Relativism and the Foundations of Philosophy

STE V E N D. H A LE S Relativism and the Foundations of Philosophy Relativism and the Foundations of Philosophy Rel...

59 downloads 1585 Views 2MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

STE V E N D. H A LE S

Relativism and the Foundations of Philosophy

Relativism and the Foundations of Philosophy

Relativism and the Foundations of Philosophy

Steven D. Hales

A Bradford Book The MIT Press Cambridge, Massachusetts London, England

( 2006 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. MIT Press books may be purchased at special quantity discounts for business or sales promotional use. For information, please email [email protected] .edu or write to Special Sales Department, The MIT Press, 55 Hayward Street, Cambridge, MA 02142. This book was set in Sabon on 3B2 by Asco Typesetters, Hong Kong, and was printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Hales, Steven D. Relativism and the foundations of philosophy / by Steven D. Hales. p. cm. ‘‘A Bradford book.’’ Includes bibliographical references and index. ISBN-13: 978-0-262-08353-9 (hc : alk. paper) ISBN-10: 0-262-08353-1 (hc: alk. paper) 1. Relativity. 2. Philosophy. 3. Methodology. I. Title. BD221.H33 2006 149—dc22 10

9 8 7

6 5

2006043837 4 3

2 1

To my colleagues at Bloomsburg University: Richard Brook, Gary Hardcastle, Wendy Lee, Scott Lowe, and Kurt Smith. Their collegiality and encouragement have made it possible for me to think and write about these issues.

Contents

Preface

ix

Introduction

1

1

Intuition and Philosophical Propositions

9

2

A Problem for Philosophical Knowledge

49

3

The Relativist Solution

4

The Naturalists’ Challenge Notes 187 References 195 Index 207

97 147

Preface

Relativism was one of the ﬁrst philosophical topics to capture my imagination. I remember grappling with the self-refutation objection back in my undergraduate days in a philosophy of science course under Robert Frederick. Even then I had the suspicion that claiming relativism to be merely relatively true wasn’t really going to work, although I had no idea why not. I set the issue to the side for a long time though, going to graduate school and pursuing other philosophical matters. I returned to relativism through the back door. As a graduate student at Brown, I was made a T.A., and the department chair at the time, Jaegwon Kim, asked all the grad students to submit a ranked listing of the courses in which they would prefer to assist. I put Existentialism dead last on my list, since I knew next to nothing about it. Naturally, that was course Jaegwon assigned me. A more senior grad student, Rex Welshon, was teaching the course. As a dutiful T.A. I struggled to understand the readings, and I became interested in making sense of Nietzsche, especially his thoughts about truth. Eventually Rex and I wound up writing some articles on Nietzsche and ultimately a book, Nietzsche’s Perspectivism (2000). Thinking about the issues Nietzsche raised led me back to the topic of relativism apart from any historical ﬁgure or context. Initially I just wanted to provide a successful solution to the self-refutation problem, but the more I thought about philosophical methodology and the role of intuitions in argument, the more I thought (to my surprise!) that a sort of relativism is right after all. Such relativism is the subject of the present book. I was one of the late Roderick Chisholm’s last two students. The other one, Dean Zimmerman, once told me that he thought that Rod must be rolling in his grave, what with me writing on Nietzsche, supporting metaphysical perdurantism, defending relativism, and no doubt engaging in other heresies. Perhaps Dean is right. Yet one of Rod’s many

x

Preface

virtues was his downright enthusiasm for criticism—for Rod the highest compliment one could pay a philosopher was to take him seriously enough to try to refute him. I don’t try to refute Rod in what follows, at least not directly, but I like to think that he would have approved of my waywardness. Certainly my own hope for this book is to be complimented in the Chisholmian sense. Besides Rod, others who have been inﬂuential, helpful, or supportive have been my Bloomsburg colleagues named in the dedication; audiences at an assortment of universities and conferences in the United States and Europe; Michael Devitt, Michael Lynch, Ernest Sosa, and my wife, Susan Hales. My sincere thanks to all of them. Special thanks to Greg Kenenitz and Ryan Wanttaja for compiling the index.

Introduction

This book is a defense of relativism. It must be admitted that most arguments for relativism are not very persuasive. Often they seem presented in a playful, ironic, Puckish sort of way that is inimical to serious philosophy.1 Such proposals are hard to evaluate because they are hard to pin down; one can’t have an earnest discussion with the class clown. Other kinds of relativism attempt to show that human reason itself is a contingent, historical artifact with no especial credentials as the royal road to truth. Thus the deliverances of reason are supposed to be truths ‘‘for us,’’ and not absolute or objective ones. Of course, this argument itself is a bit of reasoning, as would be any attempts to refute it. It seems that reason is on the agenda no matter what we might do, and so the wouldbe relativist is hard put to offer a genuine alternative.2 Many relativists cannot resist making grand and sweeping claims about which propositions are merely relatively true. They claim, for example, that everything is relative, or at least everything except the thesis of relativism is relative. Such declarations raise the hackles of those who note that, particularly in the physical sciences, humanity has made deﬁnite, evident progress in the advancement of knowledge. If relativism were true, such advance would be impossible. Who would want to ﬂy with a relativist pilot who thought the laws of aerodynamics were radically subjective, or undergo an operation by a relativist surgeon who thought modern medicine had no greater claim on truth than the theory of the bodily humors? I will not be supporting any of the preceding forms of relativism. The sort defended here is much more circumscribed and circumspect: I will defend a form of relativism that applies speciﬁcally to philosophical propositions alone. This, then, is the main thesis of what follows: Philosophical propositions are true in some perspectives and false in others.

2

Introduction

While (I hope) this thesis will be more plausibly defended than some of the more comprehensive and radical relativisms current in the marketplace of ideas, nevertheless I trust that it is sufﬁciently controversial. Obviously, much needs to be said just by way of explanation. What are philosophical propositions? What are perspectives? What is it to be true in one perspective and false in another? These questions will all be addressed. One that will be ignored is the nature of propositions in general. For the purposes of this book it doesn’t matter whether propositions are abstract objects, whether they are Fregean or Russellian, or whether they are reducible to sentences, statements, or even beliefs. The key point is that propositions are truth-value-bearers whose truth-value is perspective dependent and which are philosophical in nature. Many defenses of relativism are grounded in the philosophy of language. For example, different cultures, languages, persons, and so on are reputed to have distinct concepts, incommensurable with each other, which carve reality along different lines. Or it is claimed that concepts are vague and indeterminate, and that any precisiﬁcation is an arbitrary, artiﬁcial construct.3 Or there are problems for truth-conditional semantics for future contingents or epistemic modals that a relativistic approach is thought to solve.4 The present approach is orthogonal to these; I have no brief against philosophers of language who also want to be relativists, although I am inclined to think that the philosophical value of concepts is overrated with respect to relativism. An underexplored and more direct path to relativism is epistemological, stemming from a recognition of rationally irreconcilable methods of acquiring noninferential beliefs. Not everyone would employ the same methods of acquiring beliefs about philosophical propositions, even if everyone held their beliefs in wide reﬂective equilibrium and shared precisely the same set of concepts. Furthermore, these varying methods do not all lead to the same results. As we will see, this is the diversity that matters for establishing relativism. The book is divided into four somewhat lengthy chapters. In chapter 1 I consider the familiar methodology by which contemporary analytic philosophers ply their trade—the use of intuition. There is an important distinction between physical intuitions, which are informed expectations of future experience and of which principles will turn out to be the correct generalizations of nature, and rational intuition, which is the proper province of philosophy. Descartes, unsurprisingly, provided the canonical account of rational intuition, maintaining that it was of propositions, that the propositions known through rational intuition

Introduction

3

were necessary truths, and that intuitive knowledge is both foundational and certain. Since the linguistic turn in Frege’s Grundlagen helped found analytic philosophy, philosophers have often considered their task to be one of acquiring beliefs about how concepts can be factored into sets of necessary and sufﬁcient conditions. Whether or not conceptual analysis is at the heart of the philosophical enterprise, it is clear that rational intuition, with its corollary techniques of hypothetical examples and counterexamples, is the method by which philosophers come to have the beliefs they do. Moreover, the results of intuitive reﬂection and analysis are modal ones: often the mere metaphysical possibility of some case shows that a philosophical principle is false, a procedure that would work only if the principle in question were an ostensible necessity. Bounding the class of what is properly philosophical is extremely difﬁcult, and historically some things once thought to be a priori turned out to a posteriori after all. I have no general account of what makes a proposition a philosophical one and will rest content with pointing to a few familiar examples as paradigm cases. Yet why think that intuition provides any sort of justiﬁcation for belief? Contemporary philosophy may be saturated with intuitions and possible cases, but unless we have some reason to think that rational intuition is going to produce knowledge, then we are all spinning our wheels. I argue that, on pain of contradiction, we are compelled to accept that ‘‘the method of intuition justiﬁes some propositions’’ is selfjustifying. In other words, a form of foundationalism must be true for intuition-driven philosophy to get off the ground. The deliverances of intuition are supposed to be basic, noninferential kinds of evidence. Accepting that there is noninferential data upon which reason operates does not entail accepting that this data is known with certainty. There are plenty of good reasons for thinking that rational intuition is fallible and can be refuted both empirically (at least some of the time) and through the further use of intuition. I offer an explanation of how this works within the rationalist framework. At the end of chapter 1 I conclude that the most plausible, and widely employed, means of settling on a philosophical view is to try to achieve wide reﬂective equilibrium within one’s set of beliefs. Overall, chapter 1 presents a synoptic vision of analytic rationalism—a broad-brush overview of the structure of the rationalist, intuition-driven approach to answering philosophical questions. I sketch the general features of rationalism, arguing that its central features include the foundational reliance on rational intuition as a

4

Introduction

method of acquiring basic beliefs. These beliefs are then adjusted to achieve wide reﬂective equilibrium, a process of adjustment and reﬁnement that continues to operate as one encounters new thought experiments and arguments that depend for their appeal on intuition. While I am giving a sympathetic reading of rationalism in chapter 1, let me emphasize that I’m not defending it. In one sense the whole book might be viewed as an assessment of rationalism: I am presenting a challenge to rationalists, one that becomes clear in chapter 2. Once the case has been made for the widespread use and virtues of rational intuition as a belief-acquiring method, in chapter 2 I argue that there are other belief-acquiring traditions whose doxastic structure is isomorphic to rationalism. That is, there are other means besides intuition that people use to gain beliefs about philosophical topics. Moreover, these methods are foundational ones that produce beliefs used as noninferential data, which are then systematized and adjusted in an effort to achieve wide reﬂective equilibrium. The methodological difference between intuition-driven philosophy and these alternative methods is solely at the level of what provides the basic beliefs. The two other methods I consider are Christian revelation and the ritual use of hallucinogens. Although one may be tempted to summarily dismiss these ways of forming beliefs as nonrational, I believe that they are strikingly analogous to intuition and much harder to write off than one might think. I try to provide the most charitable reading of revelation and hallucinogen use that I can while still remaining true to the historical record of those traditions. If revelation and the ritual use of hallucinogens were simply alternative, equally legitimate paths to discovering philosophical truths, there would be no problem. However, the results of these approaches often conﬂict with the results of rational intuition. More exactly, there are key philosophical propositions that are core elements of the Christian religion or are widely accepted on the basis of a culturally sanctioned ingestion of psychotropic plants and yet are generally eschewed by philosophers employing rational intuition. If revelation or hallucinogens deliver a belief that p and intuition recommends a belief that not-p, which should we prefer? Naturally, advocates of intuition will try to show the superiority of their own method. I consider four arguments to this conclusion. The ﬁrst is that intuition is the best method for gaining philosophical beliefs because the set of beliefs that it generates is more stable, cohesive, and comprehensive than the networks of beliefs produced by the competition. Therefore, intuition is our best bet for arriving at philosophical beliefs in reﬂective equilibrium. The second argument is that the use of rational

Introduction

5

intuition is our standard justiﬁcatory practice, and we are entitled to use intuitions as evidence until confronted with powerful reason not to do so, a reason that we do not have. The third argument is the Reidian one that our epistemic faculties are all equally trustworthy, and so reliance on rational intuition is just as reasonable as our trust in sense perception. The fourth argument is that all parties agree on the value of reason: reason is used to operate on our basic beliefs and develop more comprehensive philosophical theories. That is, reason is the fundamental way we have to acquire epistemically virtuous beliefs; it is the standard package. Other methods are optional features. So in case of conﬂict, these optional methods are the ﬁrst to go. It is intuition that remains. I argue that, for one reason or another, all of the preceding arguments fail. The noninferential results of rational intuition have no more credibility than the deliverances of revelation or the ritual use of hallucinogens. This is a serious problem to which there are only three possible outcomes. (1) Skepticism: since we don’t know which is the right way to gain beliefs about philosophical propositions, we have no philosophical knowledge. It would be mere luck if we picked the method that actually led to truth. (2) Nihilism: there are no philosophical propositions. Putatively philosophical propositions are either noncognitive, or, more likely, covertly empirical. (3) Relativism: the truth of philosophical propositions is relative in some way to doxastic perspectives. I argue that the skeptical alternative should be rejected because it results in a paradox. Nihilism I largely bracket and return to in chapter 4, with a fuller consideration of epistemological naturalism. In chapter 3 I defend the relativist alternative. The ﬁrst task is to consider the best-known argument against relativism, namely, the charge that it is self-refuting. I argue that ‘‘absolutely’’ and ‘‘relatively’’ function as kinds of modal operators, and just as no one would defend ‘‘everything is possible’’ while all would ﬁnd ‘‘everything true is possibly true’’ to be the most modest of assertions, so too relativists ought to abandon ‘‘everything is relative’’ and instead embrace ‘‘everything true is relatively true.’’ Treating relativism in this way provides a nice solution to the selfrefutation problem. ‘‘Everything is relative’’ is inconsistent, whereas ‘‘everything true is relatively true’’ is not. Moreover, the well-known relativist relation of commensurability is shown to be formally on a par with the accessibility relation in modal logic. As a result, Davidson’s critique of conceptual schemes has no bite on the present account. Of course, even if there is a formulation of relativism that is internally consistent, that alone is no reason to be a relativist. Just as one could think that everything true is possibly true and still maintain that

6

Introduction

all truths are necessary ones, so too one could agree that everything true is relatively true and still consistently believe that all truths are absolute ones. However, I argue that there are good reasons to be a relativist about philosophical propositions. The key steps along the way are (1) philosophical propositions are knowable (skepticism is false); (2) there is no way to rationally prefer one method of acquiring basic beliefs about philosophical matters over another; and (3) different belief-acquiring methods lead to different, mutually inconsistent beliefs. The only way to make sense of all these things is to conclude that the very truth of philosophical propositions is perspective relative. The chapter concludes by considering and rebutting a variety of possible objections to the account. The ﬁrst is the idea that even if there are numerous ways to gain beliefs about philosophical propositions and these lead to different results, nevertheless convergence and agreement is in principle possible, and so relativism is not needed as a solution after all. Another objection is that relativism is inconsistent with the pursuit of wide reﬂective equilibrium, which I have considered a crucial component of all legitimate means of coming to have philosophical knowledge. The third objection is that I have failed to demonstrate relativism about truth. The most that can be inferred from the epistemic arguments presented is an epistemic conclusion, that justiﬁcation or warrant is perspective dependent. The fourth objection is a subtle revisit of the selfrefutation objection. If all philosophical propositions are merely relatively true, then the axioms and theorems of the relativist logic I develop are merely relatively true. However, I am committed to the view that those axioms and theorems must be absolutely true (true in all perspectives). The ﬁnal objection is that relativism is committed to an epistemic theory of truth. Since the epistemic theory of truth is wrong, so is relativism. All of these arguments are found wanting. In chapter 4 I examine the most sweeping objection: philosophical propositions are not merely relatively true, because there are no philosophical propositions. This is the contention of the naturalists, who hold that all propositions are ultimately empirical and that only the methods of science can yield knowledge. There are several ways to understand naturalism. One is as a primary methodological assumption that simply assumes that the scientiﬁc method is the correct way to gain knowledge. Another is to treat naturalism as a research program; that is, given various failures of rationalist epistemology, perhaps the thing to do is turn to the natural and social sciences as a way to save the business. Some push this point harder—intuition-driven epistemology has proven to be an un-

Introduction

7

mitigated disaster, and so the only course of action is to throw it out and start again with what does have a track record of success: empirical science. If naturalism is simply an undefended methodological assumption, then it is difﬁcult to see why fans of nonnaturalistic belief-acquiring methods should climb on board; no argument has been provided. The other forms of global naturalism suffer from two defects. The ﬁrst is that the alleged failures of rationalist epistemology are largely due to a failure to conclusively refute skepticism about the external world and to provide a self-evident foundation for science. Even granted this complaint, it does not show that rationalist epistemology has failed to produce knowledge about speciﬁcally philosophical issues like the nature of the mind, the scope of moral obligation, and the like. The second failure of global naturalism is that the scientiﬁc method itself seems riddled with normative principles. As an example, I consider Ockham’s Razor and argue that it cannot be justiﬁed on the grounds that it is truth conducive or even pragmatically helpful. Its value is irreducibly normative. But if normativity is an ineluctable component of science itself, then the naturalistic approach is not going to eliminate the normative in favor of a purely scientiﬁc picture of the world. There are more local naturalistic arguments that bear on the theses of this book as well. Some naturalists have recently argued that social science can actually debunk rational intuition as a legitimate beliefacquiring method. If intuition, revelation, and the ritual use of hallucinogens are all in the same boat together (as I maintain), and intuition has no value, then all can be dismissed. Another naturalistic approach is to question empirically the usefulness of ethnographic data for demonstrating that different cultures have interculturally inconsistent beliefs and attitudes. If such inconsistency cannot be demonstrated, then I cannot make out the case that different basic belief-acquiring methods lead to mutually inconsistent beliefs, and so the motivation for relativism disappears. I argue that neither the debunking approach nor ethnographic skepticism is ultimately successful. I concede that the relativist conclusions defended in this book are disturbing. But we should not grasp at ﬂimsy arguments in a desperate attempt to avoid them. However, I do not yet know of good arguments that would permit a rejection of the relative truth of philosophical propositions. The matter is more subtle and difﬁcult than sometimes supposed. It seems that intuition-driven philosophy does produce knowledge, but not knowledge of absolute truths.

