors. James is clear that with things we know are not so, we cannot apply the will-to-believe. James’s hypothetical schoolboy gets faith wrong, then, when he says “Faith is when you believe something you know ain’t true.” The will-to-believe is restricted to genuine options, specifically limned by liveness: In concreto, the freedom to believe can only cover living options which the intellect of the individual cannot by itself resolve; and living options never seem absurdities to him who has them to consider. (WB 32)

Because the schoolboy’s faith case is recognized as an absurdity, it cannot be live; and if it cannot be live, it cannot be a candidate for the will-to-believe as WTB*. Consequently, because James’s case is heavily restricted, the objects of faith are going to be limited. What, then, of miracles? Of belief, say, in the trinity? Other mysteries? These are not cases of belief on the basis of insufficient evidence, but against one’s evidence. Does the Jamesian restriction of liveness make it impossible, say, for Saul to become Paul? (Christianity was so dead for Saul, he preferred Christians dead.) Again, my suspicion is that what might be the more radial Kierkegaardian take on faith, belief in absurd things, is a step too far for James’s defense of faith.35 Consequently, the Jamesian restriction makes the standard fare for the faithful inaccessible. Independently of this point is the scope of what will pass muster as contents of faithful belief on the will-to-believe. That faith is an expression of the courage behind the Truth Norm is James’s primary hortatory focus. We are not made to wait, to have a stopper on our hearts. We are made to act. James holds that because these matters are forced, we must recognize that others will make different decisions, choose other options. In light of this, an ethic of tolerance must be observed: No one of us ought to issue vetoes to the other, nor should we bandy words of abuse. We ought … delicately and profoundly to respect one another’s mental freedom. (WB 33)

Reading William James’s “The Will to Believe”

179

This objective of tolerance is what James had announced in the Preface, that he has “preached the right of the individual to indulge his personal faith at his personal risk” (WB 8). Recall that pluralism is the term of art to euphemize the situation of facing an inconsistent and conflicted mess of claims. Given that liveness is subjectsensitive, and people’s natures vary widely, the breadth of acceptable contents for faith will be conflicted. We will, then, accept that, if we are to allow not just evidence but our wills to lawfully determine beliefs, that we will have lawful and interminable inconsistency. The question is how tolerance follows from this view. James sees, once we have this commitment, an “intellectual republic,” one in which the inhabitants live with an “inner tolerance,” wherein we “live and let live” (WB 33). But how does this follow? If there are no reasons, but only motives with faith, then why not provide others with motives to convert? On what basis do mere standing motives deserve respect and tolerance? Why not anarchy, instead of tolerance? Consider, if liveness is a matter of education, nature, and enculturation, then why not educate and enculturate those who reject your views? They happen to be that way, but they can just with a few changes be made to be this way. Or consider the core commitments of the Christian who holds that Jesus is the light of the universe, the king of kings, the son of God. Is it possible to hold that belief in the way it must be held to be properly related to Jesus and yield the kind of life it is supposed to, and then also say that the matter is merely one of liveness of subjects and that those who reject Jesus in favor of some other favored nonsense are within their rights to do so? For the tolerance to follow, the religious believer must believe in a way that is, well, less than religious. It may be a commitment internal to the religious view that tolerance is required, but that’s not a consequence of pluralism. That’s just a happy coincidence about the particular religious view. Tolerance must follow from the pluralism, not from one of the substantive commitments of the tolerated religions. Bertrand Russell presses this objection against James by noting that the WTB justifies “an immense multiplicity” of competing beliefs between different people: One gathers… from [James’s] instances that a Frenchman ought to believe in Catholicism, an American in the Monroe Doctrine, and an Arab in the Mahdi…. It seems odd that … he should maintain acceptance of his doctrine should diminish persecution; for an essential part of each of the above creeds is that people who think otherwise must be taught their place. (1996: 87)

If the live religious views are intolerant, how does the WTB (or WTB*) constrain them? Only if they are properly deflated or reconstructed can any of the elements

180

Evidentialism and the Will to Believe

of James’s program work on them. Otherwise, the will-to-believe is less an invitation to toleration than it is to endless strife. In his letters, James himself concedes that his own views on the matter comprising those of religious belief are less than orthodox: I believe myself to be (probably) permanently incapable of the Christian scheme of vicarious salvation, and wedded to a more continuously evolving mode of thought. (1920:149)

Jamesian pluralism is, then, a double curse. First, James’s account of how the values arise so that we can acknowledge their plurality deflates the commitment to the contents. Once we see religious belief as merely expressive of our passional nature, we may allow it as an excusable exception to the evidentialist norms, but the content of those beliefs are deflated. It is no longer clear why we were so worried about them in the first place. Second, once we’ve come to see these religious views as mere expressions of what is live and what is not for these people, then is it not putting a stopper on the heart of the jihadist or crusader to tell them that they’ve gone too far? But they, too, cast their lots, hoping to be on the winning side, to throw the last stone. Winning a religious war is the way to ensure you throw the last stone, you know. James has given an argument for religious pluralism, and it has been based on the background thought that we have individual rights to overbelieve in cases that meet the will-to-believe conditions. The first trouble is that pluralism itself has no clear normative consequences. Anarchy or totalitarianism, just as well as tolerance, can follow from the fact of pluralism. The second trouble is that James’s case for the plurality of religious views itself reduces them to the stipulated “essence” of religion. This reduction is also a deflation. A further concern is whether this reduction is really pluralism. There, of course, is an irony to a view that, on the one hand, calls itself pluralist, but nevertheless takes religious commitment to have a singular essence. In saying this, the plurality is acknowledged, but what is plural is what’s accidental to any of the religious views on offer, as they must have the essential features identified to count (and be tolerated) as religious in the first place. Call this the meta-monism problem for Jamesian pluralism—to recognize the plurality of religions, they must be recognized as expressions of the same essential thing.36 It’s not tolerance if you think, in the end, the matters of disagreement aren’t that important.

3

The Ethics of Belief and Philosophy of Religion

In this closing chapter, I want to pursue a few issues that the Clifford–James debate occasions. Some of these matters are purely epistemological, others bear on pragmatism’s scope, and still others on philosophy of religion. I will organize my discussion here in terms of what I call lingering questions.

Question 1: Must evidentialism be an ethical doctrine? Recall that Clifford’s argument for evidentialism was of the basic form: overbelief either directly or indirectly yields action that endangers others. One shouldn’t endanger others, so one shouldn’t overbelieve. Again, my thought is that the nonendangerment principle can be attributed to either deontologists or consequentialists, but that Clifford is most likely a stoic deontologist. Regardless of what kind of ethical argument he provides, it is clear it is an ethical argument for evidentialism.1 What makes Clifford distinct from his Enlightenment predecessors is that they give primarily cognitive justifications for the versions of the evidentialist rule. Hume invoked the theoretical virtue of wisdom, and Locke speaks of rationality. This cognitivist, as opposed to practicalist, case for evidentialism has continued, as we see it with contemporary thinkers such as Richard Feldman (2000 and 2006), Jonathan Adler (2002), Nishi Shah (2006 and 2008), Robert Talisse (2009), and myself (2007, 2006a, and 2006b). I’ll present my own version of the cognitivist case for evidentialism, but then, I think, it is worth noting that the special practical cases for evidentialism, that from ethical considerations, are consistent with the cognitivist and (assuming they are also correct) further determine the correctness of evidentialism. Here is the basic thought behind the cognitivist view. You simply can’t maintain a belief as something you hold as true if you view yourself as holding it

182

Evidentialism and the Will to Believe

on anything else but evidence for its truth. So, consider any of the weird cases that evidentialists must face. Perhaps the faithful husband case: Larry loves Linda, but she’s been spending lots of time with Sam. She’s come home with hotel matchbooks in her pockets. Sam calls at strange hours. But because he loves Linda, Larry believes in her and tamps down his suspicions. He believes Linda is faithful, and he saves his marriage.

Or consider a more extreme example, the save the world case: Larry’s got one chance to save the world from the aliens. They will blow the place up unless he can bring himself to believe, contrary to all the evidence they provide private-detective-style, that Linda is faithful. Larry is able to force himself to believe Linda faithful, and he saves the world.

For sure, Larry’s got practical reasons to believe Linda’s faithful.2 But these practical reasons can’t provide Larry with the circumspect motivation to believe that Linda’s faithful. He can wish she’s faithful, prefer that he believe he believed it, but the fact that having the belief is practically preferable cannot be the reason Larry gives to himself when he considers why he believes it. Such a reason would be in the form of what I call: Conflicted Thought (CT): There is no evidence for p’s truth, but I believe that p because believing that p is practically preferable to not. The problem is that any time one makes those reasons explicit to oneself in CT fashion, the belief is no longer a belief.3 So Larry can force himself to say Linda is faithful, or hope, or wish it, but he can’t believe it when he sees it in light of the reasons given above. Again, he may, in Pascalian fashion, bring himself to focus his attention on the evidence for Linda’s faithfulness and do what he can not to think of evidence for her cheating, but that would be for him to (like with the ship owner in Clifford’s cases) try to rationalize the evidence before him. That may bring about a belief, but that just shows that belief has internal norm— beliefs must be evidence-sensitive. Presenting yourself with CT does the trick to remind you of that. The main trouble with my line of argument is that there is a serious hole— what, then, would be the reason not to rationalize one’s evidence?4 Brian Zamulinski’s criticism of the cognitivist’s arguments (specifically those from Jonathan Adler, which mine parallels) is that the evidentialist’s norm ends up being merely verbal with the deployment with CT-style critical reflection. If we take anything like [the cognitivist’s] approach, however, it does not follow that we ought to regulate our beliefs … The most we could do is stop calling the

The Ethics of Belief and Philosophy of Religion

183

propositional attitudes in question “beliefs” on the ground that they really are another sort of thing. (2004: 81)

This criticism seems correct, as far as it goes—if we see the commitments in light of CT, they no longer can be called “beliefs.” The trouble, as Zamulinski notes, is that this may be a difference that makes no difference. Here is the reply. If these propositional attitudes are not beliefs, they are not determiners of action in the way that beliefs are, and so we have no longer the same problems with their ethical upshot. Wishing that Linda is faithful is not just a different propositional attitude for Larry, it’s a different practical attitude. Zamulinski’s challenge, however, can be reformulated. Again, if Larry wished he believed Linda faithful, then he might plot against himself to rationalize his way to the belief. Now, that’s a belief, it passes the CT, but it’s still defective. The rationalizing prevents us from thinking the CT in its bubble-bursting form. Consequently, the Zamulinski challenge is that we must have ethical reasons not to rationalize like this, as, like Clifford, he holds that there are bad consequences for those who rationalize and those about whom they rationalize. I don’t dispute Zamulinski’s argument, though I think it should be made with a deontological edge. My reply is that it is hasty to rush to practical reasons here, as the CT schema only needs some tweaking to prevent the rationalized version we see getting around it. Consider the way that the following thoughts are conflicted in CT fashion, but have rationalization (and other distortions) as part of what yields the conflict: ●

●

●

●

●

I believe that p, but only by gerrymandering the evidence to support it. I believe that p, but by ignoring all the counterevidence. I believe that p, by attending only to supporting evidence. I believe that p, but all my evidence has been filtered so that it will always support p. I believe that p, but on the basis of insufficient evidence that I’ve trumped up to look, when in my uncritical and desperate moments, to strongly support p.