1

Intuition and Philosophical Propositions

What do we want philosophy to do for us?1 What is the point of the pursuit of philosophy? The only real answer is that we (we modern Western intellectuals) hope that philosophy can answer certain questions, questions that science cannot. How should we live? What is the good life? Are there gods? What is the nature of peculiar things like properties, surfaces, color, events, and art? Pilate’s question: what is truth? What can we know, and how? Science has generally seemed ill equipped to address these sorts of issues; whether a Duchamp ready-made is genuine art or a clever put-up job just doesn’t seem soluble though observation, experimental testing, or mathematical explanation. Philosophy has traditionally addressed such issues through a different methodology, namely, through an appeal to common sense, ‘‘what we would say,’’ or a kind of intellectual intuition. This sort of methodology goes as far back as the Platonic dialogues. In book 1 of the Republic, Cephalus proposes that justice is treating others honestly and paying back what one has received from anyone (Republic, 331). Socrates rebuts Cephalus by claiming that it would be unjust to return weapons to a friend who had gone mad. Given that, it cannot be that Cephalus’ understanding of justice is correct. Socrates offers no other reason or argument to support his counterexample than ‘‘everyone I presume would admit’’ that it is true. In other words, he assumes that everyone shares his intuition that it is not in the interest of justice to return weapons to an insane person. All one need do is reﬂect on the claim, and through a kind of intellectual inner vision one comes to see both its truth and how it confutes Cephalus’ deﬁnition of justice. Indeed, Cephalus and the other interlocutors do just that (although Polemarchus tries to hold out against Socrates for a bit before capitulating). Socrates is aware of his own method and thinks it to be no more than the drawing-out of commonsense ideas possessed by everyday folk. He

10

Chapter 1

remarks to Theaetetus, ‘‘being only ordinary people, we shall prefer ﬁrst to study the notions we have in our own minds and ﬁnd out what they are and whether, when we compare them, they agree or are altogether inconsistent’’ (Theaetetus 154e). Theaetetus, of course, concurs. This way of trying to gain philosophical knowledge has not changed too much in the last 2,500 years. The idea that the intuition of philosophers is nothing other than common sense, distilled and puriﬁed, is a recurring theme, from Thomas Reid to G. E. Moore to Roderick Chisholm. Today it is difﬁcult to imagine the pursuit of philosophy without the use of intuition. Why have we all been persuaded that knowledge must be more (or other) than justiﬁed true belief? Because Edmund Gettier showed us cases in which a person has a justiﬁed true belief, but our intuitions are that the person still does not have the relevant knowledge. Why do we think that existence is not a perfection? On the grounds that it is counterintuitive to suppose that existence is a property that a thing may or may not have. Why are we convinced that torturing babies for pleasure is wrong? Either because it seems intuitively immoral, or because it falls out from more general moral principles that we ﬁnd intuitively plausible. Why do we think that predicate logic is an improvement on sentential logic? Because there are intuitively valid arguments (such as the famous (1) All men are mortal, (2) Socrates is a man, therefore (3) Socrates is mortal) whose validity cannot be captured by symbolization into sentential logic. Without appeal to intuition it is hard to see how we might even begin to settle such issues as whether compatibilism is an adequate notion of freedom, or whether intentional harm is morally worse than foreseen harm, or if parts are essential to their wholes, or whether multiple realizability arguments undercut token-identity theories of the mind. Indeed, philosophical ethics appears hopeless without intuition. For example, Leon Kass’s argument against human cloning is no more than a raw appeal to intuition. He writes: Repugnance is the emotional expression of deep wisdom, beyond reason’s power fully to articulate it. Can anyone really give an argument fully adequate to the horror which is father-daughter incest (even with consent), or having sex with animals, or mutilating a corpse, or eating human ﬂesh. . . . we are repelled by the prospect of cloning human beings not because of the strangeness or novelty of the undertaking, but because we intuit and feel immediately without argument, the violation of things we rightfully hold dear. (Kass 1997, p. 21)

Cloning is wrong because we directly intuit that it is. QED. While one can always dress up this intuition in argumentative clothing, such argu-

Intuition and Philosophical Propositions

11

ments themselves will depend for their ultimate force on moral intuitions and so will be no more than embroidery upon Kass’s brute intuiting. Even those who claim that intuition is a mysterious, heretofore unexplained faculty cannot resist the repeated use of it. As an example, Paul Boghossian writes that ‘‘no one has succeeded in saying what this faculty [of intuition] really is nor how it manages to yield the relevant knowledge’’ (2001, p. 6), yet in the very same article he asserts that ‘‘there is no intuitive sense’’ in which certain epistemic facts are analogous to other kinds of facts (p. 3), worries about ‘‘intuitive constraints’’ concerning justiﬁcation (p. 11), claims that circular arguments could prove ‘‘intuitively unjustiﬁable’’ conclusions (p. 11), discusses principles that are ‘‘intuitively too strong’’ (p. 25), and maintains that there are reliably formed beliefs that are ‘‘in some strongly intuitive sense, epistemically irresponsible’’ (p. 30). Moreover, he often uses language that amounts to the use of intuition in other terminology: in one instance he writes that ‘‘I don’t see . . . how this idea [that there are presumptively but defeasibly justiﬁed beliefs] could plausibly apply to the case at hand’’ (p. 7). Boghossian is not merely reporting his personal mental limitations here, something that would be of fairly narrow interest outside his circle of intimates. He is making an argument based on how things seem to him and his own common sense—in short, on rational intuition. These observations are not meant to pick on Boghossian especially, simply to note that even writers ofﬁcially skeptical of intuition seem to ﬁnd appeals to it irresistible. On the other hand, many philosophers are up front in admitting that their arguments are ultimately grounded in intuition and that philosophy essentially depends on it. For example, David Chalmers explicitly states that ‘‘all these arguments [on the nature of the mind] are based on intuition’’ (Chalmers 1996, p. 110). His defense of this claim is a familiar one: ‘‘I have tried to make clear just how natural and plain these intuitions are, and how forced it is to deny them’’ (ibid.). Daniel Dennett contends that much of what philosophy does is bandy about considerations that are meant to promote one or another intuition, what he calls ‘‘intuition pumps.’’ ‘‘The point of such thought experiments,’’ writes Dennett, ‘‘is to entrain a family of imaginative reﬂections in the reader that ultimately yields not a formal conclusion, but a dictate of ‘intuition’ ’’ (Dennett 1984, p. 12). Saul Kripke is wonderfully explicit about how crucial intuition is. In Naming and Necessity he writes, ‘‘I think having intuitive content is very heavy evidence in favor of anything, myself. I really don’t know, in a way, what more conclusive evidence one

12

Chapter 1

can have about anything, ultimately speaking’’ (Kripke 1980, p. 42). The bold central thesis of Laurence Bonjour’s recent (1998) book is that we must accept rational intuition ‘‘more or less at face value as a genuine and autonomous source of epistemic justiﬁcation and knowledge,’’ and that ‘‘philosophy is a priori if it has any intellectual standing at all.’’2 Of course, it could be that philosophy ought to have even less intellectual standing than it does. We need to give reasons to think that intuition is a way of gaining any sort of knowledge at all. Unfortunately, this foundational issue has not generally received the attention is deserves. In 1984, Dennett wrote that we fail to recognize the centrality of intuition owing to ‘‘a fairly vigorous institution of professional repression’’ (1984, p. 17). Since then philosophers have become increasingly concerned about the justiﬁcatory status of intuition. In their editors’ preface to Rethinking Intuition, Michael R. DePaul and William Ramsey write, ‘‘we do not think it is an overstatement to say that Western analytic philosophy is, in many respects, undergoing a crisis where there is considerable urgency and anxiety regarding the status of intuitive analysis. It is becoming increasingly clear that philosophers need to provide a clear and strong justiﬁcation for the important role intuitions play in their research’’ (DePaul and Ramsey 1998, p. x). Jaakko Hintikka has even suggested a moratorium on publishing papers that appeal to intuition without some theory of how intuition can provide the right kind of information or insight. He asserts that ‘‘blind faith [in intuition] is below the intellectual dignity of philosophers for whom unexamined intuitions should not be worth intuiting’’ (Hintikka 1999, p. 130). Clearly a detailed examination of how—or whether—intuition can provide knowledge is overdue. Two Kinds of Intuition There seem to be two sorts of intuition. One kind of intuition is scientiﬁc or physical intuition, and the other is philosophical or rational intuition. Physical intuition is a hunch, an expectation of how the world works based on one’s previous experience, and is to be conﬁrmed or disconﬁrmed by future experience. Intuition about empirical facts has gotten bad press because human beings turn out to be frequently in error about the natural world. The universe seems intuitively Newtonian to us, not relativistic. The sun’s apparent motion through the sky is best explained intuitively by the hypothesis that it really does move relative to a stable Earth. The thesis that a benevolent creator is responsible for the order

Intuition and Philosophical Propositions

13

and complexity we see evident around us has tremendous intuitive appeal. The idea that after ﬁve bad poker hands one should bet the house on the sixth (after all, one is due some good cards) is irresistible to novice gamblers. Yet such hypotheses generated by our empirical intuitions must eventually face the impartial tribunal of experience, and all these seemingly plausible inferences have been laid waste by science. Some recent writers have attempted to rehabilitate the reputation of empirical intuition. Evelyn Monsay has argued that there is an important difference between sensible intuition (i.e., naive commonsense) and physical intuition, which is a more sophisticated, informed sort of subconscious intellection in the practice of normal science. While still fallible, this sort of intuition plays an important role in creativity and originality in scientiﬁc work (Monsay 1997). Richard Feynman agrees. He contrasts ‘‘the inexperienced and crackpots,’’ who make simple, wild guesses about the consequences of an experiment or the outlines of a future physical law, with veteran scientists who, in advance of rigorous proof or experiment, are able to imagine what will be the case with accuracy (Feynman 1965, p. 171). Imagination, guessing, and intuition all contribute to the advance of science, but Feynman notes that it is ‘‘imagination in a terrible strait-jacket.’’ Empirical intuition is restricted by one’s prior knowledge of what’s physically possible and perhaps subconscious induction from prior experience about what results will follow from an experiment. Moreover, the expectations of even Feynman’s intuitions are capable of being proven wrong by empirical experience.3 Philosophers do not rely on this sort of intuition. Our intuition is rational, not empirical; it is the pure light of reason that shines upon philosophical propositions. We are never hauled before the tribunal of experience. Instead, as David Lewis observes,4 our intuitions are tested against one another, and their logical entailments drawn out and presented as evidence. Weaker, less well-rooted intuitions get trumped, and ultimately eradicated, when one squarely faces the implications of one’s deeper, more ﬁrmly held intuitions. The familiar system of example, counterexample, and straight-up appeals to naked intuition is entirely closed. Furthermore, the most cursory survey of the history of philosophy shows that it is thought experiments, not empirical ones, that constitute the great leaps forward. Think of Plato’s cave; Descartes’s evil genius; Locke’s prince and the cobbler; Rousseau’s state of nature; Nietzsche’s blond beasts of prey; Quine’s linguists translating ‘‘gavagai’’; Goldman’s papier-maˆche´ barns; Searle’s Chinese room; Putnam’s Twin Earth; Rawls’s original position; Nozick’s experience machine;

14

Chapter 1

Thomson’s kidnapped violinist; and no end of runaway trolleys, brain transplants, and teleportation. These are the common currency of our profession, and not a one is fundamentally empirical in nature, or even indicative of scientiﬁc insight. They all appeal in some fashion to our rational intuitions. The contemporary understanding of ‘‘intuition,’’ like many concepts in epistemology, ﬁnds its origins in Descartes. In the Rules for the Direction of the Mind, he writes, Let us now review all the actions of the intellect by means of which we are able to arrive at a knowledge of things with no fear of being mistaken. We recognize only two: intuition and deduction. By ‘‘intuition’’ I do not mean the ﬂuctuating testimony of the senses or the deceptive judgment of the imagination as it botches things together, but the conception of a clear and attentive mind, which is so easy and distinct that there can be no room for doubt about what we are understanding. Alternatively, and this comes to the same thing, intuition is the indubitable conception of a clear and attentive mind which proceeds solely from the light of reason. (Descartes 1628 [1985], Rule 4, §368)

Descartes claims that he is the ﬁrst to use ‘‘intuition’’ in this technical sense. He writes, ‘‘In case anyone should be troubled by my novel use of the term ‘intuition’ and of other terms to which I shall be forced to give a different meaning from their ordinary one, I wish to point out here that I am paying no attention to the way these terms have lately been used in the Schools’’ (Rule 4, §369). As examples of things one can know by rational intuition he lists ‘‘that [one] exists, that [one] is thinking, that a triangle is bounded by just three lines, and a sphere by a single surface, and the like’’ (Rule 4, §368). Subsequent writers have generally taken Descartes to mean that what one knows through rational intuition are necessary propositions. It is possible that one fails to exist, or that one think, but not when one is considering those claims. It is a necessary truth that ‘‘if I am cognitively grasping the proposition that I am thinking, then it is true that I am thinking.’’ Similarly, it is a (putative) necessary truth that ‘‘if I am thinking about whether I exist, then it is true that I exist.’’ Likewise it could not be the case that a triangle fail to be bounded by three sides or a sphere by a single surface. Information about such claims is not gleaned through experience (the ﬂuctuating testimony of the senses) or any kind of imaginative guessing, but through the faculty of intuition, which utilizes the light of reason. For Descartes, rational intuition is a method that one can employ to acquire beliefs about certain kinds of

Intuition and Philosophical Propositions

15

propositions. It is ‘‘only though intuition’’ that we arrive at ﬁrst principles, and it is upon these that deduction operates to yield additional knowledge. The conception of rational intuition ﬁrst laid out by Descartes has changed little. He claims that: (1) (2) (3) (4)

Intuition is of propositions. The propositions known through intuition are necessary truths. Intuitive knowledge is foundational. Intuitive knowledge is indubitable.

To get a better grasp of rational intuition, let us consider these in turn. Propositions and Concepts Since the linguistic turn of the early twentieth century, analytic philosophers have focused on concepts as the sort of objects about which we have intuitions. The philosophical enterprise itself has been recast as conceptual analysis, the attempt to formulate sets of necessary and sufﬁcient conditions for concepts like ‘‘knowledge,’’ ‘‘causation,’’ and ‘‘free will.’’ Ancient debates have been reinterpreted as, at heart, debates over deﬁnitions. For example, in Euthyphro, when Socrates asks the eponymous hero to provide him with moral instruction and share his knowledge about things divine, contemporary philosophers interpret Socrates’ request as asking for the deﬁnition of ‘‘piety.’’ It is the concept of piety that is at issue, a concept to be probed by consulting intuitions that presumably are shared by the debating parties. Thus when Euthyphro, on his way to court to prosecute his father for murdering a serf, replies that piety is doing as he is doing, namely, prosecuting murderers, Socrates rejoins that this is an unsatisfactory reply, as he is looking for a general deﬁnition of ‘‘piety’’ that would allow him to evaluate various different acts as pious or impious. Euthyphro tries again, this time deﬁning ‘‘piety’’ as those acts which the gods love. Socrates notes an ambiguity here: are the pious things those which only some of the gods love, or those which all the gods love? If the former is correct, then Hera might consider an act to be pious, whereas Hephaestus might regard the same act as impious. This would make obeying the gods a difﬁcult matter indeed. Euthyphro replies that pious acts are those which all the gods love. Socrates proceeds to examine difﬁculties with this deﬁnition also, raising the now-famous question of whether acts are pious because the gods love them, or whether the gods love them because they are pious.