In this case, we can see that the CT has a meta-application, too, and it can be deployed to ensure that the assessments are not merely verbal about what we call “beliefs,” but about what we can reflectively endorse. Consequently, I do not think the ethics of belief must be grounded in practical rationality. It most certainly can be, and I think that the practical justification of evidentialism is a welcome co-collaborator in its justification. Other practical

184

Evidentialism and the Will to Believe

justification for evidentialism, beyond the ethical one we see with the Clifford– Zamulinski–Wood tradition, are pragmatic and social justifications. The rough pragmatic justification is closely allied to Clifford’s, which is that if one wants one’s plans to succeed, one should take one’s means to one’s ends. Having true beliefs and not having false ones is instrumental in having one’s plans come to fruition, and following one’s evidence is the best means to that. So one should, for pragmatic reasons, be an evidentialist.5 The social argument is that beliefs are commitments, moves in a social game of giving and asking for reasons. One rule in a social game such as this is that one have reasons for commitments. If you have beliefs, you can share what counts in their favor with others. So if you have beliefs without backing, you’re cheating in the game of language use, and you don’t treat the other players in that game with the respect they deserve.6 The consequence, as I see it, is that there is a wide variety of practical cases to make for evidentialism, and there is nothing intrinsic to the pragmatist tradition, then, that makes it so that pragmatists mustn’t be evidentialists. Contemporary neo-pragmatist work embraces the evidentialist tradition, and I believe it is clear that this strain of pragmatism is the right-headed Peircan part of the family. Consequently, as I noted earlier, I see the Clifford–James debate as internal to the nascent pragmatist tradition, not a pragmatist criticism of an anti-pragmatic line of thought.7 Clifford was committed to the view that belief was a guide for action, and this view is a pragmatist touchstone. In the end, I take a pretty big-tent view of pragmatism, that it is a wide and varied tradition with wide and varied influences. The fact that its list of founders is populated with selfidentified empiricists, Hegelians and positivists, and its current practitioners self-identify as Foucauldians, Sellarsians, and phenomenologists too is evidence that pragmatism is a very, very broad tradition.

Question 2: Can practical reasons trump theoretical reasons? From the Larry cases before, it’s easy to see that sometimes, practical reason conflicts with theoretical reason. There are inconvenient truths and convenient fictions aplenty. So, what should happen when those two kinds of reasons conflict? Notice here a separate motive for cognitivist evidentialism—that it can escape from special versions of problems like this. The reasoning behind the cognitivist tradition runs that if you hold, like Clifford, that evidentialism is

The Ethics of Belief and Philosophy of Religion

185

exceptionless, then it’s best, unlike Clifford, to be cognitivist. This is because if you’re an ethicalist (or broader practicalist), then the bindingness of evidential norms is derived from practical reason. But there’s nothing intrinsic to, say, the norm that requires that we not endanger others, that guarantees that it never can be trumped by another practical reason. Another norm, perhaps one that should save the world from destruction, can trump it. Let’s put this line of reasoning into motion. Surely, someone who holds that evidentialism has serious counterexamples would also hold that the ethical consequences of endangering others with one’s credulity in Larry’s save-theworld case pales in comparison to the ethical consequences of aliens blowing the place up. Were Larry to somehow rationalize the belief to himself and yield the belief, we surely shouldn’t blame him. It may be too strong to hold that he’d be blameworthy were he not to do so. Now, if you’re a cognitivist, these practical deliberations are beside the point, as they trade on a coinage that the cognitive goods do not recognize. In the case of deliberating about consequences or ranking our duties, we have a rough idea of the one kind of thing we are out to maximize gains in and minimize losses of. But cognitivists hold that if evidentialism is justified entirely by other kinds of goods, the trumping can’t happen quite as easily. These values are, we might say, disparate. There are the values of practical rationality and then there are the values of cognitive integrity. The practical values regulate what we do and the cognitive values determine what we believe. The Jamesian case, from the cognitivist perspective, is, then, a category error. I believe this is an appealing element of cognitivism regarding the evidentialist norm. But it is too strongly stated, and it is because there can be excused practical exceptions to the evidentialist norm. This makes my ultimate view what I call modest evidentialism, where the rule is exceptionless in terms of our duties, but we can excuse some cases of breaking it. Consider the case of the faithful friend: Carrie has been charged with murder and the evidence is piling up that she did it. But Carrie needs a friend while in jail, and Susan’s her very best and only friend. Susan can’t be of comfort to Carrie if she believes Carrie a murderer. Susan, then, resolves to studiously ignore the evidence as it rolls in and attend only to her best memories of Carrie. Thereby, she’s able to believe her likely innocent and be of comfort to her.

The demands of evidentialism (precisely on the prohibition of evidential sloth) and the demands of loyalty conflict here. Is Susan obligated to believe Carrie innocent? No. Is she obligated to believe her guilty? Yes, but were she to fail

186

Evidentialism and the Will to Believe

to do so and hold her innocent or studiously avoid holding a view, her belief is excusable in the circumstances. Notice, by the way, that this thought works out only in cases of Susan playing the role of loyal friend, not jury member. Special role-related duties can determine whether there are excusable cases of overbelief, despite the fact that the ultima facie norm of evidentialism is never trumped.8 If there are excusable failures in following the evidentialist norms, then two things must be clear. First, how do practical reasons excuse in those cases where one contravenes a theoretical reason? A longstanding cognitivist line is that they never can. If they are of different coinages, then how does one bear at all on the other? Second, doesn’t this open a reversion back to Jamesian willings-tobelieve? To begin, what does cross-categorical assessment look like? The main concern about such assessment is what might be called the apples-to-oranges problem. Put simply, because they are different kinds of goods, we have no systematic or principled way of articulation on the basis of which we have a reason for taking a practical norm to be an excusable reason to contravene a theoretical norm. Or vice versa. Richard Feldman, a defender of the cognitivist view, holds that the apples-tooranges objection seals the two realms of evaluation from each other. In “The Ethics of Belief ” (2000), Feldman has argued that in committing to the thesis that moral duties trump epistemic duties, one commits to “a sort of generic ought that encompasses moral considerations, epistemic considerations, and perhaps others, and that weighs them against one another to come up with an overall assessment” (692). According to Feldman, cross-type overriding requires a just plain ought in terms of which the two types of norms may be compared. The problem, Feldman argues, is that there is no way to make sense of the notion of the just plain ought. Feldman establishes the unintelligibility of just plain ought by two analogies. The first analogy is that we have two dolls—one short and squat, another tall and thin. A child asks us which one is “bigger.” We cannot tell which property properly “trumps” the other for one doll or the other to be “just plain bigger.” As a consequence, Feldman proposes that “there is no such thing as ‘just plain bigger’” (693). The second analogy is that we have two people, A and B. A is stronger than B, and B is wealthier than A. Given this, we ought not to assume that there’s some proper way of combining these two measures into one, which we might call “strealth.” There are countless ways of combining and weighing the measurements of the two properties to yield the difference, each yielding

The Ethics of Belief and Philosophy of Religion

187

different strealth rankings. If it is objected that one of these rankings is based on an inappropriate weighing of wealth to strength, Feldman proposes “we would say we have no idea what that person is talking about,” because we have no independent notion of strealth for which there are correct weightings of the two components (693). Neither of Feldman’s analogies yields the conclusion that these notions of “just plain bigger” and “strealth” are unintelligible. First, take the dolls. Let the one that’s taller be a whole lot taller and the one that’s wider be only negligibly so. I, myself, don’t see any problem with saying that the taller one is just plain bigger—it’s taller, it has more volume, and though the other’s wider, it doesn’t make much difference. If the child objected that the squat doll is nevertheless wider, I would concede that, but I would also ask whether that makes a difference in our overall assessment, given the other differences. If it does, then we may revise the assessment. That the assessment of “just plain bigger” is something we may deliberate about is one thing, that it makes no sense at all is another. The same goes for “strealth.” Certainly it would be a curious concept, but the fact of there being hard cases for comparison does not show that the concept is unintelligible. There may still be easy cases—take A to be considerably stronger than B, but B to be only negligibly more wealthy than A. Ranking A over B in strealth seems perfectly intelligible, and perhaps (depending on what we want to do with the rankings) right. Moreover, if someone objected to this ranking, we would most certainly know what she’s talking about. She’s questioning whether the wealth difference really is negligible. The point of Feldman’s analogies is to show that we do not have an independent concept of these cross-type assessments. He, I think, does not show that with “just plain bigger,” as there are cases where it seems right to use the notion. He may be right about “strealth,” but that’s partly because of how he devised the example—of course, we won’t have an independent notion if the story introduces the concept dependently. Regardless, Feldman’s objective with the analogies is to show that these cross-type assessments require an independent standard. On analogy between “just plain bigger” and “strealth,” the required notion of “just plain ought” for modest evidentialism does not make sense. Feldman argues this is because there is no independent notion by which we can articulate that kind of ought: It’s far from clear what that value would be—it isn’t to be identified with any of the more determinate kinds of value and there seems to be no uniquely correct … way to combine moral, practical, and other values…. There is no meaningful question about whether epistemic oughts “trump” or are trumped by other oughts. (2000: 694)

188

Evidentialism and the Will to Believe

The thought, then, is that we cannot compare and rank competing requirements unless they share a system of value. Given that moral oughts come in the same system, they can bear this relationship with each other, and Feldman concedes this point. It is cross-type assessment that is the issue.9 The question now is as to what our theory of cognitive norms must be beholden. On the one hand, Feldman’s argument is that evaluative (ultima facie) relations between norms must be articulated in a shared value. Since cognitive and practical norms do not share the same values, a norm from one type cannot override that of another. On the other hand, it is clear that from the save-theworld and faithful friend cases, there are reasonable comparisons we can make between these normative judgments. Feldman’s requirement does not make it false to say so, but makes it nonsense or a category mistake. This hardly seems correct. Our theories and intuitions aren’t in reflective equilibrium, and in this case, I think the theory should go. Feldman’s requirement of a shared notion of value extends to all overall cross-categorical evaluations. Figure skating, for example, can be judged with regard to technical difficulty and with regard to artistic elements, but is it nonsense to also ask for an overall evaluation? Beers are judged with regard to how well they instantiate their respective types, but it is still possible to ask which is the best overall beer. Is it nonsense that kennel clubs have awards for best in show? One way to save the appearances here is to make a distinction between the commensurabilty and comparability of goods. Goods are incommensurable when they do not share a third good by which one may articulate the rationality of preferring one to another. They are commensurable when they do. Goods are incomparable when there is no right or rational decision to be made between the two, and they are comparable when there is. Feldman’s argument is that because cognitive and practical goods are incommensurable, they are incomparable. This is mistaken. The fact of incommensurability of goods does not yield incomparability. Surely friendship and money are incommensurable. But it is still reasonable to prefer friendship to money. One may not be able to exchange money for friendship, but this is because they are not fungible. That is, they are not rationally exchangeable—paying someone to be a friend renders the friendship something else. But one may, nevertheless, value friends more than money, and even wish one could trade one’s money for friends, despite their being incommensurable and infungible. This is not category error. In fact, it is the position many inhabit—the billionaire who says she’d trade it all for a good