16

Chapter 1

Thus Euthyphro can be seen as illustrating a classic bit of conceptual analysis. It begins with Euthyphro offering a deﬁnition, or analysis, of a concept, that is, explaining the problematic concept ‘‘piety’’ in more familiar terms. This proposal is then tested to see if it is genuinely explanatory (Euthyphro’s ﬁrst deﬁnition of ‘‘piety’’ was not), unambiguous and resistant to equivocation (Euthyphro’s second deﬁnition was not), and has acceptable implications (an issue they pursue with the third definition). Socrates, as is well known, argues that the gods love pious acts because they are pious; it is not the antecedent love of the gods that makes acts pious. His defense of this view is grounded in intuitions about the nature of love. Frank Jackson has recently defended the view of philosophy as conceptual analysis. For Jackson, ‘‘conceptual analysis is the very business of addressing when and whether a story told in one vocabulary is made true by one told in some allegedly more fundamental vocabulary’’ (Jackson 1998, p. 28). Jackson claims that one tests proposed conceptual analyses through the use of rational intuition. Once an analysis is proposed, it is must survive the method of possible cases. Various hypothetical scenarios are imagined, ones that need no more than bare logical possibility, and we consult our intuitions as to whether the analysis is applicable in the imagined scenario. Thus Jackson writes that Gettier ‘‘invited us to agree with his intuition that [Gettier cases] should not be described as cases of knowledge,’’ and that ‘‘Hilary Putnam and Saul Kripke refuted at least some versions of the description theory of reference by appeal to intuitions about possible cases’’ (1998, pp. 28–29). What are the philosophical questions to be answered by this method? Jackson writes that when we inquire into free will, for example, ‘‘what we are seeking to address is whether free action according to our ordinary conception . . . exists and is compatible with determinism’’ (1998, p. 31, emphasis in original). As Socrates said, we are ordinary people and must ﬁrst study the notions we have in our own minds. Jackson suggests that we can identify our ordinary notions of philosophical concepts through introspecting our intuitions about what seems most obvious and central about free action, determinism, belief, and so on. Intuition seems to be a universal faculty that, when coupled with reﬂection on possible cases, reveals the nature of one’s concepts. Not only is the faculty of intuition common among nondefective thinkers, but Jackson also believes that conceptual content of philosophical concepts is uniform among the learned and the vulgar alike. At least, it is largely uniform; he acknowledges that there may be people who have oddball,

Intuition and Philosophical Propositions

17

deviant ideas of philosophical terms. They just use words like ‘‘knowledge’’ in a way that is different from the norm. But in the end he is convinced that ‘‘often we know that our own case is typical and so can generalize from it to others’’ (1998, p. 37). There are several difﬁculties with Jackson’s vision of philosophy as conceptual analysis through the use of rational intuition. The ﬁrst is that concepts themselves are notoriously vague and slippery entities that may not be up to the rigorous demands placed on them. Many philosophers have wearied of the endless post-Gettier debates over the correct analysis of knowledge, and it may well be that this muddle is a result of the fact that ‘‘knowledge’’ does not have a univocal meaning, one that all competent speakers can come to see after a parade of possible cases and counterexamples. Yet Jackson writes that ‘‘we are simply concerned with making explicit what is, and what is not, covered by some term in our language’’ (1998, p. 51), as if matters are cut-and-dried and the true nature of knowledge, the real, objective fact of its meaning in normal English, is just waiting for a clear explanation by a clever philosopher. I think we ought to be quite skeptical of such conceptual optimism. Russell was closer to the truth when he wrote: A certain difﬁculty as regards the use of words is unavoidable here, as in all philosophical inquiries. The meanings of common words are vague, ﬂuctuating, and ambiguous, like the shadow thrown by a ﬂickering street-lamp on a windy night; yet in the nucleus of this uncertain patch of meaning, we may ﬁnd some precise concept for which philosophy requires a name. If we choose a new technical term, the connexion with ordinary thought is retarded; but if we use the common word with a new precise signiﬁcance, we may seem to run counter to usage, and we may confuse the reader’s thoughts by irrelevant associations. It is impossible to lay down a rule for the avoidance of these opposite dangers; sometimes it will be well to introduce a new technical term, sometimes it will be better to polish the common word until it becomes suitable for technical purposes. (Russell 1914a, pp. 128–129)

Concepts are our tools for communication and understanding the world. It is important that we understand how to employ these tools, but it is equally important to understand their limits. Ordinary, pre-analytic concepts may not be up to the relentless demands of conceptual analysis.5 A second objection to Jackson is that whatever intuition is, it is not the simply the humble Socratic procedure of uncovering the ordinary folk notions of ‘‘free action,’’ ‘‘belief,’’ and so on. Socratic commonsensism is premised on the assumption that there is some folk theory of freedom, knowledge, morality, and so on, that, if not universally accepted, is

18

Chapter 1

at least very widely the object of tacit agreement; through the Socratic method, reﬁned into the method of imaginary cases and counterexamples in the pursuit of conceptual clariﬁcation, we can uncover our commonly shared theory. This theory might need revision or amendment to make it consistent, but, so it is assumed, there is still much widespread agreement. The problem here is the blithe assumption of universality. While Jackson takes conceptual analysis to elucidate what governs our classiﬁcatory practice, he thinks that this social practice can be discovered from the armchair. No social science, no polls or studies are needed. Or, at most, a foray into a classroom is all that’s required: ‘‘everyone who presents the Gettier cases to a class of students is doing their own bit of ﬁeldwork, and we all know the answer they get in the vast majority of cases’’ (Jackson 1998, p. 37). Philosophers who take scientiﬁc methodology seriously have rightly taken Jackson to task on this point. Stephen Stich and Jonathan M. Weinberg have criticized Jackson for failing to control for sample bias (the participants in his polls are advanced philosophy students at elite universities), and experimenter bias (no blinding of the experimenter, who has strong antecedent beliefs about how things should turn out). Furthermore, polling through a public show of hands or some other public expression tends to suppress dissenting opinions (Stich and Weinberg 2001). Stich and Weinberg go on to argue that when genuine polls are conducted, ones that meet the requirements of authentic social scientiﬁc methodology, what is revealed is that there is no common, underlying folk theory of philosophical concepts. Instead what one ﬁnds is systematic variation of the intuitions of the folk across ethnic, educational, and economic class backgrounds. Stich and Weinberg take their arguments to undermine the use of intuition in philosophy entirely; I will examine their claims in greater detail in chapter 4. There I will argue against their more extravagant thesis that the method of intuition is worthless. However, it does seem that they have provided compelling reasons for rejecting the Socratic, commonsense model of philosophy as unveiling a commonly shared folk theory of concepts like ‘‘knowledge,’’ ‘‘free action,’’ ‘‘duty,’’ and the like. A third objection to Jackson is that it is unclear why he thinks that to analyze a concept is to use a more fundamental vocabulary, as he claims. It is unclear what ‘‘fundamental’’ even means in this context: that sentences involving the deﬁniendum could be eliminated in favor of sentences using only the deﬁniens? that molecular concepts are composed out of atomic ones? As Nietzsche commented in The Gay Science (§355),

Intuition and Philosophical Propositions

19

what we want from knowledge is the reduction of the strange to the familiar. Euthyphro’s deﬁnition of piety in terms of what the gods love is supposed to provide a way to judge whether an act is a pious one, namely by seeing whether it is loved by the gods, something that is presumably easier to ﬁgure out. What the gods love is supposedly less mysterious than which acts are pious. The JTB analysis of knowledge is the hypothesis that the strange concept ‘‘knowledge’’ can be explicated in terms of the supposedly more familiar concepts ‘‘justiﬁcation,’’ ‘‘truth,’’ and ‘‘belief.’’ It is not that ‘‘justiﬁcation,’’ ‘‘truth,’’ and ‘‘belief’’ are elements of a more fundamental vocabulary than ‘‘knowledge’’; it is simply that those terms are supposed to be more familiar ones. This is the reason one looks up deﬁnitions in the dictionary—not to discover a more ‘‘fundamental’’ vocabulary, but to reduce (in the Nietzschean sense) the strange word to more familiar ones. This is even clearer when one considers that there are instances of one-term ‘‘analyses.’’ For example, utilitarians argue about the nature of the good based on competing intuitions, and come to identify the good with some nonnormative property like pleasure or preferencesatisfaction. For a hedonistic utilitarian, the purported ‘‘analysis’’ is that X is the highest good iff X is pleasure. Yet this is no longer an explication of a complex concept in terms of its component concepts, but the brute positing of the truth of a proposition about the nature of the summum bonum. The good is identiﬁed with pleasure—a proposal that may be helpful to someone ignorant of moral language yet familiar with pleasure, or that may play a theoretical role in a more comprehensive moral theory, but not a claim that reduces the summum bonum to a more fundamental cluster of concepts. In this case it does not seem that rational intuition is delivering a conceptual analysis of goodness, but is instead directed upon the proposition that the highest good is pleasure. Whether intuition is ultimately of the nature of concepts or of propositions is not crucial to the central theses of the present book, although I think there are some reasons to be less than sanguine about Jackson’s insistence that rational intuition is devoted to conceptual analysis. What is important is that intuition is a method of acquiring beliefs. Jackson seems to think that intuition delivers de re beliefs about concepts that they are such-and-such; I will focus on intuition as a way to gain beliefs about a certain class of propositions. This is the way that many philosophers have understood rational intuition, from Descartes (as we saw earlier) to Hintikka, who writes that ‘‘philosophers’ intuitions have typically pertained to the truth or acceptability of various propositions’’

20

Chapter 1

(Hintikka 1999, p. 134). The question to consider, then, is just what kinds of propositions intuition supposedly makes credible. Philosophical Propositions and Necessity Rational intuition is traditionally distinguished from empirical intuition by claiming that rational intuition generates beliefs about putative necessary truths, whereas the most that empirical intuition produces is beliefs about contingent truths. This is, of course, Descartes’s view, and it is one that has gained a wide following among contemporary philosophers. For example, George Bealer writes, ‘‘For you to have an intuition that A is just for it to seem to you that A’’ and then adds that he means a priori intuitions about what is necessary (Bealer 1996, p. 5). Ernest Sosa concurs that intuitions are ‘‘a priori intellectual seemings, which present themselves as necessary’’ (Sosa 1996, p. 151). Bonjour writes, ‘‘It is common to refer to the intellectual act in which the necessity of [an analytic] proposition is seen or grasped or apprehended as an act of . . . rational intuition’’ (Bonjour 1998, p. 102). Chisholm also argues that rational intuition yields knowledge of necessities, and traces this view to Aristotle (Chisholm 1977, pp. 38–39). There are two points to be made here: ﬁrst, not all necessary truths are known by rational intuition. Some necessary propositions are known as a result of logical inference from other necessities. One could derive that necessarily q from the premises that necessarily (if p then q) and that necessarily p. According to the sense in which ‘‘intuition’’ is generally understood, the truth of the premises might be known or justiﬁed through intuition, but the conclusion that necessarily q is not. Whatever intuition amounts to, it is not an inference procedure. To pick another example, one could derive ‘‘necessarily, if p then p’’ from the empty set using modal conditional introduction. In such a case, a belief that ‘‘necessarily, if p then p’’ is justiﬁed as a result of a valid derivation. On the other hand, one could just directly apprehend the truth of ‘‘necessarily, if p then p’’ without the derivation; one can just see that it is impossible for ‘‘if p then p’’ to be false without further proof. It is this latter case that is supposed to be an example of the use of rational intuition. Second, it is generally considered that there are necessary truths that are not philosophical propositions. For example, logical and mathematical truths are traditionally taken to be necessary. Intuitionists in mathematics hold that mathematical axioms, like philosophical proposi-

Intuition and Philosophical Propositions

21

tions, are justiﬁed through intuition, but I will not enter into that debate here. Part of what is traditionally taken to separate philosophical propositions from empirical ones is that the former and not the latter are either metaphysically necessary or impossible. What marks out speciﬁcally philosophical propositions from mathematical ones is a bit harder to say; certainly many mathematicians (e.g., Pythagoras, Descartes, Leibniz, Russell) drew no such distinction or saw the foundations of mathematics as contiguous with philosophy and philosophical logic. For the purposes of this book, all that matters is that we can pick out philosophical propositions ostensively. Rational intuition is supposed to bear upon propositions like these: Intended harm is morally worse than foreseen but unintentional harm. Psychological facts supervene on microphysical facts. All of the parts of a whole are essential to it. Gettier cases prove that knowledge is something other than justiﬁed true belief. If S knows that P and S knows that P implies Q, then S knows that Q. If determinism is true, libertarian free will is impossible. ‘‘The F is G’’ is equivalent to ‘‘There is an F, there is no more than one F, and anything that is an F is G.’’ Multiple realizability arguments refute token-identity theories of mind. Apparent organization in the universe is best explained by positing an omniscient, omnibenevolent, omnipotent God. Physical objects wholly exist at each moment of their existence and endure through time.

All of these propositions are taken to be implicitly prefaced with ‘‘j’’. Necessarily, if determinism is true, libertarian free will is impossible; there is no possible world in which determinism is true and yet there are agents that are free in the libertarian sense. The nature of these necessary propositions, whether they are a priori, analytic, or whatever, is irrelevant. All that matters here is that these are the sorts of propositions whose truth-value is supposedly apprehended by rational intuition or not at all. (N.B.: I am not asserting that any members of the preceding list are true; I am simply observing that they are generally taken to be putative metaphysical necessities whose truth-value is demonstrable solely though rational intuition and philosophical reason.)

22

Chapter 1

In traditional analytic philosophy we use possible cases—even examples that are physically impossible in our universe—to try to refute propositions like those above. If those propositions were not taken to be necessary truths, then refutation by demonstration of what’s possible would not work. The usual pattern of argument is for someone to assert j p, and for an interlocutor to rejoin ‘‘but wait; Uq, and Uq ! Usp; therefore it is not the case that j p.’’ For example, in the debate over secondary qualities, dispositionalism about color is the view that colors are dispositional properties, much like solubility is a dispositional property. Salt is soluble in water, even when it is in solid form and is not actively dissolving. Under normal conditions, salt will dissolve in water (assuming the water is not already saturated, the salt sample is not super dense, and so on). So salt has a dispositional property, the property of solubility, and tennis balls have a dispositional property, the property of yellowness. A typical dispositionalist deﬁnition of color is the following: x is yellow ¼ df. x appears yellow to normal observers in normal conditions. This sort of deﬁnition is a familiar bit of analytic philosophy. The deﬁnition is meant to express a true proposition if it justiﬁed by intuition, and it is to be rejected if an appropriate counterexample is found. Justin Broackes has devised the following counterexample to color dispositionalism. Imagine a yellow object that also emits death rays. These rays are incredibly fast acting—so fast, in fact, that anyone who looks at the object is killed before he can see its color. Ex hypothesi, the object will be yellow, but will have no disposition to produce experiences of yellow in normal observers. Thus colors are not dispositions. Since there are no death rays, and no yellow objects that emit them, why is this a counterexample? Because it seems logically possible that there could be such an object. The bare logical possibility of a phenomenon won’t refute an empirical truth, though. Richard Dawkins has said that he would consider that Darwinism has been refuted at a stroke if a rabbit fossil were found that radiometrically dated to 2,000 million years old. Yet Dawkins doesn’t think Darwinism has been refuted because we can imagine the bare possibility of a two-billion-year-old rabbit. The difference is, of course, that ‘‘x is yellow ¼ df. x appears yellow to normal observers in normal conditions’’ is supposed to be a necessary truth. If it is a necessary truth, then it is impossible for there to be a yellow object with no disposition to produce yellow sensations. Since it is possible (argues Broackes), yellow (and mutatis mutandis the other colors)

Intuition and Philosophical Propositions

23

isn’t a disposition. Notice that Broackes’s counterexample may not even be physically possible in the actual world; logical possibility is all that’s needed to play the game of intuition-driven analytic philosophy. Thus, unless philosophical propositions are metaphysically necessary, Broackes’s counterexample refutes dispositionalism just as much as the mere possibility of a two-billion-year-old rabbit refutes Darwin—that is, not at all. The role of science, of empirical observation, in the assessment of philosophical propositions is, then, easy to discern. The most deﬁnitive way to show that something is possible is to show that it is actual. For example, suppose one were to offer a simple wavelength theory of color: Colors are identical with speciﬁc wavelengths of light; yellow is identical with 577 nm light, red is identical with 650 nm light, and so on. This proposal is supposed to be a metaphysical necessity—there are no other possible worlds in which color is something other than light wavelengths. It’s not as if a wavelength theory is true in the actual world, whereas dispositionalism or subjectivism is true in other possible worlds. Showing that there are metameric color matches to 577 nm light scientifically refutes the wavelength theory. That is, a spot of light that has just two components, monochromatic 540 nm light (which most people see as green) and monochromatic 670 nm light (which most people see as red), will appear indistinguishable from a spot of pure 577 nm light. Yet the 540 nm–670 nm blend contains no 577 nm light at all, as a photospectrometer would reveal. Thus the 540 nm–670 nm spot turns out on the wavelength theory not to be yellow, while appearing manifestly yellow, indeed just as yellow as 577 nm light. Therefore, by reductio, the wavelength theory is false.6 In this case it is the scientiﬁc discovery of metamers that refutes the wavelength theory, not, as in the case of color dispositionalism, a mere reﬂection on a logically possible case. This is not to suggest that all philosophical propositions can be falsiﬁed by empirical discovery. Normative propositions—at least if cognitivism is true—seem particularly resistant to science. Unlike the metaphysics of color, it is difﬁcult to see what empirical theory or discovery could refute a claim that ‘‘the highest good is pleasure,’’ or ‘‘one ought never to treat others merely as means,’’ or ‘‘Mozart was a better composer than Salieri,’’ or ‘‘Samuel Smith’s oatmeal stout tastes better than Miller Lite.’’ Social scientiﬁc polls could establish what most people believe about these propositions, even what most informed people believe about them, but barring more general philosophical reasons to accept noncognitivism about the normative, such surveys will be unable to

24

Chapter 1

rebut such propositions. Furthermore, many philosophers since G. E. Moore have argued that in the case of ethical claims, naturalism fails to capture their action-guiding nature. Ethical reasoning engages our preferences and desires in a way that mere descriptive facts cannot, so no purely naturalistic treatment of morality could be successful.7 In a similar vein, Kim (1988) argues that epistemological naturalism fails because it is powerless to address the ineliminable normative component in epistemology. He contends that epistemology, in particular the concept of justiﬁcation, is essentially normative, and thus a Quinean ‘‘naturalized epistemology’’ is not a form of epistemology at all. Whatever the Quinean program is, without a normative element it is an entirely separate subject and not properly viewed as a ‘‘successor’’ subject. One might give up normative inquiry and pursue science instead, Kim suggests, but that is not to use science as a means of settling the normative. Science also seems completely powerless to demonstrate the truth of a philosophical proposition. For example, although proposals about the nature of color obviously have to be consistent with scientiﬁc ﬁndings, it is hard to see what scientiﬁc discovery would settle what color actually is. Furthermore, if philosophical propositions are metaphysical necessities, then it is no surprise that the scientiﬁc method cannot show their truth. Science is in the business of discovering physical law and hence what must be the case in the actual world, what is physically necessary given the laws of nature. Yet even ﬂawless knowledge of natural law and physical necessity will not yield up what must be the case no matter what the laws of nature actually are, what must be true in every possible world. Noninferential Beliefs So far I have argued that traditional rationalist philosophy relies on rational intuition to acquire beliefs about an ostensively delimited class of philosophical propositions. These propositions, if true, are metaphysically necessary, which serves to explain how they can be refuted through appeals to possible cases and nonempirical counterexamples. The nature of rational intuition, as a belief-acquiring faculty, is left largely unexplained; this doesn’t make intuition magical or mysterious, as its detractors at times charge. One needn’t have a theory of perception to be an empiricist. Nevertheless, the question of how rational intuition is supposed to provide knowledge cannot be left unaddressed. A. J. Ayer

Intuition and Philosophical Propositions

25

wrote, ‘‘philosophers who ﬁll their books with assertions that they intuitively ‘know’ this or that moral or religious ‘truth’ are merely providing material for the psycho-analyst’’ (Ayer 1936, p. 120). Are there reasons to reject Ayer’s skepticism? Bonjour (1998) has given the following argument for the claim that there must be some a priori knowledge. Even the opponents of a priori knowledge agree on basic rules of logical inference; for example, they agree that there is such a thing as a logically valid argument. The most robust of naturalists would accept this argument: 1. 2. 3.