The Ethics of Belief and Philosophy of Religion

189

friend isn’t speaking nonsense; she’s just saying she prefers friendship to money, that it’s just plain better to have friends. Take some other cases of incommensurable goods yet yielding rational preference rankings. Money and education are incommensurable, but yet it still makes sense to prefer an education to having a sum of money. It even makes sense to exchange the money for the educational experience. If it did not make sense to have this rational ranking, tuition would be nonsense. Further, take a case where a practical good trumps a cognitive good. Vivisection is morally objectionable, and that overrides a good many cognitive goods that come from pursuing that research program. The Tuskeegee syphilis experiments were unethical, and that is the reason why inquiry is morally constrained. Relatedly, Philip Kitcher has recently argued that research concerning the natural abilities of the races should be avoided, because its conclusions (regardless of their content) put minorities in harm’s way.10 The point here is that there are cases where practical and cognitive goods can be compared and rationally ranked despite the fact that they are incommensurable. Feldman may be right that cognitive and moral values are incommensurable, but it does not follow that they are incomparable. The fact that we have intuitions about the save-the-world, faithful friend, and the other cases above and that it not only makes sense but seems correct that some decisions are better than others is evidence of such comparability. That is, it not only makes sense to say that Larry is perfectly reasonable for believing Linda faithful to save the world, it is true. Articulating how one judgment is rational and another is not is, of course, difficult in cases where the goods are incommensurable. We cannot point to the maximized shared good as the criterion for assessment. In these circumstances, only a kind of right judgment can pick out the correct option. We can subject these judgments to scrutiny and challenge, and the ones that are correct are the ones that survive that kind of reflection. The second problem for modest evidentialism is explaining how this isn’t another Jamesian WTB*-style view. In a way it is similar, as James, too, acknowledges the evidentialist norms as “healthy,” but holds that there are exceptions with a right to overbelieve and preferable specific overbeliefs. There are two important differences. First, I acknowledge no right to these overbeliefs, whether they are Larry’s commitments in faithful husband or save the world or Susan’s commitment in faithful friend, is nevertheless cognitively wrong, but it is practically excusable.11 Susan’s belief, particularly, in faithful friend, is improperly

190

Evidentialism and the Will to Believe

held, but the practical reasons bearing on her belief mitigate whatever blame we would have for her. Now, notice, that this is, again, a matter of role-related duties. Susan has two roles here: believer and friend. For her to live up to one, she fails the other. But notice also that circumspection here is important—Susan holding that belief is excusable only insofar as the matter is her being a friend. Were she to have duties as a juror or other position, she would be in practical error. The faithful friend case has Carrie already in jail. Were Carrie just living next door and Susan had all that evidence, things would be different. Context matters, and so many of these cases of cross-categorical assessment require judgment. But that it’s complicated does not make it mysterious.

Question 3: Can religion be pragmatically reconstructed? Recall that one of my objections to James’s defense of religious faith was that he was guilty of the old switcheroo. He promised to save something like conversation at “good old orthodox college(s)” (WB 13), but he serves up something I think very few religious believers would recognize as what they wanted defended from atheists and agnostics. So he fails the orthodoxy requirement in reconstructing religion into the religious hypothesis he can save. He called that reconstructed part the essence of religion, but that seems very thin. It’s the sharpshooter who paints bullseyes around holes he’s shot in the target. The lesson, as I take it, is that if we are to save religious commitment by reconstructing it in defensible fashion (so that it can pass the lawfulness requirement), we are likely to fail the orthodoxy requirement. That’s the dilemma. Either you get the beliefs in recognizable but indefensible form or you get beliefs with acceptable contents that seem to have very little religiousness about them. James tried to split the difference, but he ended up on the latter horn of the dilemma. Too bad. My view is that it’s not so bad on that side of the dilemma, so long as you’re honest that you’re saving a reconstructed version of religion. That’s where James went wrong in “The Will to Believe”—it was false advertising that what the religious believers in his audience hold was faith was what was to be justified. He pulled the switcheroo. But if we were to have a kind of direct take on religion, one that made the reconstructive strategy clear at the outset, we would not have this problem. John Dewey’s strategy of A Common Faith (1934) is precisely along these lines, and I think that the real value of “The Will

The Ethics of Belief and Philosophy of Religion

191

to Believe” and James’s subsequent work was to provide a blueprint for later pragmatist reconstructions of belief.12 We, the later inheritors of the pragmatist tradition, also inherit James’s strategy. This is what I take to be the best lineage from James’s work.13 The difference, I think, is that we must be clear that we are not going to be saving what the religious believers think matters. That’s what reconstruction is. James’s essay is a marker of a broader reconstructive trend in the pragmatist tradition with religious belief. Peirce rejected a number of commitments internal to religious life in light of his pragmatic criterion (especially the transubstantiation thesis in “How to Make Our Ideas Clear” 1960: 257; CP 5.401). But many of the pragmatists in the subsequent tradition are considerably more irenic. John Dewey later followed James’s lead and reconstructed the content of belief in God as “the ideal ends that at a given time and place one acknowledges as having authority over his volition and emotion, the values to which one is supremely devoted, as for as these ends, through imagination, take on unity” (1934: 42). Cornel West (1989) and Richard Rorty (1997) follow in this line, too, as they similarly take a more ethical-Romantic interpretive agenda with religious commitments over their purportedly metaphysical contents. And this line of thought in the pragmatist tradition intersects with the Wittgensteinian and expressivist traditions with religious belief in D.Z. Phillips (2001) and Michael Hodges (1990) and in Hodges’s collaborative work with me (2006, and forthcoming). These are broadly practical-hermeneutic programs of reconstruction. It is important to note that as these programs proceed, each must pause to note that they will be revisionary interpretive theories—they will be presenting accounts of religion that will take the believers’ beliefs as data, but will not take what the believer believes that she believes as the criterion for interpretation. As a consequence, pragmatic reconstructive programs must acknowledge that they are reconstructive, they do not save the faith of the faithful, at least as the faithful see it. They, rather, save what the reconstructors take as the relevant and defensible parts of these faiths. What remains is often quite a surprise to the true believers, as it looks surprisingly like the humanism and secularism they abhor. For programs that suspend the orthodoxy requirement, this is not an internally damaging criticism. But when one presents the program as one to be amenable to those who want to maintain the discussions of the means of salvation and the relations between God’s attributes, such a turn is damaging. This is the case for James’s “The Will to Believe.” For better or worse, the religious beliefs it takes in for reconstruction come out unrecognizable, and so they fail the orthodoxy requirement.

192

Evidentialism and the Will to Believe

Can the problems that plague pragmatist reconstructions, the narcissism and contradiction problems, be resolved? I think the answer is a tentative yes. Here’s a rough sketch. We shouldn’t be out to reconstruct religious beliefs. They aren’t worth saving. They are indulgent overbeliefs. What we should reconstruct, instead, is religious expression and practice. If we take the broad view that religious language is a natural expression of awe in the face of a grand universe, a sense of deep piety and gratitude, then it most certainly is not simple narcissism. The justifying reconstruction of religion, then, is to capture the expressive function of holding something beyond yourself sacred and finding it precious. Sometimes this is manifest in creation stories or parables of good deeds, but the truth or falsity of those stories is not what’s to be communicated, but rather the endorsement of a form of life committed to the values those stories exemplify. Thus, an expressivist account of religion. Again, I do not think this falls into a form of narcissism, and I believe it captures the deep and defensible core to religious life. It’s not only what’s salvageable, but it’s also what’s worth saving. Whether this view admits of the contradiction problem is something worth considering at length. However, I cannot do that here. One thing to say, though, is that given the core commitments of piety and gratitude, we do have the means for teasing apart and arbitrating differences. So I do not see an expressivist reading of religious expression on the road to insipid pluralism. If it is, that’s a reason to reject it. A final concern is whether expressivism here has the same sharpshooter’s fallacy as James’s reconstructive strategy. Probably so, as I take it that what passes the test of justification itself is what’s valuable in religion. The test of justification reveals what’s valuable. To dialectical opponents, for sure, this looks like painting bullseyes. Fair enough, I reply. What do you expect when an anti-theist tries to sympathetically reconstruct religion?

Question 4: What about the power of positive thinking? What’s remarkable about “The Will to Believe” is that James makes the case for the thought that our beliefs, even our overbeliefs, can make a positive difference. What we believe, even as individuals, matters. For sure, no sports coach ever gave a pep talk that ran that the team stinks, the players are talentless, and that they are overmatched before the big game. Or, at

The Ethics of Belief and Philosophy of Religion

193

least, no coach should ever give that pep talk. It’s not a pep talk. Acknowledging that we have many doxastically efficacious commitments is important, especially given that negative thoughts can yield their own verification. However, the circumstances are not forced. One needn’t believe one way or the other to do one’s job or perform difficult tasks. And what makes the likelihood of success cut in one’s favor more consistently is not belief that one will succeed, but in practice and attention to where one might fail. One’s energies are better directed, instead of maintaining one’s self-esteem, toward actual improvement. Consider the Dunning–Kreuger effect. Those who do not know do not know that they do not know. Those with little skill do not know they have little skill. The Dunning–Kreuger effect is a familiar enough pattern, named after the Cornell psychologists Justin Kreuger and David Dunning. In their landmark essay (1999), they reported that they found that undergraduates scoring in the lowest quartile in grammatical and logical abilities tests systematically (1) overrated their abilities in these areas, (2) even after receiving feedback that they had not done as well as they had thought, nevertheless considered themselves to have done comparatively well on the tests, and (3) failed to recognize the performances of others who did well as successful (1999). In a follow-up study, Dunning, Kreuger and their coauthors note: “poor performers do not learn from feedback suggesting a need to improve” (Ehrlinger, Johnson, Banner, Dunning, and Kruger 2008). Some other examples: ●

●

●

Medical students receiving the lowest grades (D+) among their peers in clinical clerkships in obstetrics and gynecology overestimated their grades by two full grades, thinking on average they did B+ work (Edwards, et al. 2003). In a 2009 survey of roughly 2500 American citizens, of the group that claimed to have “very good” understanding of American politics only half the respondents could name all three branches of the Federal government, only 54 percent knew that the power to declare war rests with Congress rather than the president, and only 57 percent could properly identify the role played by the electoral college, with many thinking it “supervises television debates” (Intercollegiate Studies Institute 2008). Ninety-four percent of college professors (responding to an AAUP poll) believe they do “above average work” (Cross 1977).