If naturalism is true, then there is no a priori knowledge. Naturalism is true. Therefore, there is no a priori knowledge.

However, one can conditionalize this argument and arrive at a sentence that is truth-functionally true, namely S: if [(if naturalism is true, then there is no a priori knowledge) and (naturalism is true)], then there is no a priori knowledge. S is knowable a priori; it is a logical truth that needs no empirical support or prior reasoning on its behalf. Therefore, concludes Bonjour, there is some a priori knowledge. I am sympathetic to Bonjour’s argument, but the a priori knowledge of truth-functional truths alone is too anemic to support the pretensions of rationalist philosophy. Logical truths are all well and good, but knowledge of them is not going to give us content-rich beliefs about substantial philosophical propositions. As Ayer suggests, it is moral and religious truths that are supposedly apprehended by intuition. What about them? Why should the fact that A seems true to me be evidence for A, or a justiﬁable reason to accept A? Another way of putting the question: why should our phenomenological sense of self-evidence be reliable? If a proposition seems self-evident to me, why should this seeming be accurate? Descartes and Sartre claimed that rational intuition yields certainty,8 but we need not suppose this. In fact, there are good reasons for thinking that intuition is quite fallible, as will be discussed in the next section. Nevertheless, a fallibilistic intuitionism still faces a challenge once posed by Ayer: ‘‘unless it is possible to provide some criterion by which one may decide between conﬂicting intuitions, a mere appeal to intuition is worthless as a test of a proposition’s validity’’ (Ayer 1936, p. 106). If intuition offers prima facie justiﬁcation of belief, and two people have conﬂicting intuitions, then they have two conﬂicting prima facie justiﬁed beliefs. In itself this is not too uncommon; but I take it that Ayer’s point is that if there is no way for the two opponents

26

Chapter 1

to settle their differences except by appeal to yet more intuitions which again conﬂict, then they have reached an epistemic impasse. If they cannot both be right, the method of ‘‘pure reason’’ has proven useless in ﬁnding the truth. But there is a more basic, and more troubling, problem than Ayer’s close to hand. He fears the irresolvability of conﬂicting root intuitions. What’s worse is that unless we have reason to think that the beliefacquiring method of intuition provides some degree of justiﬁcation for the beliefs that are premised on it, appeal to intuition is a total nonstarter to begin with. Establishing that intuition provides justiﬁcation proves exceedingly difﬁcult because it does not seem that rational intuition can be justiﬁed empirically, yet if we attempt to justify it intuitively, we encounter a circularity of a particularly vicious sort. That is, before we can even ask to what degree or under what circumstances intuition provides justiﬁcation of ostensibly necessary propositions, we must establish that intuition does justify at least some propositions at least some of the time. Let us put the key question this way: is the proposition ‘‘the method of intuition justiﬁes some propositions’’ true? To answer this question, we need to see what reasons we have to accept the proposition, what arguments might be adduced on its behalf; in short, we need to determine whether we are justiﬁed in accepting the contention that intuition justiﬁes some propositions. It is in attempting to justify this last that the deep problem of intuition arises most acutely. The Problem of Intuition 1. If a proposition is epistemically justiﬁed, then it is justiﬁed either a priori or a posteriori. (Premise) 2. If a proposition is epistemically justiﬁed a priori, then its justiﬁcation depends on the method of intuition justifying some propositions. (Premise) 3. If the proposition ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed, it is not justiﬁed a posteriori. (Premise) 4. ‘‘The method of intuition justiﬁes some propositions’’ is epistemically justiﬁed. (Premise) 5. Nothing is self-justifying. (Premise) 6. If ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed, it is justiﬁed a priori. (From 1, 3) 7. If ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed, then its justiﬁcation depends on the method of intuition justifying some propositions. (From 2, 6)

Intuition and Philosophical Propositions

27

8. The justiﬁcation of ‘‘the method of intuition justiﬁes some propositions’’ depends on the method of intuition justifying some propositions. (From 4, 7) 9. Thus ‘‘the method of intuition justiﬁes some propositions’’ is not epistemically justiﬁed. (From 5, 8) 10. ‘‘The method of intuition justiﬁes some propositions’’ is and is not epistemically justiﬁed. (From 4, 9) Bealer accepts most of the central claims of the problem of intuition argument (PI). He argues that intuition is not infallible, but that nevertheless it has a kind of modal reliability.9 It is worth noting that this modal link actually falls out of the preceding reﬂections. According to lemma 8 of PI, the justiﬁcation of ‘‘the method of intuition justiﬁes some propositions’’ depends on intuition. Yet the deﬁnition of ‘‘intuition’’ tells us that the rational intuition of a proposition requires that the proposition seem to be necessarily true. Thus if we ﬁnd it intuitive that intuition justiﬁes some propositions, then the modal tie must also seem present to us—necessarily, intuition justiﬁes some propositions. To his credit, Bealer correctly acknowledges that demonstrating this reliability of intuition is a task that intuition itself must perform, as lemma 8 in PI states. He writes: ‘‘The thesis that intuitions have the indicated strong modal tie to the truth is a philosophical (conceptual) thesis not open to empirical conﬁrmation or refutation. The defense of it is philosophical, ultimately resting on intuitions’’ (Bealer 1996, p. 8). While Bealer is concerned with the possibility that this requirement leads to an undesirable circularity, he does not, I think, see how problematic it is. He writes: ‘‘There is nothing vicious about this circle, . . . denying that intuitions are evidence leads to epistemic self-defeat; it is impossible to have a coherent epistemology without admitting intuitions as evidence.’’10 My reply is to agree that Bealer is right: without admitting intuitions as evidence, we cannot construct a coherent epistemology. In fact, we cannot coherently do philosophy at all. There are two ways out of this conundrum. The ﬁrst is to insist, as Bealer does, that since we do coherently engage in philosophy, and we do construct sensible epistemologies, that the circularity of justifying intuition on the basis of intuition is not vicious. That is, there must be something wrong with the problem of intuition as I have presented it. The other alternative is to admit that the problem of intuition is deep and far-reaching, and upon its solution hangs the whole of rationalist philosophy. This, I submit, is the unpalatable truth. It is easier to see if one considers a Bealerian response to

28

Chapter 1

Hume. ‘‘Denying that the principle of induction is justiﬁed leads to epistemic self-defeat; it is impossible to have a coherent science without admitting that the future will, in a relevant way, be like the past. Therefore, the problem of induction is not really a problem.’’ There could be Bealerian answers to a host of philosophical problems. For example, if Quine’s thesis about the indeterminacy of translation is correct, then we could never genuinely translate from French to English. But we do construct genuine translations. Therefore the indeterminacy of translation is a pseudo-problem. If Cartesian-style skepticism is right, then we know nothing about an external world. But we do have knowledge of an external world; hence Cartesian skepticism is not a genuine problem. It is too easy to solve philosophical problems by this method; they are solutions by theft. While Bealer conﬁdently asserts (1996, p. 30) that no reason to discontinue using intuitions as evidence will be forthcoming, lemma 9 of PI tells us that ‘‘intuition justiﬁes some propositions’’ is not epistemically justiﬁed. Unless one of the ﬁve premises of PI is rejected, we have in hand as good a reason as we might want to discontinue using intuitions as evidence, Bealer’s protestations aside. PI doesn’t merely suggest that justifying the use of intuition via intuition has a funny odor of circularity; it shows that a logical contradiction results from the practice. This is a serious problem. I am sympathetic to Bealer’s view (and the view of many others already mentioned, like Bonjour and Sosa) that without intuition, philosophy is left high and dry. But our desperate need to count on intuition does not absolve it of the charge that its use is both justiﬁed and unjustiﬁed. I conclude that Bealer has not found a satisfactory way of escaping PI. Is there a way out? Let’s examine the premises. The ﬁrst premise states that if a proposition is epistemically justiﬁed, then it is justiﬁed either a priori or a posteriori. A priori justiﬁcation, traditionally conceived, is justiﬁcation that derives from pure reason alone and does not depend on experience.11 If understanding a proposition is all that is needed to be justiﬁed in believing it, then such a proposition is justiﬁed a priori. Propositions that are justiﬁed but not justiﬁed a priori are justiﬁed a posteriori. That is, the a priori and the a posteriori are exhaustive and exclusive categories of justiﬁcation. More positively, a posteriori justiﬁcation requires experiential input; it is the sort of justiﬁcation needed by empirical propositions, and the sort provided by science. This classiﬁcation is hardly news; the foregoing is a familiar distinction. Not everyone thinks this classiﬁcation is worthwhile, however. Radical empiricists

Intuition and Philosophical Propositions

29

like positivists and naturalizers deny that anything is justiﬁed a priori. As Michael Devitt writes, ‘‘[we should] reject a priori knowledge and embrace ‘naturalism,’ the view that there is only one way of knowing, the empirical way that is the basis of science. From the naturalistic perspective, philosophy becomes continuous with science’’ (1999, p. 96). As far as justiﬁcation is concerned, it is a posteriori or bust. To this I have two responses. The ﬁrst is that PI is a powerful aid to the radical empiricist position; it concludes that the method of justiﬁcation through a priori intuition is not only circular but also inherently contradictory. It seems a tactical error for the naturalists to strive to reject the premises of an argument that dovetails so nicely with their own conclusions. The second response is that the intended audience of PI and this essay more generally are those philosophers who already rely on a priori reasoning and appeals to intuitiveness to establish their results. They are certainly in no position to deny premise 1. It is for them that PI is such a problem. The second premise maintains that if a proposition is epistemically justiﬁed a priori, then its justiﬁcation depends on intuition. My defense of premise 2 is conceptual: it is part of the very nature of rationalism that rational intuition is used to justify a priori propositions. That’s just what rationalism does. This observation isn’t a vindication or defense of the rationalist approach; indeed, we will see in the subsequent chapter that there are alternative methods of gaining beliefs about the a priori that have just as much claim to our allegiance as intuition. Even if the connection between intuition and a priori justiﬁcation seems something like the ‘‘secret power’’ Hume saw in causation, it is nevertheless clear that there is an intimate tie between them. Rational intuition is deﬁned in terms of the seemingly necessary truths. That is, intuition is supposed to apprehend or make manifest the truth of noncontingent propositions. The justiﬁcation yielded by intuition, Bonjour writes, ‘‘depends on nothing beyond an understanding of the propositional content itself’’ (1998, p. 102).12 We grasp a proposition like ‘‘necessarily, if A implies B and B implies C, then A implies C,’’ and the promise of its truth-value is available to our intuition. If we admit a priori justiﬁcation at all, it is hard to imagine why we would think a proposition is a priori justiﬁed if not either as a direct dictate of intuition or as a logical consequence of basic intuitions. At least, it is hard to imagine for traditional rationalist philosophers. Naturalists who deny that any proposition is justiﬁed a priori are logically committed to the truth of premise 2. The premise is a conditional, and they deny the antecedent, so, of course, the conditional is

30

Chapter 1

true. Thus naturalists are logically committed to premise 2, and rationalists are conceptually committed to it. Premise 3 states that however ‘‘the method of intuition justiﬁes some propositions’’ itself gets justiﬁed, it cannot be justiﬁed a posteriori. Apparently, only Hilary Kornblith has directly argued against premise 3. He attempts to show how an a posteriori justiﬁcation of the method of intuition could be accomplished; his argument goes something like this.13 Naturalists accept externalist accounts of justiﬁcation, like reliabilism. So, for a naturalist, ‘‘the method of intuition justiﬁes some propositions’’ means roughly ‘‘the method of intuition is a reliable way of acquiring true beliefs.’’ Now, rational intuition is supposed to apprehend the truth-values of necessities. If it does so reliably well, then we may conclude that the employment of intuition produces justiﬁed beliefs, that is, that we are justiﬁed in accepting ‘‘the method of intuition justiﬁes some propositions.’’ For empirical science to evaluate or test the effectiveness of intuition, it would have to compare the results of our intuitions with the correct assessment of logically necessary propositions as determined empirically. For example, suppose I ﬁnd it intuitive that the following proposition is necessarily true: ‘‘all the parts of a physical object are essential to the whole.’’ Let us suppose that my use of intuition justiﬁes (to some extent) this proposition for me only if my use of intuition is generally reliable. If I am to establish this reliability a posteriori, then I must be able to appraise various propositions like ‘‘necessarily, all the parts of a physical object are essential to the whole’’ in an a posteriori manner. Kornblith thinks that this could be accomplished if we take philosophical investigation to be the investigation of natural kinds and couple this with a causal theory of reference for natural kind terms (1998, p. 134, passim). Kornblith’s argument here is highly programmatic, and he does not give even one worked-out example of an a posteriori appraisal of a philosophical proposition. He offers up ‘‘work on the psychology of inference, concept formation, cognitive development, and so on’’ (p. 136) as examples of a wholesome naturalistic approach, but this contradicts his own view (p. 133) that ‘‘epistemologists ought to be concerned with the nature of knowledge, not the concept of knowledge; the proper subject matter of ethics is the right and the good, not the concepts of the right and the good, and so on.’’ Empirical studies of concept formation cannot be relevant to philosophy if we care not about the concept of the good, but only the good itself.

Intuition and Philosophical Propositions

31

There is a more difﬁcult problem latent in the Kripkean strategy that Kornblith favors. Kripke argues that there are a posteriori necessities. They occupy the logical space provided by the rigid designation of proper names and natural kind terms coupled with a causal theory of reference for such terms. Naturalists who blithely suppose that the Kripkean approach is all they need to save necessity ignore two important facts, which I will brieﬂy mention. The ﬁrst is that Kripke does not claim, nor does it follow from his arguments, that all necessary truths are justiﬁed a posteriori. To establish this point would require arguments and resources beyond those Kripke provides. Second, Kripke repeatedly and explicitly relies on rational intuition to bolster his arguments. He does so, for example, to determine which properties of a thing are essential and which accidental, which phrases act as rigid designators and which do not, and so forth. That is, intuition is used to justify modal propositions in Kripke’s theory, and this justiﬁcation is not a posteriori. Naturalists will have other ways of undermining premise 3, from a posteriori criticisms of intuition to more global arguments for the view that only the methods of science justify any propositions at all. However, I will bracket these broader concerns for now, with the promise that they will be more fully addressed and rebutted in chapter 4. For now, I conclude that ‘‘the method of intuition justiﬁes some propositions’’ cannot be justiﬁed a posteriori.14 Premise 4 states that ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed. If this premise is false, then we are unjustiﬁed in believing that intuition provides justiﬁcation for anything. Given the evidence cited earlier that philosophy relies heavily on intuition, and that it cannot help but do so, the falsity of premise 4 would be a hideous turn of events for the profession. We think of intuition as a method of determining the truth of a priori propositions as epistemically preferable to reliance on palmistry, crystal balls, and Tarot cards. Yet if we are not justiﬁed in believing that intuition justiﬁes anything, it is difﬁcult to see how it is better than the occult. If two methods of discovering the truth are worthless, it is splitting hairs to ﬁgure out which is worse. Of course, the fact that everyone but the radical empiricists should be put out of a job unless premise 4 is true is no evidence that the premise is true. It will be argued in chapter 3 that the truth of ‘‘the method of intuition justiﬁes some propositions’’ is perspective dependent. Intuition can provide philosophical knowledge, but only granted a perspective, like

32

Chapter 1

that of analytic philosophy, in which it is an integral belief-acquiring method. The naturalist position that rational intuition has no legitimate role to play in procuring philosophical knowledge is set aside, pending a fuller discussion in chapter 4. For now I note that rationalist philosophers who rely on intuition in making arguments and gaining beliefs are committed to the view that intuition is an epistemically respectable way of gaining justiﬁed beliefs. Premise 5, the last assumption of the proof, states that nothing is self-justifying. What does this mean? The idea is that there are no basic propositions whose justiﬁcation stems from no source other than themselves that we are justiﬁed in accepting. That is, if we must already assume the truth of a proposition p in order to construct an argument showing that we are epistemically justiﬁed in accepting p, then p is selfjustifying. And so it is with intuition. We must assume that the method of intuition delivers justiﬁed propositions when we employ this method to show that ‘‘the method of intuition justiﬁes some propositions’’ is true. It is this kind of self-justiﬁcation that is rejected in premise 5. This I believe expresses nothing more than a familiar rejection of foundationalism, a rejection now so common that Richard Foley has declared that we are now in a ‘‘postfoundationalist era.’’15 He argues that, if we want to hang on to necessary propositions and engage in traditional philosophy, we have no choice but to believe that ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed.16 He is correct—the rejection of premise 4 spells disaster. Yet Foley does not recognize that in order to maintain premise 4, some other premise of PI must go. I have already argued that philosophers who are not radical empiricists are committed to premises 1 and 4, and that there are excellent reasons to accept premises 2 and 3. I believe that the only plausible way out is to reject premise 5. We must accept that some propositions are self-justifying and that ‘‘the method of intuition justiﬁes some propositions’’ is one. Put another way, this proposition is axiomatic for traditional intuition-driven philosophy. In adopting such an axiom, we adopt a form of foundationalism.17 Modest foundationalism in this sense expresses the thesis that there are justiﬁed propositions whose justiﬁcation depends on nothing other than themselves. Obviously, various stronger versions have been defended, maintaining that such propositions are known with certainty; that all of our knowledge rests on basic propositions; and so on. I take no stand on these, and acceptance or rejection of them is beyond the requirements of the modest foundationalism discussed here.18 We cannot, despite Foley’s

Intuition and Philosophical Propositions

33

announcement of the postfoundationalist era, abandon all forms of foundationalism and continue to engage in traditional philosophy. To sum up, the ﬁve premises of PI form an inconsistent set. I have argued that there are only two ways to avoid commitment to the elements of this set: (1) become a radical empiricist/naturalist, give up the a priori, and abandon the use of rational intuition; or (2) accept that a modest foundationalism is true and that ‘‘the method of intuition justiﬁes some propositions’’ is epistemically justiﬁed on the basis of nothing other than the method of intuition itself. The only way for a proponent of traditional a priori philosophy to avoid the problem of intuition is to reject premise 5, and by so doing, endorse a modest foundationalism. Here, then, is our choice: either a form of foundationalism is true or philosophy grounded in the use of rational intuition is bunk. Coherentists might object here that I have not demonstrated the truth of a modest foundationalism at all; rather, the credence of intuition might be justiﬁed via a chain of reasoning involving intuition itself and yet further epistemic resources. Coherentism can accommodate the selfjustifying character of intuition by admitting that its justiﬁcation is circular without conceding that it is foundational. I do not think that a coherentist reply is going to be successful. Crucial to the basic coherentist idea is the view that no proposition is special—all justiﬁably believed propositions tie together into a cohesive, mutually supporting network. The justiﬁcation of any particular claim is derivative from the global, holistic justiﬁcation that accrues to the whole. As one gains new information and tries to incorporate it into the set of justiﬁably believed propositions, adjustments must be made, with some former elements of the set stripped of their credentials as justiﬁed. Since no proposition is unshakeable in some foundationalist sense, any member of the set might get jettisoned. However, no rationalist argument will result in rejecting the proposition ‘‘the method of intuition justiﬁes some propositions.’’ To dislodge that proposition is to give up rationalism altogether. Giving up rationalism is of course what empiricists and naturalists are eager to do, but the point being made is that advocates of traditional a priori philosophy (again, most of us) have to hang on to the view that ‘‘the method of intuition justiﬁes some propositions’’ is itself justiﬁed. It’s just not up for grabs. Here’s another way of putting the point. Consider the set of justiﬁed a priori propositions (naturalists bear with me here). I have argued that each member of the set is justiﬁed, ultimately, through the use of rational intuition. A rationalist coherentist would have to argue that

34

Chapter 1

propositions previously considered to be justiﬁed on the basis of intuition can be rejected on the basis of new and improved intuition-based arguments. Now consider the proposition ‘‘the method of intuition justiﬁes some propositions.’’ To reject this proposition one would have to provide an argument, based on intuition, that gives us good reason to abandon ‘‘the method of intuition justiﬁes some propositions’’ as truly being justiﬁed—surely a self-defeating procedure if ever there was one! A rationalist cannot remove that particular proposition without giving up the game entirely. Thus ‘‘the method of intuition justiﬁes some propositions’’ is special: it is self-justifying and particularly resistant to rejection. This state of affairs seems best described as a sort of foundationalism. A coherentist might rejoin that there is a way to explain the specialness of ‘‘the method of intuition justiﬁes some propositions’’ without capitulation to foundationalism, namely by accepting a form of rulecircularity as a legitimate means of inference. Boghossian (2000) has mounted this sort of argument to defend the use of logic in argumentation. The problem there is that it seems that we are not entitled to use a rule of inference unless we are ﬁrst justiﬁed in accepting the corresponding general claim that the rule is valid. But any argument for the validity of a nonderived rule of inference will depend on that rule, which seems illegitimately circular. Boghossian’s strategy is to accept a conceptualrole semantics for logical connectives: connectives just mean what they do by participating in some inferences and not in others. To understand a logical connective is therefore to recognize which inferences it licenses and which it does not. Boghossian then proposes the following principle: (L) If M is a genuinely meaning-constituting rule for S, then S is entitled to infer according to M, independently of having supplied an explicit justiﬁcation for M. There’s nothing wrong with circularity, Boghossian claims, so long as the condition in (L) is met. For his approach to the knowledge of legitimate logical rules of inference to be exported to our knowledge that intuition is a legitimate means of gaining justiﬁed beliefs, we would have to suppose that a principle like this is correct: (L*) If M: ‘‘if intuition supports a belief in proposition P, then P is justiﬁed’’ is a genuinely meaning-constituting rule for S, then S is entitled to infer that P is justiﬁed on the basis of intuition, independently of having supplied a explicit justiﬁcation for M.

Intuition and Philosophical Propositions

35

Why accept principle (L)? One answer that Boghossian mentions is that it helps out with the circularity problem for knowledge of logic. Yet that reason makes (L) seem a bit ad hoc. His real defense of principle (L) is that it ‘‘is intuitively plausible’’ (Boghossian 2000, p. 250). To be sure, this is a traditional rationalist defense of an epistemic principle and may be an acceptable reason to accept (L). But is it a good reason to adopt (L*)? A coherentist needs to supply an explanation as to why intuition is justiﬁcation-conferring without admitting that it is foundational. The Boghossian-style answer under consideration is the conceptual-role semantics principle (L*). If the reason to accept (L*) is its intuitive plausibility, then we seem to have arrived back at foundationalism after all. Why think that if intuition supports a belief in proposition P, then P is justiﬁed? Because of principle (L*). Why think (L*) is justiﬁed? Because intuition supports it! That’s no more than saying that it is on the basis of intuition that we ﬁnd intuition to be justiﬁcation-conferring; it is selfjustifying—in short, it is foundational. I have argued that if intuition-driven philosophy has merit at all and can provide us with justiﬁed beliefs, some of the beliefs generated by rational intuition must be basic beliefs that are noninferentially justiﬁed. (In what follows, I will use ‘‘noninferential belief’’ and ‘‘basic belief’’ as synonyms.) The only proposition that clearly must be noninferentially justiﬁed on the basis of what has been argued so far is ‘‘the method of intuition justiﬁes some propositions.’’ The use of rational intuition itself is legitimated only through intuition. But what of the more robust and substantive philosophical claims that intuition claims to support? Consider this one: ‘‘Necessarily, Gettier cases prove that knowledge is something other than justiﬁed true belief.’’ The truth of this proposition is not something that is inferred from other beliefs, or that is the result of argument. Rather, once one understands what Gettier cases are and what the JTB analysis of knowledge is, rational intuition is supposed to deliver the belief that Gettier cases show that knowledge must be something other that justiﬁed true belief. As James Van Cleve writes, ‘‘There is such a thing as just ‘seeing’—by a kind of intellectual vision—that a proposition is true. . . . ‘Seeing’ of this sort is what many philosophers call ‘intuition’ ’’ (1983, p. 36). One just sees that knowledge is not mere justiﬁed true belief. Or again, to consider a moral example, if a deontologist rebuts act utilitarianism by citing a case in which one intentionally harms an innocent in order to prevent greater collective harm to others, the case is meant to prompt the intuition that act utilitarianism is mistaken in

36

Chapter 1

recommending the intentional harm of the innocent as one’s duty. Through intuition, reﬂection on the proffered case produces (or is meant to by the case-giver) a belief. Possible cases are supposed to trigger the faculty of intuition and directly produce a noninferential belief about some philosophical proposition. As Bonjour writes, ‘‘appeal to apparent rational insight is epistemologically so basic and fundamental as not to admit of any sort of independent justiﬁcation’’ (1998, p. 148). Bealer agrees: ‘‘there is no alternative but to identify intuition as a basic source of evidence’’ (1996, p. 31). This not to say that all philosophical beliefs are basic or the direct result of rational intuition. Accepting the consequences of elaborate and intricate arguments clearly produces plenty of philosophical beliefs. However, those arguments themselves frequently include premises that are not based on still more argument, but are justiﬁed through the appeal to intuition alone. In Rules for the Direction of the Mind, Descartes made four claims about rational intuition: that it is of propositions, that the propositions known through intuition are necessary truths, that intuitive knowledge is foundational, and that intuitive knowledge is indubitable. I have defended the ﬁrst three of these. Let us now turn to a consideration of his fourth claim, that knowledge produced by rational intuition is indubitable. Indubitability and Reﬂective Equilibrium Descartes’s view that intuitive knowledge is indubitable has been repeatedly criticized. The history of philosophy is littered with examples of philosophers conﬁdently declaiming that one or another proposition is, in Descartes’s words, derived ‘‘from the light of reason, so that we cannot doubt their truth’’ (Descartes 1644 [1985], pt. 3, para. 80). For example, Descartes thought the light of reason showed that Euclidean geometry and mathematics explain all natural phenomena. Kant also thought that it was an a priori necessary truth, known by intuition, that space conforms to Euclidean geometry (Kant 1787 [1965], B41). Locke thought intuition showed it to be logically impossible ‘‘that things wholly void of knowledge, and operating blindly, and without any perception, should produce a knowing being’’ (Locke 1690 [1974], bk. 4, ch. 10, §5). Frege found it intuitive that any property determines a set—a proposition that he thought necessarily true. Wittgenstein declared that it is impossible to travel to the moon (although what exactly he meant may

Intuition and Philosophical Propositions

37

be difﬁcult to say) (Wittgenstein 1969, §286). All of these ostensible certainties are now considered to be false. Einstein showed that space is non-Euclidean; Darwin demonstrated that dumb bits of matter could, operating blindly, produce a knowing being; Russell discovered a paradox in Frege’s intensional speciﬁcation of sets; and, of course, there was Apollo 11. Scientists have often latched on to such examples as a means of a wholesale rejection of the contributions of rationalist philosophy. Steven Weinberg, for example, devotes an entire chapter of Dreams of a Final Theory (1992) to explaining why philosophy is at best useless for the pursuit of science and at worst positively harmful.19 He argues that theories such as veriﬁcationist-style positivism hindered the acceptance of atomic theory at the beginning of the twentieth century and that post-Kuhnian relativism serves only to turn a scientiﬁcally illiterate public against the advance of knowledge. I suspect that he is right about these examples. However, it is also true that allegiance to any theory, including purely scientiﬁc ones, can last longer than the warrant for those theories, and that such allegiance can impede the development and acceptance of superior successors. Indeed, quantum mechanics and general relativity could not win universal acceptance until an older generation of physicists actually died off, to say nothing of the difﬁculties faced by Galileo or Darwin. The truth behind Weinberg’s criticisms is that philosophers have made two mistakes. The ﬁrst is thinking that a proposition is logically necessary and demonstrable through rationalist means when in fact it is contingent and capable of being supported or disproved only through empirical evidence. This was the problem with the Cartesian and Kantian view about Euclidean geometry. Whether the geometry of the universe is Euclidean cannot be settled by the use of pure reason; it is a matter for physics determine. Interestingly, though, current physics seems to be supporting the idea that the density of the universe is on average about one hydrogen atom per cubic meter (i.e., a value called W is equal to 1), which means that the average curvature of spacetime is ﬂat, and Euclid was more or less right after all.20 If the universe is in fact less dense than this ðW < 1Þ, then it will expand forever, and ﬁnally thermodynamic entropy will reduce everything to an undifferentiated particle soup (‘‘The Big Chill’’). In such a scenario, the geometry of the universe is best described by Gauss-Lobachevskian geometry. If W > 1, then the expansion of the universe will eventually come to a halt, reverse, and mutual gravitational attraction will compress everything into an inﬁnitely

38

Chapter 1

dense singularity (‘‘The Big Crunch’’). In that case, the universe’s geometry is Riemannian. In any event, the geometry of the universe is an empirical issue, one that turns on the value of W, not one that is discovered by rational intuition and tested through logically possible counterexamples. The second mistake philosophers have made is in undervaluing or even discounting the relevance of scientiﬁc discoveries for the assessment of philosophical propositions. The assumption that we don’t need to know anything about the world beyond our armchair is depressingly common in philosophy. As I argued earlier, this assumption is false. Even granted everything a rationalist could want—the autonomy of philosophy, the metaphysical necessity of philosophical truths, the method of rational intuition as a way to acquire justiﬁed beliefs—it is still the case that scientiﬁc ﬁndings can disprove (some) philosophical theses. Ted Sider puts the point nicely when he writes, ‘‘science is certainly relevant to metaphysics since inconsistency with a ﬁrmly established scientiﬁc theory is as good a reason against a theory as one could ask for’’ (Sider 2001, p. xiv). We saw an example of such inconsistency earlier in the case of a simple wavelength theory of color. The nature of color is supposed to be a metaphysically necessary truth; thus under the wavelength theory, j (x is yellow iff x emits or reﬂects 577 nm light). It is impossible for x to be yellow without 577 nm light being involved. Yet the empirical discovery of color metamers shows that there are most deﬁnitely yellow things that neither reﬂect nor emit any 577 nm light at all. Even though an empirical discovery thereby refutes a putative metaphysical necessity, the nature of color still seems to be a metaphysical issue, and the truth about color may well be a necessary one. This is different from the case of the universe’s geometry, where the geometrical structure of the universe is a contingent fact, one dependent on the physical facts about how much matter exists and how spread out it is. Knowing the appropriate physical facts will enable one to know which geometry of the universe is correct, whereas it is not nearly as clear that knowing all the physical facts about the chromatic vision system and mental processing will enable one to know which theory of color is correct—at least, if the truth about color is metaphysically necessary, as many assume. Naturalists may reject the distinction I am drawing here between the nature of the universe and the nature of color, but their rejection is predicated on a more comprehensive rejection of metaphysical necessity. Science is relevant not only for metaphysical claims, which may not come as a big surprise, but also for normative ones. Suppose that fully

Intuition and Philosophical Propositions

39

speciﬁed moral propositions are either necessarily true or necessarily false. This would mean, as traditional moral philosophers have thought, that having all the empirical evidence in the world about actions, intentions, and the consequences of actions will not tell us which moral theory is the right one. However, physical evidence and scientiﬁc discovery can serve to set parameters as to which moral theories could be right. For example, if ‘‘ought’’ implies ‘‘can,’’ then one’s moral duty cannot be to perform a physical impossibility. What’s physically possible and impossible in our world (that is, what the laws of nature permit and what they exclude) is a matter for science to determine. Thus, if a proposed moral theory includes a commitment to the performance of p as a general moral duty, and scientiﬁc discovery has shown that p is nomologically impossible, the moral theory is thereby refuted. Another example from ethics is consequentialism. Consequentialists maintain that moral obligation is connected with the production of good consequences. While what counts as a good consequence may be beyond the reach of science and determinable only through rational intuition, the consequences themselves are caused by agents. The nature of causation and the prediction of expected outcomes are either within or at least overlap the domain of science. Others have recognized the relevance of science for a priori philosophy. For example, Henderson and Horgan (2000) maintain that there is a priori knowledge, and they defend the use of rational intuition. They go on to argue that knowledge of the a priori is intimately tied up with conceptual competence and to provide an externalist account of concept possession (which they admit is a bit tenuous), as well as a spirited defense of the value of wide reﬂective equilibrium for philosophical reﬂection. The upshot, they conclude, is that a priori knowledge is not wholly divorced from the empirical as defenders of classical ‘‘high-grade’’ a priori would have it. I pass over their account of concept possession without comment. However, I am in obvious agreement with their assessment of the importance of wide reﬂective equilibrium for philosophical reasoning and concur that scientiﬁc evidence is often relevant to philosophical conclusions. So in the end we have excellent reasons to reject Descartes: rational intuition is not infallible and does not produce beliefs that are indubitable. There are two ways for intuition-generated beliefs to be defeated: one is by empirical discovery, which has already been discussed. The second way that a belief produced by intuition could be overthrown is through further a priori reasoning.