In short, with a vast majority of us, our overbeliefs that we are good at what we do does not make us any better, they just make us wrong and worse. Overbelief ’s

194

Evidentialism and the Will to Believe

a curse. In fact, a double curse. First, it entails the evidence blindness that yields increased likelihood of error, and second, because of the evidence blindness, it yields insulation against correction. This is a serious problem, and when we conjoin these consequences of overbelief with the will-to-believe doctrine, we have a recipe for what I call deepening dogmatism. Roughly, the thought is that if one has very strong practical reasons to be evidentially slothful, then one will not only have the Dunning-Kreuger effect, but one will also cement it. Doubt, given the will-tobelieve doctrine, is the enemy. And so there are thoughts that one must guard oneself against, even against the evidence of their truth. In these cases, it’s not that someone, say, believes in God or themselves, but it’s that they believe in believing in God or themselves.14 Once one believes in belief, no amount of evidence can dislodge the beliefs. This, by the way, is why people are so often upset with themselves when they find they have doubts about the divine. This is why James must cast about and reconstruct religion so it is something that can withstand the crucial scrutiny—he believes in belief. This turns us into loyalty enforcers, those who want to ensure orthodoxy. Now, I think this is why people who believe in God are regularly infrequent readers of the Bible. They value the belief more than the content, and this yields wild and untempered dogmatism.15 The strident belief that the hypothesis is true is a hindrance not only in error avoidance but also in truth pursuit. This is because the will-to-believe provides motivation to resist the abandonment of hypotheses that are false even when the evidence goes against them. These are the costs of being “light hearted” about errors.16 In short, with a vast majority of us, our beliefs that we are good at what we do does not make us better. They just make us wrong. And that’s not the end of it. Being wrong isn’t just bad in itself, but it has consequences.

Notes

Introduction 1

References to James’s “The Will to Believe” will be to the Harvard University Press The Works of William James: The Will to Believe and Other Essays in Popular Philosophy, F. Burkhardt, F. Bowers, and I.K. Skrupskelis Eds. My references to Clifford’s “The Ethics of Belief ” will be to. William Kingdon Clifford’s Lectures and Essays. New York: Macmillan and Co My references to these editions will be in the form (WB p#) for James and (L&E p#) for Clifford.

Chapter 1 1

2 3 4

5 6

7 8

See Kallfelz (2009) for an extension of Clifford Algebra into the question of how physical theories can keep questions of ontology distinct from formal models for objects. Stephen (1875). See for a survey of the cases, Livingston (2007: 122–3). W. B. Carpenter (1876), and Livingston’s discussion (2007: 123). For the originating essays on this issue of the core challenge of luck and moral assessment, see Nagel (1979) and Williams (1981). In Nagel’s terminology, all of Clifford’s cases are those of “resultant luck.” Versions of this defense have been made for the insignificant view of luck that have been made by Jonsen (1984) and Rosebury (1995). See Roy Baumeister (2001) and my own work on Poe’s Law for this priming effect (Aikin 2012). For recent studies on the priming elements on group estimation, especially for false memories of negative things about the outgroup, see Vitelli (2012) and Saletan (2010). See also another version posed by Clark, who argues that given this line of reasoning, “The Cliffordian inconsistently accepts Clifford’s Maxim” (1990: 109). See Price (1956) for a model for indirect voluntarism as a reply.

196

Notes

9 Brian Zamulinski rightly notes that Clifford’s argument depends on the thought that “If one adjusts one’s standards to permit one overbelief, the adjusted standards permit others” (2002: 444). In short, one must adjust the standard not only to acquire the belief, but also to retain it. The rule, then, will apply to further commitments. 10 Vorstenbosch (2000: 105) calls this “Clifford’s holistic theory of belief.” Notice further that this observation overturns Plantinga’s thesis that evidentialism must be a form of foundationalism (1983: 15 and 2000: 89). Clifford is a prime counterexample. 11 See Van Harvey (2008: 42) for the observation that most evidential norms are “role-specific intellectual responsibilities.” Because we have specific practical responsibilities, we have correlated cognitive duties. The question is whether there is a general duty for all of us as believers. 12 Richard Gale calls this the “plague theory” of epistemic negligence (1993: 335). 13 See MaCoun (2002: 124) and Sunstein (2006: 80). 14 Gigone and Hastie (1993) and Sunstein (2006: 83). 15 See Sustein (2002) and Sunstein (2006: 88). 16 See Sunstein (2006), Aikin and Clanton (2010), and Aikin (2012) for the argumentative consequences of group polarization. 17 See Aikin (2012) where I show that the circumstances of entertainment and humor yield wild distortions in views on issues—in essence, if you can make those who hold a view look ridiculous, you’ve made progress in making the view look obviously false. Humorous ad hominem argument is incredibly effective. 18 Susan Haack poses the challenge to Clifford by noting that not all self-deceptions are ones that yield larger, global errors. In fact, it’s not clear that even most do (1997: 137). 19 Vorstenbach is not alone in taking the Duck Test to be the full story with Clifford. (See also: Ammerman 1965: 264, Kauber 1972: 47, Nathanson 1982: 576, Wernham 1987: 60, Fogelin 1994: 115, Haack 1997: 135, Plantinga 2000: 89, Bergeron 2006: 68 and Harvey 2008: 42) 20 See Shaffer (2006: 47), who holds that the duty of non-malfeasance “equally is at home” in deontological and consequentialist ethics. 21 See Hollinger, who takes it that Clifford’s line here, in putting Muslims as spokespersons for religion, is supposed to yield increased scrutiny from the target audience (1997: 74). 22 Later positivists, like Ayer (1952) call this condition a verifiability in principle condition, in contrast with verifiability in practice. Though the later positivists used this as a semantic criterion for meaning, Clifford’s case proceeds with it as a purely epistemic issue. 23 Fogelin holds explicitly that Clifford’s evidentialism instigates the Pyrrhonist (1994: 115). For a brief survey of the replies to the Pyrrhonist’s challenge, see Aikin (2009

Notes

197

and 2011). For accounts of the validity of epistemic infinitism, see Klein (1999) and Aikin (2011). I have argued elsewhere (Aikin 2010c and 2011) that versions of epistemic infinitism are not uncommon to nineteenth century views on the demands of knowledge, and I think it is plausible to attribute this view to Clifford, too. Finally, I argue in my 2011 book, Epistemology and the Regress Problem, that epistemic infinitism is the only account of internalist justification that can be consistently evidentialist. 24 See Brian Zamulinski’s defense of Clifford from the double standard objection in (2004).

Chapter 2 1 See Bronson (1914: 29) for this contrast between the religious cultures of the two universities. 2 I owe this term and the observation here to James Bednar, who noted this in his 2007 dissertation, “Action, Inquiry and Evidence.” Vanderbilt University. See also Richard Gale’s account of belief-helps-make-the-truth (1980, 1999, and 2005). 3 See Suckiel (1996: 34), who takes this as especially important in replying to Clifford. In short, those for whom religious belief needs a defense are those for whom the defense will likely never make a difference. It’s only those who need the defense. 4 James himself later insists on his pluralism by holding that his views are ones that allow “free competition of the various faiths” (WB 8), and others have taken this pluralist line, too. See Dooley (1975: 88), Haack (1997: 137), Oliver (2001: 102), Welchman (2006: 236). Robert Talisse and I have criticized this Jamesian-style pluralism in our 2005 and 2011 works. Again, notice that the Jamesian line here is that we can be pluralists, but only with regard to a very restricted range of alternatives—this will be a target for criticism later, as I call this the meta-monism problem in Section X-7. 5 Robert Adams, for example, calls the virtue of faith an inclination of “greatest trust,” an “acceptance of complete dependence” (1987: 22). 6 I’ve taken the case from what Hare and Madden (1968: 117), Davis (1972: 239), and O’Connell (1997: 62) constructed similar examples. 7 See George Mavrodes’s “James and Clifford on ‘The Will to Believe’ ” for an early version of this argument. Mavrodes’s case is that James could have stopped his argument immediately after introducing the doctrine of forced options, as “James’s view of such options as forced… implies that a suspension of belief is not inadvisable, but rather impossible” (1963: 197). Ought implies can, so evidentialism is false.

198

Notes

8 My view is that those who emphasize Jamesian experimentalism in religious belief make hash out of their case that the belief is momentous. It is a costly and simply obvious error. See, for those who are exemplary in this error, Weintraub (2003: 115) and Welchman (2006: 230). 9 The division of these stages arose in conversation with Jeff Kasser, Micah Hester, and Michael Slater after Micah’s Presidential Address to the William James Society at the 2012 meeting of the American Philosophical Association (Hester Forthcoming). First, Slater’s book (2009) distinguishes the psychological stages and the epistemological stages of James’s argument. Kasser’s substantive view, that James’s meta-epistemic stage of “The Will to Believe” is a form of expressivism, required Jeff to distinguish the psychological stage, which he takes is not a form of expressivism from the normative epistemological stage, which is an application of epistemic rules. So, with Kasser, there are three stages (see his essay with Shah in 2006 and his forthcoming essay in William James Studies). My addition has been to note that a logical stage must precede the psychological. So, with my view, there are four. An early version of this distinction can be found in Nathanson (1982: 573), who sees “The Will to Believe” developing in psychological and axiological stages. 10 On this point, John E. Smith notes that “at the heart of [James’s] view is a consistent voluntarism” (1963: 41). 11 This is a way of handling, for example, Wernham’s “heretical” interpretation that the WTB doctrine is a duty to believe. He is right in observing that the doctrine has two parts: first, establishing the right to believe, and then to show that it’s foolish not to believe. But that it is foolish not to believe only establishes the preferability of belief, not the obligation. 12 See, for versions of this distinction, Davis (1972) and Johansen (1975). William Gavin’s interpretation, by contrast, takes undecidability to be a function of vagueness (1992). I cannot address Gavin’s line of argument with any detail here, but note that were this the case, then James’s case would not address Clifford’s in the slightest. 13 See Kauber (1972 and 1974) for this arc of the dialectic. 14 See Henry Jackman’s (1999) case that James is an externalist on these grounds. And see Ernest Sosa’s defense of meta-knowledge for externalists (2003). 15 This view that pragmatism is a form of skepticism was exceedingly well articulated by Popkin (1964), who noted that Carneadean skeptical models for credence have modern analogues in the pragmatists. A version of this skeptical arc of pragmatist epistemology is defended in Talisse and Aikin (2008). 16 See Jon Kvanvig’s (2011: 53) claim that “it’s close to a truism that there is both evidence and what we make of it.” Cf. Gage (2014). 17 See Ellen Kappy Suckiel’s application of this view about truth: “[A] belief is justified—and hence pragmatically true—if it successfully fulfills its function;