40

Chapter 1

Undercutting the justiﬁcation of a belief acquired through rational intuition as a result of other beliefs generated by rational intuition is no more mysterious than defeating the justiﬁcation of a perceptual belief via other perceptions. Even if we consider perceptual beliefs to be foundational, this does not mean that they cannot be shown to be false through further beliefs also acquired through perception, even the perceptions of others. An analogous procedure holds in the case of rational intuition. Sometimes the purely a priori approach to rebutting the deliverances of intuition produces quite deﬁnite results; it was pure reason by which Russell discovered his paradox and so showed it to be false that there is a set that corresponds to every property. The derivation of a logical paradox or formal contradiction is a powerful reason to reject the supposition that allowed the derivation. In other cases additional intuition-based reasoning tends to make a host of propositions plausible, ones whose joint plausibility is greater than the intuitive motivation for some claim with which they are inconsistent. The attentive reader has no doubt noticed that the account I am presenting of philosophical method is related to the concept of reﬂective equilibrium. This notion has its origins in Goodman and Rawls. Goodman was concerned with how one might justify inferential rules: Principles of deductive inference are justiﬁed by their conformity with accepted deductive practice. Their validity depends upon accordance with the particular deductive inferences we actually makes and sanction. If a rule yields unacceptable inferences, we drop it as invalid. Justiﬁcation of general rules thus derives from judgments rejecting or accepting particular deductive inferences. This looks ﬂagrantly circular. I have said that deductive inferences are justiﬁed by their conformity to valid general rules, and that general rules are justiﬁed by their conformity to valid inferences. But this circle is a virtuous one. The point is that rules and particular inferences alike are justiﬁed by being brought into agreement with each other. A rule is amended if it yields an inference we are unwilling to accept; an inference is rejected if it violates a rule we are unwilling to amend. The process of justiﬁcation is the delicate one of making mutual adjustments between rules and accepted inferences; and in the agreement achieved lies the only justiﬁcation needed for either. All this applies equally well to induction. (Goodman 1954, pp. 63–64; italics in original)

Goodman is attempting to explain the justiﬁcation of deductive and inductive inferences, both for the inferences themselves and for the rules of inference. In Stich’s assessment of Goodman, he curiously ignores Goodman’s own contention that rules and particular inferences alike are justiﬁed by being brought into agreement with each other. Instead,

Intuition and Philosophical Propositions

41

Stich interprets Goodman as offering a test for inference procedures: a system of inferential rules is justiﬁed just in case it passes the reﬂective equilibrium test (Stich 1990, p. 78). Stich then proceeds to criticize this idea by noting that there is much evidence that even upon reﬂection many people accept provably poor inference procedures like the gambler’s fallacy, thinking that the probability of a sequence of events is higher than that of any of the components in the sequence, and so on.21 I will set aside the matter of inference rules. My interest is in the idea of reﬂective equilibrium as procedure for adjusting and maintaining one’s system of beliefs. This is the use of the term as made famous by Rawls: When a person is presented with an intuitively appealing account of his sense of justice (one, say, which embodies various reasonable and natural presumptions), he may well revise his judgments to conform to its principles even though the theory does not ﬁt his existing judgments exactly. . . . From the standpoint of moral philosophy, the best account of a person’s sense of justice is not the one which ﬁts his judgments prior to his examining any conception of justice, but rather the one which matches his judgments in reﬂective equilibrium. As we have seen, this state is one reached after a person has weighed various proposed conceptions and he has either revised his judgments to accord with one of them or held fast to his initial convictions (and the corresponding conception). (Rawls 1971, p. 48)

This is the sort of procedure that I think is followed in the most promising sort of analytic, intuition-driven philosophy. The early modern philosophers like Descartes and Spinoza hoped that philosophy could be set upon the same kind of deﬁnitive, unassailable basis as mathematics and physics, with certain foundations and rigorous deductions leading to substantive philosophical conclusions about the mind, the divine, and the nature of knowledge. This has proven to be an unreasonable hope. I have argued that rationalism must take the deliverances of rational intuition as basic beliefs that are nevertheless defeatable ones. Philosophy takes these intuition-spawned beliefs and attempts to systematize them in an effort to show how they are explanatorily connected, conceptually linked, and on the whole hang together in some coherent way. Reﬂective equilibrium thus has two components. The ﬁrst is the idea of taking one’s considered philosophical intuitions and developing general theories that explain and unify the intuitions and can provide guidance on future philosophical questions. This is not quite the same thing as developing an inference procedure in Stich’s sense. For example,

42

Chapter 1

utilitarianism is supposed to capture prevalent intuitions about the importance of producing good consequences with our actions and also be capable of providing answers to moral questions that may not yet have been asked. For any question of the form ‘‘in situation X, ought agent A to do Y?’’ utilitarianism has, in principle, an answer based on the outcome of action Y and the outcomes of other actions that A might perform instead. In this sense the theory has predictive power. Reﬂective equilibrium requires not only that the theories we develop accommodate widespread intuitions, but also that our intuitive beliefs themselves be subject to rejection in light of truly excellent theories. So if we have a moral theory that makes good sense of the vast majority of our moral intuitions and has predictive power of the sort just mentioned, then its results ought to trump the occasional wayward intuition. Intuitive beliefs are like data points: further philosophizing is an attempt to give the best account of how to draw a line though them; the data inform the linedrawing project, while at the same time some data points are dismissed as outliers. The second feature of reﬂective equilibrium is that it is wide. That is, in developing philosophical theories we need to do more than consider only our abstract, a priori intuitions. We must also take into account the ﬁndings of other ﬁelds of inquiry in forming an overall picture of the world. Our philosophical beliefs should ﬁt together with our best scientiﬁc and social scientiﬁc theories. Rawls defends the idea of wide reﬂective equilibrium, arguing that moral philosophy and the best theory of justice must incorporate and take notice of psychological theories of the person and of agency along with economic concepts of rationality, decision making, and game theory.22 Sosa endorses the Rawlsian position, extending it (as I have) to the philosophical enterprise as a whole: ‘‘philosophical reﬂection enhances rational justiﬁcation through wide reﬂective equilibrium’’ (Sosa 1991, p. 266). Philosophy may be autonomous, but it is not isolated. This chapter has been an attempt to sketch out a version of rationalist philosophy that coheres with and explains common practice in a plausible and sympathetic way. I have argued that the force of intuitions drives most present-day philosophizing, and that the use of these intuitions is best seen as a belief-acquiring method that delivers beliefs about philosophical propositions. The nature of philosophical propositions I have indicated largely by ostension, although unless they are putative metaphysical necessities, the familiar use of merely possible counterexamples and cases has no evidentiary value. The beliefs that rational

Intuition and Philosophical Propositions

43

intuition produces are noninferential, basic ones that are nevertheless fallible and revisable. The resulting network of beliefs is then adjusted and modiﬁed to attain wide reﬂective equilibrium, with some intuitions ultimately rejected and others retained as we systematize our philosophical views and make them congruent with our scientiﬁc knowledge. In this manner philosophers reach a rational consensus about the nature of the mind, the scope and limits of knowledge, the ontology of objects, the existence and nature of the divine, moral obligation and permissibility, and the rest of our domain. Some Objections Objection 1 Kai Nielsen (1993) has argued that reﬂective equilibrium is equivalent to a coherence conception of justiﬁcation and that it is therefore antithetical to a foundationalist treatment of intuition such as the one presented here. This is mistaken. At least, it is mistaken for the present account of a fallibilist foundationalism. Rational intuition is the source of prima facie justiﬁed yet impeachable noninferential beliefs upon which reason operates to develop a more comprehensive system of philosophical beliefs. To achieve reﬂective equilibrium, one may need to reject some of the basic, intuition-produced beliefs with which one began. To be sure, one could not abandon all of one’s intuitive beliefs and continue to engage in rationalist philosophy. One could have philosophical beliefs based solely on the authority of others, but this is not to participate in the practice of philosophy. Rationalist philosophy is a doxastic practice, a way of knowing, grounded in rational intuition that aims toward comprehensiveness and reﬂective equilibrium among one’s philosophical views. As Rawls writes, ‘‘if the scheme on the whole seems on reﬂection to clarify and to order our thoughts, and it if tends to reduce disagreements and to bring divergent convictions more in line, then it has done all that one may reasonably ask’’ (Rawls 1971, p. 53). There may be few bedrock philosophical beliefs for rationalist philosophy, but nevertheless there are many basic beliefs in the sense of noninferential ones delivered by intuition, beliefs subject to reappraisal and even rejection in an effort to attain wide reﬂective equilibrium.23 What is bedrock for rationalists, and lays fair claim to being foundational, is a belief in the virtue of rational intuition as a means of gaining knowledge and a recognition of the need for some epistemic standards of appraisal and inference. To give that up is to give up rationalism altogether.

44

Chapter 1

Objection 2 In a related strand of thought, Catherine Z. Elgin has argued against the epistemic value of rational intuition. She acknowledges that such intuitions are supposed to be about modal matters but denies that they have any reliability or even any access to a modal realm. She has two arguments. The ﬁrst is that ‘‘my intuitions may conﬂict with yours, with my earlier intuitions, even with my concurrent ones. And they may conﬂict with considered judgments and conﬁrmed theories. . . . Intuitions then are not intrinsically reliable’’ (Elgin 1996, p. 54). This is a very bad argument. Elgin seems to be assuming that unless intuitions are universal and infallible, they have no epistemic value at all. This principle, applied across the board, would yield a quite comprehensive skepticism. My perceptions, memories, and interpretations of physical phenomena may conﬂict with yours, with my earlier perceptions, memories, and interpretations, and even with my concurrent ones. Should we then conclude that perceptions, memories, and interpretations of physical phenomena are all intrinsically unreliable and should be dismissed as legitimate methods of acquiring knowledge? If so, then it seems that we have little or no knowledge of anything at all. If perceptions, memories, and so on should not be so dismissed, what makes them any different from rational intuition? Elgin seems to confuse indubitability with reliability, and in dismissing the ability of intuition to produce the former she infers the latter. Elgin’s second strain of argument concerns alethic modality. Elgin agrees with the common view, defended here, that rational intuition generates beliefs about putative necessities. However, she denies that there are necessary truths. Therefore intuition has no epistemic role to play. Her argument is that a genuine necessity would be ‘‘undeniable’’ and yet ‘‘we can almost always ﬁnd a scenario in which a seemingly necessary truth looks merely possible’’ (Elgin 1996, p. 57). She thinks that to use modal locutions is to do no more than distinguish what is conceded and what is in question given some particular context of inquiry. As contexts of inquiry change, so do the propositions listed under the category of ‘‘necessary.’’ Therefore there are no authentically necessary truths that are so in all contexts of inquiry. First, I am less than sanguine that it is so easy to describe a plausible ‘‘context of inquiry’’ in which either ðp ! pÞ or j ðp ! pÞ looks merely possible. One could always assert that they aren’t necessary truths, but it hard to see how they could turn out to be false after some form of inquiry. How might one argue for their falsity without implicitly assuming their truth? Second, there may be a self-refutation problem

Intuition and Philosophical Propositions

45

lurking nearby. Elgin claims that which propositions are necessarily true is relativized to contexts of inquiry. Presumably, then, the modality of the proposition that ‘‘which propositions are necessarily true is relativized to contexts of inquiry’’ is itself relativized to contexts of inquiry. This means that there are contexts in which ‘‘which propositions are necessarily true is relativized to contexts of inquiry’’ might be false, that is, contexts that sanction possible worlds in which it is false. In such a context and world, the modality of propositions is invariant across contexts, something that doesn’t look good for Elgin’s modal relativism. The correct relationship between modality and relativism is difﬁcult and subtle, and will be explored in detail in chapter 3. Particularly strange about Elgin’s arguments is her support of the value of reﬂective equilibrium as the best way to tend to our beliefs. Certainly part of the appeal of reﬂective equilibrium is the recognition that however we have arrived at our beliefs and no matter what sorts of propositions they are about, they must be brought into some form of coherent system as far as possible. Nothing about this procedure implies that beliefs arrived at through rational intuition are ineligible for inclusion in our best picture of the world, or that intuition-based beliefs are incompatible with reﬂective equilibrium. Moreover, a system of beliefs in wide reﬂective equilibrium could well include the view that various propositions are necessarily true while also including the recognition that others might deny that there are any necessities. Believing that a proposition is necessarily true does not require also believing that everyone else agrees. Like Nielsen, Elgin seems tempted by the view that reﬂective equilibrium is at odds with foundationalism, and additionally, she is eager to saddle foundationalism with the traditional bugbears of indubitability and certainty. As we have seen, neither is required for an account of philosophical knowledge. Objection 3 It may further be objected that I have not shown that rationalism produces any knowledge at all. A skeptic might object that all I have done is describe a closed, self-vindicating process by which philosophers satisfy themselves. Philosophers decide that how things seem to them is how they are, and ‘‘investigate’’ an unreal world of abstracta and duties and oughts. As Nietzsche complained about Christianity: Neither morality nor religion come into contact with reality at any point. Nothing but imaginary causes (‘‘God,’’ ‘‘soul,’’ ‘‘ego,’’ ‘‘spirit,’’ ‘‘free will’’— or ‘‘unfree will’’): nothing but imaginary effects (‘‘sin,’’ ‘‘redemption,’’ ‘‘grace,’’

46

Chapter 1

‘‘punishment,’’ ‘‘forgiveness of sins’’). A trafﬁc between imaginary beings (‘‘God,’’ ‘‘spirits,’’ ‘‘souls’’); an imaginary natural science (anthropocentric, complete lack of the concept of natural causes); an imaginary psychology . . . an imaginary teleology . . .—This purely ﬁctitious world is distinguished from the world of dreams, very much to its disadvantage, by the fact that the latter mirrors actuality, while the former falsiﬁes, disvalues and denies actuality. (Nietzsche 1895 [1968], §15)

Rationalist philosophy is no more than a language game with implicit rules and acceptable moves and strategies, and it has no more connection to reality than Rorschach testing—which is to say, along with Nietzsche and Ayer, that the only truths intuition-driven rationalism reveals are psychological ones about those who participate. Skepticism is at the same time the easiest objection to make to an epistemology and the hardest to put to bed. One way to ﬂesh out the skeptical response above is to claim that the method of rational intuition and subsequent reﬂection cannot be shown to be more trustworthy than competitor methods that deliver different results about the same propositions. There are philosophical propositions, but we can’t know them because we don’t know which belief-acquiring method is the best one to use. This is a difﬁcult and serious objection that I will consider in the following chapter. There I will argue that a skeptical response to such a concern leads to a paradox and so is an untenable approach. Another way to endorse skepticism is to argue that there are no philosophical propositions at all (nothing but imaginary causes, effects, and beings), and so rationalist philosophy is much ado about nothing. This is of course the approach of the naturalists, one that has been touched on in this chapter and will receive a fuller treatment in chapter 4. With these promissory notes, I am temporarily setting aside the skeptic’s contention that rationalism is not a way of acquiring knowledge and proceeding with the provisional assumption that there are philosophical propositions and that rational intuition is a means of coming to know them. Objection 4 It might also be objected that in presenting a meta-epistemology of rationalist philosophical practice I have failed to provide an adequate treatment—indeed any treatment—of the familiar terms of epistemology. That is, there has been no discussion of the nature of justiﬁcation (e.g., whether it is to be understood internally as a capacity to give reasons for one’s beliefs or externally as a reliabilist link with truth), no analysis of knowledge (e.g., solving the Gettier problem, addressing Car-

Intuition and Philosophical Propositions

47

tesian skepticism), no discussion of the validity of closure principles for knowledge, and so on. To this charge I plead nolo contendere. These tasks are ones appropriate to ﬁrst-order epistemology. While it is the case that a fully ﬂedged epistemology needs to be able to account for itself and needs to be able to accommodate knowledge of propositions about epistemology, it is not so clear that a metaphilosophical assessment of methodology must ﬁrst settle every issue in epistemology. For philosophers to debate each other at all, there must be some shared, neutral procedure by which they can do so. If one insists that any explanation of this procedure incorporate speciﬁc commitments to, say, contextualism, internalism, virtue theory, proper functionality, or whatever, is to preclude the possibility of meaningful interaction among opposing camps on these issues. If one presupposes that proper functionality is integral to the very concept of rationalist philosophy, then of course it becomes very difﬁcult to offer intuition-based arguments against the view. Any such argument would be dismissed as incompatible with the very nature of a priori reason. It may turn out that substantive epistemological theses must be true for rational intuition to be a method of gaining knowledge at all. In fact I have argued that a version of foundationalism must be correct for it to be a successful methodology. However, it requires argument to show that rationalism is possible only under a particular conception of knowledge or justiﬁcation; it may be the case that the only robust epistemic claims that must be true for rationalism to be successful are ones concerning the source and structure of a priori knowledge. So pending arguments contrariwise, I prefer to be as neutral as possible on the nature of knowledge, justiﬁcation, and so on. Furthermore, I do not see that the chief theses of this book—in particular the problem for philosophical knowledge posed in chapter 2 or the solution to it presented in chapter 3— implicitly entail any speciﬁc view about the nature of knowledge or of justiﬁcation.