Notes

18

19 20 21

22 23 24 25

26

199

that is to say, if it enables the believer to account of and predict his experience in ways that most satisfactorily fulfills his ends” (1982: 10). Gail Kennedy notes that the Jamesian line here requires that if there are not absolute truths prior to our believings, our beliefs, when experiences support them, become true (1958: 581). My colleague Jose Medina holds that this consequence is favorable, as “James’s pluralistic conception of truth makes room for a robust notion of objectivity, which is at the same time a notion of justice” (2010: 126). I will challenge both thoughts on pluralism. See Talisse and Aikin (2005 and 2011) for criticisms of the pluralist program in pragmatism. See Aikin 2010a for a defense of the thought that knowledge must be distinct from true belief. As George Mavrodes rightly notes, one can maintain this only if one can remain “lighthearted in thinking about errors” (1963: 195). Others noting the passional background for this point are Nathanson (1982: 573), Clark (1990: 109), and Axtell (2001: 327). There is evidence for both of these thoughts. See Gary Bartlett (2010) for an overview of the problem of escalation in spanking. This exaggeration of the requirements for evidentialism is corrected and noted by Wernham (1987: 72–73), Hollinger (1997: 70–71), and Zamulinski (2002: 449). Ammerman, too, worries that the “she loves me” thought can carry one too far (1965: 266). See Thomas Gabor’s Everybody Does It! (1994) for wide research showing that knowledge of cheating on taxes, tests, or even simple economic exchanges increases likelihood of cheating. Recently Rettinger and Kramer (2009) have shown correlation between knowledge of cheating and oneself cheating. Hence, “cheating is contagious.” This knowledge contributes to what is called “selective moral disengagement,” wherein one does what one knows what one ought not do, but for reasons distinctive for the situation. See also Bandura (2002) for the variety of forms of moral disengagement, not all but some dependent on knowledge of performances of others. Ruth Weintraub notes in her “Non-Fideistic Reading” of James’s “Will to Believe” that the friendship cases and all those with doxastic efficacy are “not grist for the fideist’s mill” because they causally contribute to their own truth and so undo the skeptical background conditions for faith (2003: 105). Richard Gale, too, says that the belief helps create the fact cases aren’t counterexamples to Clifford’s doctrine, because if A knows that his belief will help create the truth, A knows when he has the belief that the truth is more likely (1980: 3). Smith (1963: 75) and Nathanson (1982: 576) also note that James’s examples are more in the service of a broader notion of epistemic rationality.

200

Notes

27 See Augustine’s Tractate on St. John’s Gospel 29.6, and Anselm’s Proslogion I 28 See Richard Rorty’s observation that pragmatists, in thinking about religious commitment, “fluctuate” between “God” meaning the culmination of possible human futures and the external guarantor of those futures (1997). 29 I’ve defended this view in my essay “The Problem of Worship” (2010b) and in my book with Robert Talisse, Reasonable Atheism (2011). 30 See Feldman (2006: 26), who notes that James’s argument from “the good” of having the hypothesis true could be run just as well for atheism. Note that my argument needn’t depend on the matter indexing a cognitive good of truth, but also other practical and ethical goods. 31 Versions of this challenge to James’s argument and to defenses of it have generally been in the form that James equates action and belief. Bertrand Russell was one of the first and most influential critics along these lines (1996: 84). See also Miller (1942: 547) and Wernham (1987: 64). 32 See Dooley (1975: 89), who notes that the case of the will-to-believe is the coordination of the Truth Norm with doxastic efficacy. Suckiel develops this thought by pointing out that given God’s personality, “He is more likely to reveal himself to individuals who already show some measure of religious belief ” (1982: 83). William Dean holds, too, that Given James’s view, God must be constructed and thereby must evolve with the needs of the believers who bear this relation to him (2004: 332). 33 See Schellenberg (2006) for a fully developed line of this argument. 34 Allen Wood (2008: 14) develops the thought that with the WTB, one shows not only a lack of respect for oneself but also for the objects of one’s belief. Paul Moser (2010 and 2008) has pursued a version of the “cognitive idolatry” argument against a broad class of views, and I think it should be applied, too, to the will-to-believe. 35 Consequently, Reck is off target when he calls James “the American Kierkegaard” (1967: 83). 36 With Robert Talisse, I’ve given an argument that pragmatism cannot consistently maintain a full bore pluralism. Further, pragmatic reconstruction, especially in the form of deflation, cannot be a form of toleration (2005 and 2011). For a version of the meta-monism worry in discussions of pluralism in philosophy of religion— that is, maintaining a pluralism on the basis of thinking that all religions express a singular truth—see Aikin and Aleksander (2013).

Chapter 3 1 This has occasioned some commentators to say that Clifford has confused two notions of “ought.” See, for example, Haack, who says “Clifford never distinguishes ‘epistemically wrong’ from ‘morally wrong’” (1997: 135). See also Feldman (2006).

Notes

201

2 Versions of cases like these abound. See, for example, Suckiel (1982: 76), Jack Meiland (1980), John Heil (1983), and Marvin Kohl (1990). 3 Versions of this CT-style belief-ruination has been given by Price (1956), Adler (2002), and Shah (2006 and 2008). 4 Both Shah (2008: 498) and Booth (2008: 230) commenting on Shah, make the distinction between the question whether to believe p and whether to bring about the belief that p. Evidentialism should, I think, bear on both questions. Surely it seems that were we to use purely practical reasons to answer the second question, it would be ruinious of our belief for the same reason CT thoughts are. 5 A classic case for this pragmatic argument for evidentialism can be found in Price (1956: 5). 6 Richard Rorty pauses to consider this case for evidentialism when he notes that the wrongness of overbelief is equivalent to the wrongness of pretending to participate in a common project while refusing to play by the rules: “To be rational is to submit one’s beliefs—all one’s beliefs—to the judgment of one’s peers” (1997: 6). See also Robert Brandom’s account of “rationalistic pragmatism” (1994 and 2000). 7 David Hollinger notes that Clifford and James share a great deal of “intellectual ground” (1997: 70). For an example of what I will call “big tent pragmatism,” see Margolis (2010), and for contemporary versions of the pragmatist appropriation of evidentialism, see Misak (2000) and Talisse (2009). 8 The point is that once we’ve made the CT-style reasons explicit, we can see that these beliefs are defective if held on the basis of nonevidential reasons. We may not endorse these beliefs but we can nevertheless forgive those who hold them. See also Aikin (2006b), Shah (2008), and Booth (2008) for versions of this thought. 9 In this respect, Feldman’s evidentialism bears a resemblance to unexceptionable evidentialism, in that the rule is not ever capable of being overridden. But this is not because Feldman holds that the rule is categorical (as Clifford and Locke held). Instead, it is because moral considerations do not bear on cognitive rules, and vice versa. Feldman seems open, at least in principle, to the evidentialist requirement being a prima facie rule capable of being trumped by other cognitive rules, but it is unclear from the discussion what rules would override the requirement. 10 See Kitcher 2001. I believe Kitcher’s argument, however, is mistaken, as Kitcher’s use of the utilitarian considerations of the consequences of race-related research improperly weighs the consequences researchers taking his advice. The point here is not to show that Kitcher is right, but to show that there are intelligible relations of overriding between ethical and cognitive norms. That Kitcher is wrong is one thing, that Kitcher is speaking nonsense is another. See Talisse and Aikin (2007). 11 Rorty rightly notes that “we don’t mock a mother who believes her sociopathic child’s essential goodness even when it is visible to no one else” (1997: 12). See also Amesbury (2008: 31).

202

Notes

12 Explicitly humanist Jamesian, Patrick Dooley, states the reconstruction in stark terms: “Religious belief is rational since it fulfills our theoretic demands by answering the ultimate questions and makes relevant our practical life…. Further, the moral actions religious belief stimulates help produce internal belief in God by making the world more moral and thereby increasing our sense of the presence and purpose of God” (1972: 146). John E. Smith, similarly, states the reconstructive program: “James sought to translate religious ideas such as God, conversion, salvation, guilt, into a ‘cash value’ of experiences” (1963: 77). 13 As James-commentator Phil Oliver rightly notes, “transcendence need not imply the supernatural” (2001: 2). 14 See Daniel Dennett’s account of belief-in-belief (2006: 203). Dennett holds that this explains why religious zeal is no predictor of religious literacy. Further, he holds that once one believes in belief, one will have “contemplation effects” associated with doubts. Just as one sees another as blameworthy for even considering a bad action (but nevertheless not doing it), one will see others as blameworthy for their doubts. 15 Bertrand Russell, in conjunction with the fact that the will-to-believe doctrine yields inconsistent outcomes but encourages strenuous commitment, holds that the consequence is fierce dogmatism. He holds that it is “very questionable wisdom… accepting (the will-to-believe) as a defense of orthodoxy.” Complete belief “will justify persecution” (1996: 83). See also Nathanson (1982: 576) for what he calls the “snowball effect” with overbelief, and Hick (1963: 60), who worries about the practitioners of the WTB “becoming more firmly entrenched in their current prejudices.” 16 In this respect, I agree with Miller’s (1942: 553) and Zamulinski’s (2004: 79) view that contrary to the widely held opinion that the Cliffordian resistance to James and overbelief is too strong, rather, given the consequences of the alternatives, it is too weak.

Bibliography

Adams, Robert M. (1987) The Virtue of Faith. In The Virtue of Faith. New York: Oxford University Press. 9–21. Adler, Jonathan (2002) Belief ’s Own Ethics. Boston: MIT Press. Aikin, Scott (2006a) Contrastive Self-Attribution of Belief. Social Epistemology. 20. 93–103. ——— (2006b) Modest Evidentialism. International Philosophical Quarterly. 46. 327–344. ——— (2007) Evidentialism for Everyone. Think. 15. 37–44. ——— (2008) Evidentialism and James’s Argument from Friendship. Southwest Philosophy Review. 2008. 24. 1–14. ——— (2009) Don’t Fear the Regress: Epistemic Infinitism and Cognitive Value. Think. 23. 55–61. ——— (2010a) Knowledge Is Merely True Belief, Rebutted. European Journal of Analytic Philosophy. 2010. 6. 7–15. ——— (2010b) The Problem of Worship. Think. 25. 101–113. ——— (2010c) The Problem of the Criterion and a Hegelian Model for Epistemic Infinitism. History of Philosophy Quarterly. 27. 248–255. ——— (2011) Epistemology and the Regress Problem. Routledge. ——— (2012) Poe’s Law, Group Polarization, and Argumentative Failure in Religious and Political Discourse. Social Semiotics. 22. 400–410. ——— and Jason Aleksander (2013) “Religious Pluralism, Exclusivism and Third Options: The Case of Nicholas of Cusa’s De Pace Fidei.” International Journal of Philosophy of Religion, 74. 219–235. ——— and Caleb Clanton (2010) Developing Group Deliberative Virtues. Journal of Applied Philosophy. 27. 409–424. ——— and Robert B Talisse (2011) Reasonable Atheism. Amherst, NY: Prometheus Books. Amesbury, Richard (2008) The Virtues of Belief: Toward a Non-Evidentialist Ethics of Belief Formation. International Journal for Philosophy of Religion. 63. 25–37. Ammerman, Robert (1965) Ethics and Belief. Proceedings of the Aristotelian Society. 65. 257–266.

204

Bibliography

Aurelius, Marcus (2003) Meditations. Trans. Gergory Hays. New York: The Modern Library. Axtell, Guy (2001) Teaching William James’s “The Will to Believe”. Teaching Philosophy. 24. 325–345. Ayer, A.J. (1952) Language, Truth, and Logic. New York: Dover. Bandura, Albert (2002) Selective Moral Disengagement in the Exercise of Moral Authority. Journal of Moral Education. 31. 2. 101–119. Bartlett, Gary (2010) An Argument Against Spanking. Public Affairs Quarterly. 24. 65–77. Baumeister, Roy (2001) Evil: Inside Human Violence and Cruelty. New York: Barnes and Noble Books. Bergeron, Melissa (2006) The Ethics of Belief: Conservative Belief Management. Social Epistemology. 20. 67–77. Booth, Anthony R. (2008) A New Argument for Pragmatism?. Philosophia. 36. 227–231. Braithwaite, R.B. (1933) The Nature of Believing. Proceedings of the Aristotelian Society. 33. 129–146. Braithwaite, Valerie and Michael Wenzel (2008) Integrating Explanations of Tax Evasion and Avoidance. In The Cambridge Handbook of Economic Behavior. Ed. Alan Lewis. Cambridge: Cambridge University Press. 304–333. Brandom, Robert (1994) Making It Explicit. Cambridge: Harvard University Press. ——— (2000) Articulating Reasons. Cambridge: Harvard University Press. Bronson, Walter (1914) The History of Brown University: 1764–1914. Providence: Brown University Press. Brown, Alan Willard (1947) The Metaphysical Society. New York: Columbia University Press. Carpenter, W.B. (1876) On the Fallacies of Testimony in Religion to the Supernatural. Contemporary Review. 27 (January). In Livingston. 2007. Cicero, Marcus Tullius (1913) De Officiis. Trans Walter Miller. Cambridge: Harvard University Press. ——— (1942) Paradoxa Stoicorum. Trans H. Rackham. Cambridge: Harvard University Press. Clark, Kelly James (1990) Return to Reason. Grand Rapids: William B. Eerdmans Publishing. Cross, P. (1977) Not can but will College Teaching be Improved?. New Directions for Higher Education. 17. 1–15. Davis, Stephen (1972) Wishful Thinking and “The Will to Believe.” Transactions of the Charles S. Peirce Society. 8. 231–245. Dean, William (2004) Religion. In The Blackwell Guide to American Philosophy. EdA. Marsoobian and J.Ryder. Malden, MA: Blackwell. 325–342. Dennett, Daniel C. (2006) Breaking the Spell. New York: Penguin Books. Descartes, René (1996) Meditations of First Philosophy. Trans. Cottingham. New York: Cambridge University Press.

Bibliography

205

Dewey, John (1934) A Common Faith. New Haven: Yale University Press. Dooley, Patrick (1972) The Nature of Belief: The Proper Context for James’s “The Will to Believe” Transactions of the Charles S. Peirce Society. 8. 141–151. ——— (1975) Pragmatism as Humanism. Chicago: Nelson-Hall. Edwards, R.K., K.R Kellner, C.L Sistrom and E.J Magyari (2008) Medical Student Selfassessment of Performance on an Obstetrics and Gynecology Clerkship. American Journal of Obstetrics and Gynecology. 188. 1078–1082. Ehrlinger, Joyce, Kerri Johnson, Matthew Banner, David Dunning and Justin Kruger (2008) Why the Unskilled are Unaware. Organizational Behaviour and Human Decision Processes. 105. 98–121. Epictetus (1928) The Discourses. Trans. W. A. Oldfather. Cambridge: Harvard University Press. ——— (1973) The Enchiridion. Trans. Nicholas White. Indianapolis: Hackett. Feldman, Richard (2000) The Ethics of Belief. Philosophy and Phenomenological Research. 60. 235–256. ——— (2006) Clifford’s Principle and James’s Options. Social Epistemology. 20. 19–33. Fisch, Max (1986) Peirce, Semeiotic, and Pragmatism. Bloomington: Indiana University Press. Fogelin, Robert (1994) Pyrrhonian Reflections on Knowledge and Justification. New York: Oxford University Press. Gabor, Thomas (1994) Everybody Does It!. Toronto: University of Toronto Press. Gage, Logan (2014) Evidence and What We Make of It. Southwest Philosophy Review. 30. 20–25. Gale, Richard (1980) William James and the Ethics of Belief. American Philosophical Quarterly. 17. 1–14. ——— (1993) On the Nature and Existence of God. New York: Cambridge University Press. ——— (1999) The Divided Self of William James. Cambridge: Cambridge University Press. Gavin, William J. (1992) William James and the Reinstatement of the Vague. Philadelphia: Temple University Press. Gigone, David and Reid Hastie (1993) The Common Knowledge Effect. Journal of Personality and Social Psychology. 65. 971–973. Haack, Susan (1997) The Ethics of Belief, Reconsidered. In The Philosophy of Roderick M. Chisholm. Ed. Lewish Hahn. Lasalle, IL: Open Court. 129–144. Hare, Peter and Edward H. Madden (1968) William James, Dickinson Miller, and C.J. Ducasse on the Ethics of Belief. Transactions of the Charles S. Peirce Society. 4. 115–129. Harvey, Van (2008) The Ethics of Belief and Two Conceptions of Religious Faith. International Journal for Philosophy of Religion. 63. 39–54. Heil, John (1983) Believing What One Ought. The Journal of Philosophy. 80. 752–765. Hester, Micah (Forthcoming) Presidential Address to the 2012 Meeting of the William James Society. William James Studies. Hick, John (1963) Philosophy of Religion. Englewood Cliffs, NJ: Prentice Hall.

206

Bibliography

Hodges, Michael (1990) Transcendence and Wittgenstein’s. Tractatus. Philadelphia: Temple University Press. Hodges, Michael and Scott Aikin (2006) Wittgenstein, Dewey, and the Possibility of Religion. Journal of Speculative Philosophy. 20. 1–19. ——— (Forthcoming) St. Anselm’s Ontological Argument as Expressive. Philosophical Investigations. Hollinger, David (1997) James, Clifford, and the Scientific Conscience. In The Cambridge Companion to William James. Ed. Ruth Anna Putnam. New York: Cambridge University Press. 69–83. Hume, David (1993[1740]) An Enquiry Concerning Human Understanding. Indianapolis: Hackett. Intercollegiate Studies Institute (2008) The Diverging Influences of the College Degree & Civic Learning on American Belief. Wilmington, DE: Intercollegiate Studies Institute. Jackman, Henry (1999) Prudential Arguments, Naturalized Epistemology, and the Will to Believe. Transactions of the Charles S. Peirce Society. 34. 1–37. Johansen, Arnold (1975) “The Will to Believe” and the Ethics of Belief. Transactions of the Charles S. Peirce Society. 11. 111–127. Jonsen, H. (1984) Morality and Luck. Philosophy. 59. 323–330. Jordan, Jeff (1996) Pragmatic Arguments and Belief. American Philosophical Quarterly. 33. 409–420. Kallfelz, William M. (2009) Clifford Algebra: A Case for Geometric and Ontological Unification. Sarrbruecken: VDM Verlagsservicegesellshaft. Kasser, Jeff (Forthcoming) Prospects for a Jamesian Expressivism. William James Studies. ——— and Nishi Shah (2006) The Metaethics of Belief: An Expressivist Reading of “The Will to Believe” Social Epistemology. 20. 1–18. Kauber, Peter (1972) William James’s Ethics of Belief. In Dissertation at State University of New York. Buffalo. Ann Arbor: University of Michigan Press. ——— (1974) Does James’s Ethics of Belief Rest on a Mistake?. The Southern Journal of Philosophy. 12. 201–214. Kennedy, Gail (1958) Pragmatism, Pragmaticism, and the Will to Believe – A Reconsideration. The Journal of Philosophy. 55. 578–588. Kitcher, Philip (2001) Science, Truth and Democracy. Cambridge: Cambridge University Press. Klein, Peter (1999) Human Knowledge and the Infinite Regress of Reasons. Philosophical Perspectives. 13. 297–325. Kohl, Marvin (1990) Skepticism and Happiness. Free Inquiry. 10. 41–43. Kruger, Justin and David Dunning (1999) Unskilled and Unaware of it. Journal of Personality and Social Psychology. 77. 1121–1134. Kvanvig, Jon (2011) The Rational Significance of Reflective Ascent. In Evidentialism and Its Discontents. Ed. Trent Doughtery. New York: Oxford University Press.