2

A Problem for Philosophical Knowledge

I have argued that the familiar way that most contemporary philosophers acquire beliefs about philosophical propositions is through the logical analysis of rational intuitions, intuitions that must be taken as foundational, basic data. In this chapter I will show that there are other belief-acquiring methods besides rational intuition that generate beliefs about philosophical propositions. At ﬁrst blush it may seem odd to present evidence for something so obviously true. But philosophers can be surprisingly blinkered. For example, after praising the widespread belief-acquiring methods of observation and deduction, one well-known philosopher promptly concludes that ‘‘[there is] nearly universal agreement about which epistemic principles are true.’’1 This person is a good representative of the insular armchair theorist who spends too much time in the cramped, hermetic philosophers’ world where snow is always white, no one can ﬁgure out that Hesperus is Phosphorus, the present King of France is forever bald, Goldbach’s conjecture stays unproven, and these topics are endlessly fascinating. Out in the wider world there are many, extremely diverse ways to gain beliefs, with partisans who insist that their favored method yields justiﬁed beliefs or even knowledge and detractors who dismiss the same method as sophistry or magic. It is risible to think that there is anything like universal agreement about which epistemic principles are true. Furthermore, the alternative belief-acquiring methods to be discussed presently are very similar to rational intuition along these dimensions: they supposedly produce basic, noninferential beliefs; the basic beliefs generated are not the ﬁnal word, but must be rationally evaluated for consistency, explanatory cohesiveness, and the like in order to achieve reﬂective equilibrium; and they are taken seriously by a great many people. I mention this last because I am not simply inventing bizarre philosophers’ scenarios which might be airily dismissed, but relating empirical

50

Chapter 2

facts about how people really come to believe what they do in the real world. The problem is that these alternative methods frequently produce beliefs that are inconsistent with those sanctioned by rationalist philosophy, and so rationalists bear the onus of defending rational intuition as the most preferable method. The two alternative belief-acquiring methods to be discussed are Christian revelation and the ritual use of hallucinogens. They will be treated in sections (1) and (2) respectively. For each of these methods I will be at pains to show how the beliefs they deliver are regarded as foundational by their adherents and that they are used as data which must then be systematized and brought into reﬂective equilibrium. In addition I will argue that the doxastic results of both revelation and the use of hallucinogens are inconsistent with the beliefs most rationalist philosophers hold to result from the application of rational intuition. In section (3) I will examine four arguments designed to show the relative superiority of rational intuition over the competitors of revelation and psychoactive drugs. I will conclude that none of these arguments is successful and that there is an irreconcilable conﬂict among the outputs of different epistemic methods that produce beliefs about philosophical propositions. There seem to be only three possible responses to this conﬂict: skepticism, nihilism, and relativism. These will be canvassed at the end of the chapter. Christian Revelation In this section I will attempt to show that, according to mainstream traditions in Christian theology and scholarship, (1) revelation is an epistemic method that yields beliefs about a class of philosophical propositions; (2) the beliefs generated by revelation are foundational ones, upon which reason then operates to produce a more elaborate theology; and (3) revelation and rational intuition are apt to produce inconsistent results—that is, one method might produce the belief that p, whereas the other might produce the belief that not-p. Not only is there a substantial tradition within Christian scholarship to this conclusion, but there are many mainstream Christian views that a majority of rationalist philosophers reject. Revelation Is an Epistemic Method of Acquiring Beliefs about Philosophical Propositions It is an easy matter to show that, historically understood, ‘‘revelation’’ denotes an epistemic method of coming to acquire certain beliefs. There

A Problem for Philosophical Knowledge

51

are a good many controversies about Christian revelation, even for those sympathetic to the concept. One is whether God revealed various truths to some special people in the Middle East about 2,000 years ago and has pretty much kept quiet since, or whether all of us have an innate capacity to sense the divine and just need to tap into this ability. John Calvin defended the latter view. Calvin argued that revelation is a universally available method of gaining beliefs about the divine; whatever training or education might be required to interpret or understand the dictates of revelation, all nondefective persons have a sense of the divine (sensus divinitatis) just as all nondefective persons have sight, hearing, and the like. Indeed, Calvin found the evidence for the sensus divinitatis overwhelming, writing, ‘‘that there exists in the human minds and indeed by natural instinct, some sense of Deity, we hold to be beyond dispute, since God himself, to prevent any man from pretending ignorance, has endued all men with some idea of his Godhead, the memory of which he constantly renews and occasionally enlarges, that all to a man [are] aware that there is a God’’ (Calvin 1559 [1845], 3.1). Calvin’s argument for the universality of the sensus divinitatis is the widespread acceptance of some sort of religion in all cultures and locations. Even the fact of pagans, heretics, and animists Calvin took as a sign of a crude use of the sensus divinitatis groping toward the truth of the Christian god. This is, of course, a terrible argument. In the ﬁrst place, animists and pagans are unlikely to be impressed by a claim that they have ‘‘really’’ been subconsciously reaching out for Yahweh. In the second place, superstitions of all kinds are spread throughout the world, and this is no evidence that there are hidden nonphysical causal forces at work in our lives. The view that the alignment of the stars controls or predicts our destiny has been widely believed since human beings ﬁrst walked out of the jungles, and yet it is entirely false. Cultures all over the world are replete with similar systematic errors; everywhere and everywhen one ﬁnds the popularity of the gambler’s fallacy, a belief in telepathy and extrasensory perception, and the view that cold weather causes colds.2 The apparent fact that there is a universal tendency to form religious beliefs does not show that there is a truth-conducive human faculty that allows us to apprehend the divine, any more than our propensity to accept the gambler’s fallacy shows that there is a reliable human faculty to estimate probabilities, and mutatis mutandis for the other examples above. There might well be a sensus divinitatis, but Calvin’s argument doesn’t show it. More recently Richard Swinburne (1992)3 has defended a position more akin to the view that God revealed some things to a select crowd a

52

Chapter 2

long time ago, and theology’s task since then is a hermeneutical one, to decipher and interpret the writings of God’s legitimate prophets. Swinburne’s argument for this view is based on his a priori assessment of the character and motivations of God. He ﬁnds it a priori likely that, if there is a God, he will share knowledge by revelation, that this revelation will be attended by miracles, that it will concern human imperfections and how we have wronged God thereby, that it will show the need for our atonement, and that it will provide information about how God became incarnate and how we can use this fact to plead atonement.3 Furthermore, Swinburne thinks it a priori plausible that the content of revelation will be somewhat ambiguous and opaque—God doesn’t want salvation to be too easy; we ought to have to work at it (1992, pp. 94–95).4 Thus we need the continuing church and theological traditions to help us understand the original revelation. In short, Swinburne’s theology is carefully constructed and tailored so that the actual history of Christianity turns out to be more or less how things had to go. He then concludes that there is ‘‘only one serious candidate for having a body of doctrine which is to be believed on the grounds that it is revealed, and that is the Christian revelation’’ (p. 95). How the Whale Got His Throat, How the Camel Got His Hump, How the Rhinoceros Got His Skin, How the Leopard Got His Spots, and How Christianity Became a World Religion—God wanted it exactly that way, wrote Swinburne: just so! Whether Calvin or Swinburne is right about the nature of revelation, they are both well within the central teaching of Christianity that God revealed something to someone, at some point, and this revelation was the method by which the recipient of revelation came to acquire knowledge about the content of the revelation. What truths are supposedly apprehended through revelation? Different writers make a variety of claims. Cardinal Newman noted that ‘‘The Catholic Church claims, not only to judge infallibly on religious questions, but to animadvert on opinions in secular matters which bear upon religion, on matters of philosophy, of science of literature, of history, and it demands our submission to her claim’’ (Newman 1892, p. 257). The Catholic Church’s pretensions to infallibility on such a wide range of topics are ultimately grounded in its contention to be in possession of God’s revelation, and in its claim to be the historical (and most trustworthy) continuer of the original Christian community. However, not all theologians make such sweeping claims on behalf of revealed knowledge. Primarily the propositions supposedly delivered by revelation are about the philosophy of mind (e.g., that human beings have souls, that these souls have certain

A Problem for Philosophical Knowledge

53

properties such as location, that we survive our deaths via our souls), ethics (e.g., that we have certain duties of charity or forgiveness, that we are mired in original sin, that the death of an innocent person can atone for our moral failings), metaphysics (e.g., that there is a Heaven and a Hell, that there are invisible spirits that affect our lives such as angels and demons), and the nature of the divine (e.g., that there is a God, that God loves us, that God is triune). In short, revelation produces beliefs about philosophical propositions, many of the same ones that secular philosophers assess through the methods of rational intuition and logical argument. Revelation-generated Beliefs Are Foundational and Basic Historically, revelation has been understood as a method that produces epistemically basic, noninferential beliefs. Etienne Gilson noted as much in his ﬁne study of the history of revelation in the Middle Ages, writing, ‘‘To such men as St. Anselm and St. Augustine, religious faith is there, objectively deﬁned in its contents by Revelation, as a reality wholly independent from their own personal preferences. . . . just as scientists accept observable facts as the very stuff which they have to understand, those religious geniuses accept the data of Revelation as the given facts with they have to understand’’ (Gilson 1938, p. 32). Contemporary theologians and philosophers of religion defend this traditional view. Richard Niebuhr likens revelation to Cartesian ﬁrst principles, writing, ‘‘in dealing with revelation we refer to something in our history to which we always return as containing our ﬁrst certainty. It is our ‘cogito ergo sum’ ’’ (Niebuhr 1962, p. 140). He goes on to assert that the content of revelation is ‘‘self-evidencing.’’ Nothing else is epistemically prior to the data delivered by the method of revelation, as Karl Jaspers also maintains. He writes that the possibility of revelation cannot be deduced a priori, and that revelation ‘‘precedes all reasoning’’ (Jaspers 1967, p. 27): ‘‘The truth of revelation is established only by revelation itself’’ (ibid.). Jaspers does not think this is evidence of vicious circularity so much as a statement of the foundational, self-justifying nature of religious revelation. Among contemporary philosophers of religion, Plantinga and Alston are the best-known defenders of the view that beliefs produced by revelation are noninferential. Alston defends the idea of a beliefacquiring method he calls ‘‘Christian Mystical Practice’’ (CMP). CMP is a matter of forming perceptual beliefs about God in a way that is concordant with the chief traditions of mainstream Christianity. Alston stresses conformity with the Christian tradition because he, like many

54

Chapter 2

other Christian writers, is concerned with how one can exclude the exotic teachings of, say, James Jones or the Heaven’s Gate cult who claim to have received a revelation from God, while keeping the teachings of Jesus or Paul who also claimed to be in possession of revealed truth. Alston gives a lengthy, general defense of the need for one’s beliefs to ﬁt into socially established doxastic practices, but this is not relevant to the current discussion. Crucial for our purposes is that Alston defends the existence of a mystical experience of God that gives rise to propositional beliefs (that God is doing X or desires Y); it is this experience that I have been including under the general umbrella of ‘‘revelation.’’5 In addition, Alston’s CMP generates what he calls properly basic beliefs (Alston 1991, p. 196). Plantinga is even more explicit and detailed in his assertion that the beliefs formed by revelation (his language: the Internal Instigation of the Holy Spirit) are noninferential. He writes: Christian belief is basic; furthermore, Christian belief is properly basic, where the propriety in question embraces all three of the epistemic virtues we are considering. On the model, the believer is justiﬁed in accepting these beliefs in the basic way and rational (both internally and externally) in so doing; still further, the beliefs can have warrant, enough warrant for knowledge, when they are accepted in that basic way. (Plantinga 2000, p. 259)6

There are various ways that Plantinga’s treatment of revelation differs from Alston’s, and from those of other Christian writers like Swinburne, and he is careful to lay out all these distinctions. Everyone has his or her own theory about why revelation-based beliefs are basic, and why it is reasonable to hold them. These differences are irrelevant here, though. What is striking among all of these writers is their considerable agreement: revelation is a method of gaining beliefs about a class of philosophical propositions, these beliefs are noninferential and basic, and they are often (details vary) epistemically warranted, justiﬁed, reasonable to maintain, and so on. Reason Is Used to Develop Foundational Beliefs into a System It is to be expected that philosophers of religion like Alston, Plantinga, and Swinburne use reason and logical argument to build on the basic beliefs delivered by revelation and develop a more comprehensive theistic metaphysics and axiology. In doing so, they fall squarely within the mainstream of Christian theology. Gilson, for example, argues that the theological tradition since at least Augustine is that given the idea of

A Problem for Philosophical Knowledge

55

revelation as providing basic knowledge, reason must then operate to provide understanding of the divine. In Augustine’s formulation, ‘‘understanding is the reward of faith. Therefore seek not to understand that thou mayest believe, but believe that thou mayest understand’’ (Gilson 1938, p. 19). Niebuhr expands this, claiming that our moral knowledge too is the result of taking the data of revelation and interpreting these data through reason (Niebuhr 1962, p. 171). In his encyclical Fides et ratio (1998), Pope John Paul II defends the use the reason not only to ‘‘successfully intuit and formulate the ﬁrst principles of being’’ (§4) but to take the deliverances of revelation and develop them into a systematic philosophical theology. He writes, ‘‘Revelation . . . introduces into our history a universal and ultimate truth which stirs the human mind to ceaseless effort; indeed, it impels reason continually to extend the range of its knowledge until it senses that it has done all in its power, leaving no stone unturned’’ (§14). Once revelation has delivered its data, the task of fundamental theology is to integrate this information with the insights of pure a priori philosophy and develop a more comprehensive, systematic view of ‘‘the meaning and ultimate foundation of human, personal and social existence’’ (§5). The views of Augustine, Niebuhr, and John Paul II on the relationship between revelation and subsequent reasoning sound very similar to the relationship between intuition and rational reﬂection. In traditional intuition-driven philosophy, we take our intuitions as data and then subject them to critical scrutiny, testing them for logical consistency, coherence, explanatory connectedness, and so on. Our intuitions are tested against one another, and their logical entailments drawn out and presented as evidence. Weaker, less well-rooted intuitions can be overruled, and ultimately eradicated, when one squarely faces the implications of one’s deeper, more ﬁrmly held intuitions. This is the familiar procedure of example and counterexample. Augustine, Niebuhr, and John Paul II seem to be advocating a parallel procedure for the dictates of faith or revelation. Instead of the method of intuition delivering foundational data (albeit fallible and subject to review by reason), it is the method of revelation that produces foundational beliefs which are then interpreted and made systematic by reason and argument. The Conﬂict Thesis In this section I will argue that the methods of revelation and rational intuition come into conﬂict; that is, they produce inconsistent results. My argument will be twofold. First, I will argue that there is a well-

56

Chapter 2

established tradition in Christianity (although not the only tradition, as we will see) according to which revelation and natural reason contradict each other. Second, I will argue that a strong majority of contemporary secular philosophers—those who rely on rational intuition—disagree with many of the revelation-based teachings of the largest Christian denomination, Catholicism. These are historical, inductive arguments. Some readers of this book may be dissatisﬁed that I do not offer a more rigorous, deductive proof that intuition and revelation conﬂict. I wish that I could; unfortunately such a thing is not possible. Christianity is an old, complex, and multifaceted religion, rife with internal debates among theologians. Likewise philosophers rarely (if ever) speak with one voice about what is self-evident by the lights of rational intuition. Showing that there is an unequivocal, deﬁnite inconsistency among the beliefs generated by Christian revelation and those produced by philosophical intuition is therefore as easy as nailing Jell-O to the wall. So we will have to rest content with an historical argument. There are two main strains of thought in the history of Christianity with respect to reason and revelation. The ﬁrst is what we might call the separate spheres doctrine. According to this view, rational intuition and philosophical reasoning alone produce beliefs about different propositions than those produced by revelation. Intuition and revelation don’t come into conﬂict because they generate beliefs about separate subject matters, in the same way that the ear does not disagree with the mouth about the taste of Bordeaux. If revelation yields a belief that p, p is inaccessible by rational intuition; intuition has nothing to say about it. The second strain of thought is the conﬂict thesis. According to the conﬂict thesis, rational intuition and revelation can and do produce inconsistent results, with one generating a belief that p and the other not-p. Different Christian writers then have a variety of ways of addressing this problem. Perhaps the earliest Christian writer to explicitly endorse the conﬂict thesis was the early third-century Roman and Church Father Tertullian. In his De anima, Tertullian considers the thesis that there is a supreme principle of life and intelligence in the soul, a faculty that directs action and causes motion (Tertullian c. 210 [1950], ch. 15). Tertullian notes that this vitalist claim is rejected by empirically informed medical doctors (he names Dicaerchus, Andreas, and Asclepiades) who argue that ﬂies, wasps, and locusts will live after their heads have been removed, and that you can cut out the hearts of she-goats, tortoises, and eels and they will still move. Since the seat of the soul was supposed to be in either the heart or the head, the removal of those items should

A Problem for Philosophical Knowledge

57

also have removed the supreme principle of action. The physicians reason that ‘‘there is no supreme principle, because, if there were one, life could not continue in the soul after the seat of that principle had been removed.’’ Tertullian rejects this argument because ‘‘we know from Revelation that there is a directive faculty in the soul which itself resides in a special place in the body.’’ In other words, the data of revelation trump whatever interpretations of physical phenomena conﬂict with it. So far Tertullian has argued that empirical methods, insofar as they conﬂict with the data of revelation, must be in error. He also thinks that the arguments of the rationalist philosophers can come into conﬂict with revelation and that they too must be rejected for this reason. He cites Plato’s claim (Timaeus 69d) that the soul is in the head, Heraclitus’ view that the principal part of the soul can be stirred from without, Moschion’s idea that the soul is suffused throughout the whole body, Xenocrates’ position that that the soul is in the crown of the head, and Epicurus’ claim that the soul is in the structure of the breast as examples of erroneous philosophical positions about the soul. They are mistaken for one reason: they contradict revelation, which, according to Tertullian, shows that the soul is in the heart (although he isn’t above listing the philosophers, like Protagoras and Chrysippus, who agree with him). Here we have a case where rationalist methods of belief acquisition yield p and revelation, at least according to one of the church fathers, yields not-p. Tertullian does not argue that he himself has employed the method of revelation to arrive at beliefs about the soul. That is, God did not personally reveal to Tertullian the location of the soul. Rather Tertullian’s beliefs were inferential ones, arrived at by accepting the testimony of biblical authors. He cites Proverbs, Matthew, Psalms, Romans, and John, as well as the apocryphal The Wisdom of Solomon in this connection. It is the authors of these books who acquired noninferential beliefs about the soul through the method of revelation; Tertullian’s own method of gaining beliefs about the soul’s location in the heart was to accept the authority of these biblical writers. Whereas Tertullian argued that revelation and intuition led to inconsistent results in the philosophy of mind, Martin Luther thought that they produced conﬂicting results with respect to propositions about divinity and ethics. In his ‘‘Disputatio de homine’’ (1536), Luther argues that reason is a ﬁne thing ‘‘in comparison with other things of his life, the best and something divine’’ (§4), but that it is useless in theological matters. In §11 he writes that ‘‘if philosophy or reason itself is compared