Bibliography

207

Lightman, Bernard (1987) The Origins of Agnosticism: Victorian Unbelief and the Limits of Knowledge. Baltimore: John Hopkins University Press. Livingston, James (2007) Religious Thought in the Victorian Age: Challenges and Reconceptions. Continuum: London. Locke, John (1996 [1690]) An Essay on Human Understanding. Indianapolis: Hackett Publishing. Macoun, Robert J. (2002) Comparing Micro and Macro Rationality. In Judgments, Desires, and Public Policy. Ed. R. Gowda and J.Fox. Cambridge: Cambridge University Press. 121–126. Madigan, Timothy (1999) Introduction. In The Ethics of Belief and Other Essays. Ed. T. Madigan. Amherst: Prometheus Books. ix–xxii. ——— (2010) W.K. Clifford and “The Ethics of Belief.” Newcastle on Tyne: Cambridge Scholars Press. Margolis, Joseph (2010) Pragmatism’s Advantage. Palo Alto: Stanford University Press. Mavrodes, George (1963) James and Clifford on “The Will to Believe”. The Personalist. 44. 191–198. Medina, José (2010) James on Truth and Solidarity: The Epistemology of Diversity and the Politics of Specificity. In 100 Years of Pragmatism: William James’s Revolutionary Philosophy. Ed. John Stuhr. Bloomington, IN: Indiana University Press. Meiland, Jack (1980) What Ought We Believe? Or: The Ethics of Belief, Revisited. American Philosophical Quarterly. 17. 15–24. Miller, Dickinson (1942) James’s Doctrine of “The Right to Believe”. The Philosophical Review. 51. 541–558. Misak, Cheryl (2000) Truth, Politics, Morality: Pragmatism and Deliberation. New York: Routledge. Mitova, Veli (2008) Why W.K. Clifford Was a Closet Pragmatist. Philosophical Papers. 37. 471–489. Moser, Paul (2008) The Elusive God. New York: Cambridge University Press. ——— (2010) The Evidence for God. New York: Cambridge University Press. Nagel, Thomas (1979) Moral Luck. In Mortal Questions. New York: Cambridge University Press. 24–38. Nathanson, Stephen (1982) Nonevidential Reasons for Belief: A Jamesian View. Philosophy and Phenomenological Research. 42. 572–580. O’Connell, Robert J. (1997) William James and the Courage to Believe. New York: Fordham University Press. Oliver, Phil (2001) William James’s “Springs of Delight.” Nashville: Vanderbilt University Press. Peirce, Charles Sanders (1960) How to Make Our Ideas Clear. In The Collected Papers of Charles Sanders Peirce. Volume 5. Ed. Charles Hartshorne and Paul Weiss. Cambridge: Harvard University Press. Perry, Ralph Barton (1935) The Thought and Character of William James. 2 Volumes. Boston: Little, Brown, and Co.

208

Bibliography

Phillips, D.Z. (2001) Religion and the Hermeneutics of Contemplation. New York: Cambridge University Press. Plantinga, Alvin (1967) God and Other Minds. Ithaca: Cornell University Press. ——— (1983) Reason and Belief in God. In Faith and Rationality. Ed. A. Plantinga and N. Wolterstorff. Notre Dame: Notre Dame University Press. 16–92. ——— (2000) Warranted Christian Belief. Oxford: Oxford University Press. Popkin, Richard (1964) The History of Scepticism from Erasmus to Descartes. New York: Harper and Row. Price, H.H. (1956) Belief and Will. Proceedings of the Aristotelian Society. 65. 257–266. Reck, Andrew (1967) Introduction to William James. Bloomington: Indiana University Press. Reid, Tomas (1983) Inquiry and Essays. Ed. R. Beanblossom and K. Leherer. Indianapolis: Hackett. Rettinger, David and Yair Kramer (2009) Situational and Personal Causes of Student Cheating. Research in Higher Education. 50. 3. 293–313. Rorty, Richard (1997) Religious Faith, Intellectual Responsibility, and Romance. In Pragmatism, Neo-Pragmatism, and Religion. Ed. Charley Hardwick and D. Crosby. New York: Peter Lang. 3–21. Also published in The Cambridge Companion to William James. Ed. Ruth-Anna Putnam. Cambridge University Press. 84–102. Rosebury, B. (1995) Moral Responsibility and Moral Luck. Philosophical Review. 104. 499–524. Russell, Bertrand (1996[1909]) Pragmatism. In Philosophical Essays. New York: Simon and Shuster. 79–111. Saletan, William (2010) The Future of False Memories. Slate Magazine. June 4. Santayana, George (1958) Character and Opinion in the United States. New York: Double Day Anchor Books. Schellenberg, J.L. (2006) Divine Hiddenness and Human Reason. Ithaca, NY: Cornell University Press. Shaffer, Michael J. (2006) The Publicity of Belief, Epistemic Wrongs, and Moral Wrongs. Social Epistemology. 20. 41–53. Shah, Nishi (2006) How Truth Governs Belief. The Philosophical Review. 112. 447–482. ——— (2008) A New Argument for Evidentialism. The Philosophical Quarterly. 56. 481–498. Slater, Michael R. (2009) William James on Ethics and Faith. New York: Cambridge University Press. Smith, John E. (1963) The Spirit of American Philosophy. New York: Oxford University Press. Sosa, Ernest (2003) Philosophical Skepticism and Epistemic Circularity. In Human Knowledge. 3rd Edition. Ed. P. Moser and A. VanderNat. Oxford: Oxford University Press. Stephen, James Fitzjames (1875) “Remarks on the Proof of Miracles.” Paper presented before the Metaphysical Society, November 9 (In Livingston) 2007. Stephen, Leslie (1895) The Life of Sir James Fitzjames Stephen. London: Smith, Elder, and Co.

Bibliography

209

Suckiel, Ellen Kappy (1982) The Pragmatic Philosophy of William James. Notre Dame: University of Notre Dame Press. ——— (1996) Heaven’s Champion: William James’s Philosophy of Religion. Notre Dame, IN: University of Notre Dame Press. Sustein, Cass (2002) The Law of Group Polarization. The Journal of Political Philosophy. 10. 175–195. ——— (2006) Infotopia. Oxford: Oxford University Press. Talisse, Robert B. (2009) Democracy and Moral Conflict. New York: Cambridge University Press. Talisse, Robert B. and Scott Aikin (2005) Why Pragmatists Cannot Be Pluralists. Transactions of the Charles S. Peirce Society. 41. 101–118. ——— (2007) Kitcher on the Ethics of Inquiry. Journal of Social Philosophy. 38. 654–665. ——— (2008) Pragmatism: A Guide for the Perplexed. London: Continuum. ——— (2011) Three Problems for Jamesian Ethics. William James Studies. 6. 1–6. van Inwagen, Peter (1996) It Is Wrong Everywhere, Always, and for Anyone to Believe Anything Upon Insufficient Evidence. In Faith, Freedom, and Rationality. Ed. J. Jordan and D. Howard-Snyder. Notre Dame: Notre Dame University Press. Vitelli, Romeo (2012) Implanting False Memories. Psychology Today. November 4. Vorstenbosch, Jan (2000) W.K. Clifford’s Ethics of Belief Revisited. In Belief, Cognition, and the Will. Ed. Anthonie Meijers. Tilburg: Tilburg University Press. 99–111. Ward, Wilfrid (1912) The Life of John Henry, Cardinal Newman. Vol. 2. London: Longmans. Welchman, Jennifer (2006) William James, “The Will to Believe,” and the Ethics of SelfExperimentation. Transactions of the Charles S. Peirce Society. 42. 229–241. Weintraub, Ruth (2003) A Non-Fideistic Reading of William James’s “The Will to Believe.” History of Philosophy Quarterly. 20. 103–121. Wernham, James (1987) James’s Will-to-Believe Doctrine: A Heretical View. Montreal: McGill-Queens University Press. West, Cornel (1989) The American Evasion of Philosophy. Madison: University of Wisconsin Press. Williams, Bernard (1981) Moral Luck. In Moral Luck. New York: Cambridge University Press. 20–39. Wood, Allen (2008) The Duty to Believe According to the Evidence. International Journal of Philosophy of Religion. 63. 7–24. Zamulinski, Brian (2002) A Re-evaluation of Clifford and His Critics. The Southern Journal of Philosophy. 60. 437–457. ——— (2004) A Defense of the Ethics of Belief. Philo. 7. 79–96. Zenger, T. R. (1992) Why Do Employers Only Reward Extreme Performance? Examining the Relationships Among Performance, Pay, and Turnover. Administrative Science Quarterly. 37. 198–219.

Index Absolutism, epistemic 114–121, 122, 142 moral 42 Adams, Robert 197 Adler, Jonathan 49, 181, 201 Agassiz, Louis 79 Agnosticism 10, 80, 82, 85, 89, 97, 126, 170, 190 Aikin, Scott 171–2, 195, 198–9, 200, 201 Aleksanser, Jason 200 Alpine Climber Case 149–54, 158–59, 170–72 Amesbury, Richard 201 Ammerman, Robert 196, 199 Anselm, St. 159–60, 200 Apraxia 51–3, 176 Aquinas, St. Thomas 10 Arctic Ice Case 61–3 Aristotle 29, 119 Arizona 15 Ascending Case 155–60, 163–4, 172 Assumption of the Uniformity of Nature (AUN) 7, 69–78 Dilemma challenge 69–70 Non-optionality 73–4 Practical justification 73–6 Atheism 80, 89, 97, 169–170, 200 Attribution norm 59–64, 105 Augustine, St. 117, 159, 200 Australian Aboriginal Case, the 65–6 Availability heuristics 39–41, 57 Axtell, Guy 199 Ayer, A. J. 196 Bad Luck Barry 21 Balfour, Arthur James 10 Bartlett, Gary 199 Baumeister, Roy 195 Bednar, James 197 Ben the Sausage Eater 37–8, 120–1 Benificence, instinct of 67–9 Intellectual conceptions of 67–9 Bergeron, Melissa 196 Booth, Anthony R. 201

Brandom, Robert 201 Bronson, Walter 197 Brown University 82–3 Burnt child fears the fire case 69–71 Cambridge University 9 Carpenter, W.B. 11–12, 195 Celestial visitor 57–60 Certainty 80, 90, 114, 117, 119, 121, 124, 163 Chemistry case 61–2 Church, R.W. 11 Cicero 45–6 Clanton, Caleb 196 Clark, Kelly James 195, 199 Cognitive culture 64–8 Coherentism, epistemic 35, 68–9 Coleridge, Samuel Taylor 47, 50 Common knowledge effect 40 Confirmation bias 57 Conflicted thought norm (CT) 182–4 Consequentialism 42–7 Conversion fallacy (for James) ix, 7, 52–4, 90–1, 94–6, 100, 108, 129, 143, 148, 150, 169–173, 176 Cosmopolitanism 46 Credulity, Principle of 55–6 Darwin, Charles 10, 79 Davis, Stephen 198 Dean, William 200 Deepening dogmatism problem 97–8, 179–80, 194 Dennett, Daniel 202 Deontology 42–7 Descartes, Rene 48, 49, 100, 117–8, 138 Dewey, John 191 Dooley, Patrick 197, 200, 201 Double standards objection (to Clifford) 76–7, 107, 197 Doxastic efficacy 7, 25, 84–5, 105–6, 127, 137–8, 140–154, 156–160, 162–3, 166, 171–5, 193–4

Index

211

Duck Test 42–4, 196 Dunning-Kreuger Effect 193–4 Dupery 126, 193–4

Google 33 Group Polarization 39–41, 196 Guest, Mr. 82–3

Egoism 139 Empiricism 114–123 Enlightenment 65 Epictetus 20–22, 45–6 Error amplification 39–41 Error Norm 123–5, 126, 133, 169, 176 Euclid 119 Evidentialist norm 48–9 Ethicalist/practicalist 49, 181–3 Cognitivist 49, 181–4 Clifford’s Evidentialist Norm (CEN) 48–9, 99, 102–3 Integrated Evidentialist Rule (IER) 49, 50, 53, 54, 68, 78–9, 85, 90, 99 Expressivism 130–1, 180, 191–2, 198 Externalism 117, 198

Haack, Susan 196, 197, 200 Hare, Peter 197 Harvard University 79, 82–3, 85 Harvey, Van 196 Hegel, G.W.F. 119, 184 Heil, John 201 Hester, Micah 198 Hick, John 202 Hiddenness, divine 177 History case 71–2 Hodges, Michael 191 Holism, epistemic 34–5, 67–9 Hollinger, David 161, 196, 199, 201 Holmes, Oliver W., Jr. 80 Hume, David 12, 48, 51–2, 54–5, 181 Hutton, R. H. 82 Huxley, Thomas 10, 11–2, 54 Hypotheses 87 Live 87–91, 98, 132, 142, 163–7, 177

Faith 5–6, 10, 12, 13, 52, 76, 81–5, 89–90, 96–8, 101–3, 108–111, 121, 126, 130–1, 154–9, 164–5, 169–70, 174–5, 178, 199 Bringing its own verification 136–54 In someone else’s faith 104–6 Faithful husband case 182 Feldman, Richard 49, 181, 186–9, 200, 201 Fisk, John 80 Fogelin, Robert 196 Foundationalism 68, 196 Freethinkers 82–3, 85 Friendship, argument from 140–5, 150, 152–3, 156, 158, 168, 170, 173 Principle of Niceness (PN) 141–3, 156 Principle of Assumption (PA) 141–3 Axiom of Aloofness (AA) 141–3,173 Axiom of Suspension (AS) 141–5 Fungibility 188 Gabor, Thomas 199 Gale, Richard 110, 196, 197, 199 Gavin, William 198 Generalization problem (for Clifford) 32–47 Generosity 16–17 Gladstone, William 10 Goodman, Nelson 52

Illicit contraries fallacy (for Clifford) ix, 7, 73–4 Illicit conversion See Conversion fallacy Imam, Twelfth 88 Immoralism 51–2 Incommensurability 186–90 Incomparability 186–90 Inconsistency problem 121–3, 174–5, 192 Induction 51, 70–72 Infinitism, epistemic 69, 197 Insolence, epistemic 25, 33, 41, 49 Insurance 17–18, 32 Interested Inquiry General 134–5 Specific 134–6 Interpretation, internal 3 Island Case 22–27, 32, 128, 168 Jackman, Henry 198 Johansen, Arnold 198 Jonsen, H. 195 Kallfelz, William M. 195 Kant, Immanuel 65 Karla the Cowpokes Fan 33–35, 120–1

212 Kasser, Jeff 130–1, 198 Kauber, Peter 196, 198 Kennedy, Gail 199 Kierkegaard, Soren 178, 200 Kitcher, Philip 189, 201 KK thesis 115–8 Knowledge requirement 54–64, 78–9, 105 Knowles, James 11 Kohl, Marvin 201 Kvanvig, Jon 198 Lawfulness requirement 84–5, 99, 104, 154–5, 159, 175 Levels objection 132–3 Light-heartedness 126–7 Lincoln, Abraham 100 Livingston, James 195 Locke, John 48 Lucky Luke 21 Madden, Edward H. 197 Madigan, Timothy vii, 8, 45, 77 Mahdi 88–9, 102, 179 Manning, Henry Edward 11 Marcus Aurelius 46 Margolis, Joseph 201 Mavrodes, George 197, 199 Medicine Man case, the 65–6 Medina, José 199 Meiland, Jack 201 Meliorism 139–40 Memorial evidence 70–1 Metaphysical Club 80–81 Metaphysical Society 10–11, 54 Miller, Dickinson 200, 202 Milton, John 47, 50 Miracles 6, 11–13, 46–7, 54–7, 73–4, 76, 175, 178 Misak, Cheryl 201 Mitova, Veli 129–30 Mohammed 56–9, 65 Monroe Doctrine 104, 106, 112, 114, 179 Moral luck 19–25, 31, 41, 195 Moral questions 132–4, 136–9, 156–7 Moral sense 66–9 Moser, Paul 200 Moving the goalposts (fallacy) 76–7

Index Nagel, Thomas 195 Narcissism problem 122–3, 174–5, 192 Nathanson, Stephen 196, 199, 202 Natural theology 5–6 Nausen, Fridtjof 95, 97–8 Negligence 18 Newman, Cardinal Henry 11 No True Scotsman Fallacy 172 North Pole 95, 97–8 O’Connell, Robert J. 197 Objective evidence 115–6, 118–121 Oliver, Phil 197, 201 Orioles, Baltimore (baseball team) 11 Orthodoxy requirement 83–4, 162–2, 174, 177–9, 190–2 Options 87, 142, 156, 172 Forced 87, 91–6, 118, 132, 135, 138, 141–4, 156,167–170 As verbal 92–4, 118, 169 Momentous 87, 96–8, 132, 138, 142, 163–7 Reduced to liveness 98 Genuine 87, 99, 140, 156–7 Overbelief 34, 41, 87–8, 107, 112–4, 148, 153, 179–80, 191–4 Pascal, Blaise 31, 98–103, 108–9, 113, 114, 137, 165, 167, 170, 182 Peirce, C.S. 29, 80, 184, 191 Peter, St. 102 Phone number case 61–3 Philips, D.Z. 191 Plato 123–4, 138 Plantinga, Alvin 196 Pluralism 81, 122–3, 174, 179–80, 192, 197, 199, 200 Poe’s Law 195 Poison pill argument 73–5 Popkin, Richard 198 Pragmatism 80–1, 181, 184, 199–201 About belief 4–5, 27–9, 181–2, 184 Reconstruction of religion 6–7, 179–80, 190–2 Reconstruction of truth 121–3 Price, H. H. 195, 201 Providence 13–4, 16 Publicity requirement 59–64 Pyrrhonism problem (for evidentialism) 68–9, 106

Index Rationalization 14–17, 20, 25, 37, 165, 182–3, 185 Reck, Andrew 200 Regulative principles 72–5, 77, 108 Reid, Tomas 55 Religious hypothesis 6, 160–3, 180 Belief (R2, R2*, R2**) 160–3, 166, 173 Efficacy (R4) 163–5, 173–4 Eternality (R1) 160–3, 167, 180 Thou (R3) 166–7 Reversibility 87, 95–8, 156, 164–6 Road to Larissa 123–4 Romantic Conquest cases 94, 141–2, 145, 149, 154–8, 167–8, 170, 172 Rorty, Richard 161, 191, 200, 201 Rosebury, B. 195 Royal Society 9 Russell, Bertrand 179, 200, 202 Sacred tradition of Humanity, the 10, 43, 64–9, 70–4, 77, 105–6 Two-levels model 65–6, 70–1 Saletan, William 195 Santayana, George 161 Sauce for the goose 129, 144, 169 Save the world case 182 Schaffer, Michael 196 Schellenberg, J. L. 200 Scientific questions 136–7 Self-defeat problem (for Clifford) 29–30, 68, 107–8, 130 Sellars, Wilfrid 184 Seneca 45 Shah, Nishi 49, 130, 181, 198, 201 Sharpshooter’s fallacy 174–5, 190, 192 Ship Owner (SO) case ix, 12–22, 24–5, 27, 30, 32–3, 36, 41, 44–6, 53–4, 128, 169, 182 Sinclair, Upton 2 Skepticism 1, 5, 50–3, 63, 67–8, 72–3, 106–8, 116–8, 129, 137–9, 154, 198 Slater, Michael 161, 165, 198 Slope arguments 6, 38–40, 41, 48, 64, 90, 128 Content 34–9, 40, 44, 68 Habit 34–9, 44 Sloth, epistemic 25–9, 33, 49–50, 185, 194 Smith, John E. 198, 199, 201

213

Social Coordination, Argument from ix, 99, 114, 145–8, 150–2, 154–8, 166, 168, 170, 171, 172, 176, 200 Principle of Corporate Action (PCA) 146, 147–8, 151 Principle of Reliance (PR) 147–8, 151 Sosa, Ernest 198 Spectroscope case 71–2 Spinoza, Benedict 117 Stages of “Will to Believe” 99, 133, 198 Psychological 99, 101–8, 114, 198 Logical 86–95, 99, 198 Meta-epistemic 99, 114–128, 198 Epistemic 99, 131–148, 198 Stanley, Arthur 10 Stephen, James Fitzjames 11–12, 54, 82, 195 Stephen, Leslie 82 Stoics 20–22, 42, 45–7, 49, 68, 181 Strealth 186–7 Suckiel, Ellen Kappy 170–1, 197, 198, 200, 201 Sunstein, Cass 196 Switcheroo problem 85, 122–3, 174–5, 190 Talisse, Robert 181, 197, 198, 199, 200, 201 Talking-to-Rockin’ rule 86–7, 168–9 Tennyson, Alfred Lord 10 Testimony 54–64 Theosophy 88 Thomson, William 10 Thucydides 71 Tolerance 174, 178–80 Train robbers case 146–7, 157 Trolley case 93–4, 144, 168–9 Truth Norm 123–33, 151, 157, 169, 176–8, 200 Umpires 135 Undecidability 84, 99, 111–114, 126, 152, 154, 172 van Inwagen, Peter 76–7 Veracity requirement 46, 54–61, 78–9, 105 Victorian Crisis of Faith 10 Vitelli, R. 195 Voluntarism 31, 100–3, 108, 182, 198 Vorstenbach, Jan 196

214

Index

Wager, Pascal’s 31, 98, 101–3, 108, 112, 170, 182 Weintraub, Ruth 198, 199 Welchman, Jennifer 197, 198 Wernham, James 110, 166, 196, 198, 199, 200 West, Cornel 191 Will-to-believe 1, 5–7, 17, 25, 81, 84–5, 90, 93, 109–111, 144–5, 153, 163, 170, 180, 194, 200, 202 Bare thesis (WTB) 111–114, 130, 144, 149, 150, 154, 178, 198, 200 Composite thesis (WTB*) 153–4, 157, 158, 163, 170, 172, 177–9, 189 Modest (right and preferable to believe) 110, 130–1, 198

Strong (duty to believe) 110, 130–1, 198 Weak (right to believe) 104, 110, 112, 130–1, 177, 178 Wishful thinking 5, 36–8, 104, 106 Wood, Allen 49, 161, 200 Wright, Chauncey 80 Willings Broad 104–5, 108 Mere 104, 108, 148 Yale University 82–3, 139 Yankees, New York (baseball team) 11–12 Zamulinski, Brian vii, 19, 42–6, 49, 145, 182–4, 196, 197, 199, 202