58

Chapter 2

with theology, it will appear that we know almost nothing about man.’’ Later he offers the opinions that Aristotle knows nothing about ‘‘theological man’’ and that it is impious and in opposition to theology to suppose that reason can aspire to the knowledge of God. These passages are consistent with the separate spheres approach. However, it is clear from Luther’s earlier work that he endorses the conﬂict thesis, rejecting the medieval reliance on Aristotelian reason as the path to the correct understanding of morality and the divine. With respect to morality, in ‘‘Disputatio contra scholasticam theologiam’’ (1517) Luther writes that ‘‘Virtually the entire Ethics of Aristotle is the worst enemy of grace. This in opposition to the scholastics’’ (§41). Furthermore, Luther thought that the common doctrine that Aristotle’s views on happiness are congruent with Catholic teachings was mistaken (§42). Again, it is not the case that faith, revelation, grace, or the Internal Instigation of the Holy Spirit are going to yield knowledge about propositions that are inaccessible by other means, but rather that religious methods are the only trustworthy means of gaining such knowledge. Aristotelian reason might lead to beliefs about virtue or the value of eudiamonia, but these beliefs are apt to come into conﬂict with the data of revelation and so, according to Luther, must be mistaken. In the case of the well-nigh universal view among the medievals that knowledge of logic and a priori reasoning are essential prerequisites to the proper study of theology and the acquisition of theological knowledge, Luther writes, ‘‘it is an error to say that no man can become a theologian without Aristotle. This in opposition to common opinion. Indeed, no one can become a theologian unless he becomes one without Aristotle’’ (§§43–44). So here too Luther supports the conﬂict thesis—a priori reason is not the proper basis for beliefs about the nature of God and issues of divinity. That methodology must be rejected if one is to become a theologian, not because theological beliefs won’t be produced by reason, but because false ones are prone to be generated. In fact, Luther, unlike the Catholic view from Aquinas on, seemed to think that reason has no use at all in theological matters, writing, ‘‘in vain does one fashion a logic of faith. . . . If a syllogistic form of reasoning holds in divine matters, then the doctrine of the Trinity is demonstrable and not the object of faith’’ (§§46, 49). Since Luther thought that the truth of the doctrine of the Trinity was a matter of revealed knowledge alone, he rejected a ‘‘logic of faith’’ as a mistaken, wrongheaded enterprise. Luther’s contemporary Calvin held a similar view. While Calvin’s discussion of religious epistemology is mostly concerned with establish-

A Problem for Philosophical Knowledge

59

ing the presence and value of the sensus divinitatis, as discussed earlier, he too seems to accept the conﬂict thesis. In chapter 5 of Institutes of the Christian Religion, he considers the Aristotelian view of the soul as well as the Stoic and Epicurean assessments of natural theology and ﬁnds them all wanting. Their problems were not internal infelicities or conﬂicts with the light of reason, but that these views were inconsistent with revelation-derived Christian teaching. Calvin couldn’t abide the idea that the methods of rationalist philosophy could have epistemic priority over the sensus divinitatis, and so he dismissed them, writing, ‘‘the proper inference is, that the human mind, which thus errs in inquiring after God, is dull and blind in heavenly mysteries’’ (Calvin 1559 [1845], §12). This is not quite an endorsement of the separate spheres view. It is not that Calvin thinks that rational intuition fails to produce any results on the matter of the heavenly mysteries in the same way that the methods of chemistry fail to produce knowledge of good chess strategies. Calvin acknowledges that traditional philosophical inquiry will generate beliefs about the nature of the soul and the nature of God. It’s just that since philosophy hasn’t given the ‘‘right’’ answer—namely, Calvinist Christianity, arrived at through revelation—philosophical intuition must be dismissed as unreliable. More recently, Karl Barth (1936) is best known for defending the conﬂict thesis. Barth agrees that revelation presents us with insights, information, about certain propositions that contradict the ﬁndings of natural knowledge and rational intuition.7 Like Luther, Barth held philosophy in contempt, writing, ‘‘further, it is forced down my throat that the Dogmatic theologian [i.e. Barth himself] is under the obligation to ‘justify’ himself in his utterances before philosophy. To that my answer is likewise, No. . . . It cannot be otherwise than that Dogmatics runs counter to every philosophy no matter what form it may have assumed.’’8 There is but one true revelation according to Barth, that of Jesus of Nazareth, and it shines with a glittering self-evidence. The nature of God, virtue, and redemption contained within this revelation is beyond the ken of human reason. Try as we might to understand through reason alone the nature of justice, fairness, or whether there is an afterlife, we are forever stymied in our attempts because the truth of these matters is transcendent and available only through revelation. As with Calvin, when rational intuition fails to yield the revelationsanctioned answer, it is to be dismissed. Not only Protestants support the conﬂict thesis. In the nineteenth century, John Henry Newman, a Cardinal of the Catholic Church, also

60

Chapter 2

endorsed the conﬂict thesis. Newman wrote that he had no problem accepting the Doctrine of Transubstantiation ever since he started to believe that the Roman Catholic Church was the oracle of God and that the church had declared this doctrine to be part of the original revelation. Newman was well aware of prevailing philosophical opinion that transubstantiation was impossible, and he dismissed philosophical treatments of substance, coming awfully close to Tertullian’s ‘‘credo, quia absurdum est.’’ He writes, ‘‘what do I know of substance or matter? Just as much as the greatest philosophers, and that is nothing at all’’ (Newman 1892, pp. 239–240). Thus Newman acknowledges that philosophical intuition and revelation yield conﬂicting results concerning transubstantiation, and like Tertullian, Luther, Calvin, and Barth, he concludes so much the worse for philosophy. Whereof revelation has spoken, thereof philosophy must remain silent. It is this wholesale dismissal of rational intuition that troubled Locke and led him to propose a modiﬁed form of the separate spheres thesis. In chapter 18 of An Essay Concerning Human Understanding, Locke argues that there are propositions about which we might form beliefs as the result of either rational intuition or revelation. In such cases, Locke avers, the revelation could at best conﬁrm what we know through intuition. The certainty derived from clear intuition is always greater than that provided by revelation, and Locke holds that it would be in vain to urge as a matter of faith belief in something that manifestly contradicts our clear intuitions (Locke 1690 [1974], 18.5). In his assessment of Locke, Leibniz in New Essays Concerning Human Understanding is also impressed by apparent conﬂicts between philosophical principles known through intuition and claims of scripture based on revelation. He writes: ‘‘suppose that on one side we ﬁnd the literal sense of a text of Holy Scripture, and on the other a great appearance of logical impossibility or at least an admitted physical impossibility, is it more reasonable to deny the literal sense or the philosophical principle? It is certain that there are [such] passages’’ (Leibniz 1696 [1949], p. 589). Leibniz thought it uncontroversial that there are results of revelation (as seen in Scripture) and results of rational intuition that conﬂict with each other. Despite Locke’s acknowledgment of the conﬂict between intuition and revelation, he thought that there are limits to what might be known though philosophical methods, and that there is a proper place for the method of revelation. Locke claimed that there are matters ‘‘above reason,’’ such as the propositions that angels rebelled against God, or that

A Problem for Philosophical Knowledge

61

the dead shall rise and live again, and that these could be known only through revelation. So Locke thought that when intuition and revelation could pronounce on the same proposition, they were apt to conﬂict. Yet there remained a separate class of propositions that can be known by revelation alone. A recent defense of Locke can be found in Swinburne. Like Locke, Swinburne states that the methods of revelation and philosophical intuition can yield beliefs about the same subject matter; in particular he mentions ‘‘general moral truths’’ and ‘‘evidence that there is a God’’ (Swinburne 1992, p. 72). If there is a conﬂict, Swinburne argues, then we must use intuition as a test for when a putative revelation is genuine. He writes that the content of the supposed revelation ‘‘must, as far as we can judge by independent criteria, be true’’ (ibid., p. 86). Otherwise, we are to dismiss the claimed revelation as nonveridical. As an example, Swinburne considers a hypothetical prophet who claims to have certain moral knowledge as a result of revelation. Swinburne holds that if a prophet’s moral teachings are clearly wrong, as shown by the method of rational intuition, then we can dismiss the prophet as a false one not genuinely in receipt of divine revelation. Swinburne concurs with Locke that in the case of conﬂict, rational intuition is the ﬁnal arbiter. But he steps away from Locke (and Leibniz) in holding that the method of revelation is an infallible method of acquiring beliefs. If you get a false belief from revelation, then, well, it wasn’t really a revelation after all; you’re just a false prophet. This all-ornothing approach to claims of revelation may not be good news for Christianity. For example, while Bertrand Russell enunciated many reasons as to why he was not a Christian, one of these was moral defects in the character of Jesus, particularly Jesus’ advocacy of Hell and eternal damnation. Russell thought that rational intuition showed the notion of everlasting punishment to be morally pernicious and a good reason to reject Jesus as a moral paragon (Russell 1927). Following Swinburne’s criterion for evaluating prophets, Russell possessed an excellent reason to reject Jesus as a false prophet, one who did not actually have moral knowledge as the result of revelation. In any case, Swinburne also agrees with Locke that there are propositions that cannot be known through rational intuition, but instead can be known only through revelation. For example, we can come to know the nature of atonement and the mechanism by which our sins might be forgiven through the unjust killing of Jesus only by revelation. He writes, ‘‘we need historical information about how God provided an atonement,

62

Chapter 2

and practical information about how we can plead it. It is hard to see how we might get the historical information without God, either himself or another, telling us what was happening’’ (Swinburne 1992, p. 73). As with Locke, this is a modiﬁed separate spheres approach: the beliefacquiring methods of rational intuition and revelation can overlap and produce beliefs about the same propositions, but at the same time there are propositions that only revelation can provide a basis for believing.9 Evidence of the Truth of the Conﬂict Thesis So far I have argued that there is a substantial historical tradition, within Christianity, according to which the belief-acquiring methods of rational intuition and revelation yield inconsistent results. Even those, like Locke, who want to carve out a special niche for the method of revelation acknowledge that when they produce beliefs about the same proposition, revelation may yield a belief that p and intuition yield a belief that notp. Apart from these historical considerations, let us now look at what contemporary theologians believe on the basis of revelation and what contemporary philosophers believe on the basis of rational intuition and see whether they conﬂict. Of course, it is unlikely that there are many, if any, propositions that all contemporary theologians believe, even if we restrict ourselves to Christian theologians. The same is true in spades of philosophers. However, we can see whether there are philosophical propositions that, if not all secular philosophers accept, at least a sizable majority do, and whether a substantial class of theologians disagree about those same propositions. If there are such propositions, then this will provide some reason to think that the method of revelation tends to pull in one direction, and the method of intuition in another. The theological views I will focus on are the ofﬁcial views of the Roman Catholic Church. Catholicism is the largest Christian denomination, and tends to have more precisely codiﬁed doctrines than Protestant denominations whose historical roots are in a rejection of hierarchy and priestly authority. I am not suggesting that Catholic theology is correct, simply that it is the most prominent and most explicit about philosophical propositions, thus the easiest to examine. There are cases in which the revelation-derived beliefs of the Catholic Church come into conﬂict with, if not universal, at least very widespread views of contemporary philosophers who rely on rational intuition. The irony is that ever since Aquinas, the Catholic Church has emphasized the idea that theology and rationalist philosophy dovetail, with the latter aiding and enhancing the former.10 In fact this optimism is unwarranted.

A Problem for Philosophical Knowledge

63

The ﬁrst conﬂict between Catholic revelation-derived teachings and the views of most secular intuition-driven philosophers can be seen in matters of sexual and medical ethics, especially the morality of premarital sex, masturbation, birth control, voluntary sterilization, homosexuality, abortion, and euthanasia. The Catholic Church ﬁnds all of these things to be unequivocally immoral. Most secular philosophers do not. The Roman Curia’s doctrinal document on sexual ethics (Roman Curia 1975) is explicit that ‘‘Christian doctrine . . . states that every genital act must be within the framework of marriage’’ (§7). Thus premarital sex, even for the best of reasons, even when the couple intends to get married, is strictly forbidden. Not only premarital sex, but homosexual relations are also taboo. The Curia claims that ‘‘homosexual acts are intrinsically disordered’’ and ‘‘in Sacred Scripture they are condemned as a serious depravity’’ (§8). In Pope Paul VI’s encyclical Humanae vitae, he writes that ‘‘the Church . . . teaches that each and every marital act must of necessity retain its intrinsic relationship to the procreation of human life’’ (Paul IV 1968, §11). Paul VI states quite clearly that this teaching comes directly from God, who supposedly established a moral connection between sex and procreation. Indeed, he claims that ‘‘the teaching of the Church regarding the proper regulation of birth is a promulgation of the law of God Himself’’ (§20). Since God has revealed that sexual acts must contain the possibility of leading to new life (successful use of the rhythm method as a means of birth control and involuntary infertility are exceptions), masturbation and homosexuality are straightforwardly condemned under any circumstances. Ofﬁcial Catholic doctrine makes every episode of birth control, vasectomy, and tubal ligation immoral, if those actions are designed to prevent contraception.11 The doctrine of double effect would permit sterilization as a side effect of an otherwise necessary medical procedure, but that’s it. In addition, Catholics hold abortions to be wrong for a panoply of reasons, one of which is that it prevents a new life arising from coitus. Paul VI’s teaching in Humanae vitae is strongly endorsed by John Paul II. In his encyclical Evangelium vitae (John Paul II 1995) he not only rails against abortion and contraception (§13) but also against euthanasia (§15). Even cases of voluntary passive euthanasia, where a suffering or dying patient voluntarily requests to be removed from life-sustaining equipment or to have life-sustaining treatment withheld, are considered by John Paul II to be contrary to the moral law. Again, it is worth emphasizing that the moral views enunciated by Paul VI, John Paul II, and the Roman Curia are not the result of a priori reasoning, or the ferreting

64

Chapter 2

out of moral intuition through reﬂection and debate, but ultimately stem from divine revelation, as those authors explicitly acknowledge. It would be foolish to attempt to show that all intuition-based philosophy rejects these conclusions. There is probably no proposition that, once conceived, hasn’t been defended by some philosopher. Nevertheless, I do not know of a single contemporary secular philosopher who defends the idea that all premarital sex is seriously immoral or that birth control is unethical. Likewise, defenses of the immorality of masturbation are exceedingly scarce. The only philosophers who argue against homosexuality do so on prudential, not moral grounds, and even prudential criticisms (like Michael Levin’s) are rare. Abortion and euthanasia are more controversial, but it is a safe bet that a substantial majority of philosophers who rely on rational intuition in normative ethics think that quite a few cases of abortion and voluntary euthanasia are morally permissible. A second area of conﬂict concerns the doctrine of original sin. The Catechism of the Catholic Church (pt. 1, sec. 2, ch. 1, art. 1, para. 7) makes it abundantly clear that it is only through divine revelation that we can come to understand the nature of sin and recognize that it is something other than a developmental deﬁciency or psychological weakness. Furthermore, every person is infected with sin from birth, as the result of the disobedience of Adam and Eve in the Garden of Eden (as related in Genesis 3). Tempted by the serpent, they ate the fruit of the Tree of the Knowledge of Good and Evil, thus defying God’s orders otherwise, and offending him so profoundly that he punished all of their descendants. The Catechism admits that ‘‘the transmission of original sin is a mystery that we cannot fully understand. But we do know by Revelation that Adam had received original holiness and justice not for himself alone, but for all human nature’’ (pt. 1, sec. 2, ch. 1, art. 1, para. 7, §404). When Adam fell from grace, we all did. Thus every human being is inherently sinful—even newborn infants and saints—and all need redemption before God or else face damnation and eternal torment. For secular philosophers, original sin is more than a mystery that we just have to accept; it is absurd. The idea that a person can be fairly punished for the misdeeds of another, especially a hypothetical ancestor of a few thousand years ago, is anathema to every treatment of justice of which I am aware. St. Augustine was wracked with guilt over his youthful theft of some pears, raising this indiscretion up as the height of wickedness (Augustine 400 [1952], bk. 2, §§9–18). Secular philosophers these days generally ﬁnd this risible, but they would be positively stupe-

A Problem for Philosophical Knowledge

65

ﬁed at a suggestion that Augustine’s modern-day descendants (if any) ought to be arrested and prosecuted for pear-theft. The method of revelation thus yields a belief that p—every human being morally deserves punishment for the wrongdoing of Adam—and the method of rational intuition supports a belief that not-p. An additional key ethical conﬂict between the results of Christian revelation and that of rational intuition is with respect to eschatology. According to Catholic (and nearly all Christian) doctrine, Jesus of Nazareth was the sin-free son of God, whom God executed on the cross as a blood sacriﬁce to atone for the sins of humanity. (See, e.g., The Catechism of the Catholic Church, ch. 2, art. 4, para. 2.) Through the sacriﬁce of the innocent man, the sins of others (even the sins of future generations, as the Catechism makes explicit) are forgiven, and thereby justice is served. It is, of course, fairly puzzling why God couldn’t just forgive the sinners without ﬁrst assassinating his innocent child, as if even the Almighty has to obey some karmic law of balance. This aside, secular rationalists ﬁnd it ﬂatly ridiculous that justice can be done through the punishment of the innocent. Although there are circumstances under which act utilitarians are committed to the sacriﬁce of the innocent for the greater good, this fact has universally been taken to be an objection to act utilitarianism. Virtue theorists may consider a great personal sacriﬁce for the well-being of others to exemplify one of the virtues, but no virtue theory renders the sacriﬁce of some other innocent person as a virtuous act. The conviction and execution of the innocent is unanimously regarded as the very standard of injustice. Nietzsche (1895 [1968]) well expresses the rationalist assessment of the cruciﬁxion of Jesus: ‘‘the guilt sacriﬁce, and that in its most repulsive, barbaric form, the sacriﬁce of the innocent man for the sins of the guilty! What atrocious paganism!’’ (The Antichrist, §41). Ethics is not the only area of conﬂict. Catholic teachings also conﬂict with the majority view in the philosophy of mind. According to The Catechism of the Catholic Church (pt. 1, sec. 2, ch. 1, art. 1, para. 6, §366), human beings each have a soul which animates their bodies and ‘‘is immortal: it does not perish when it separates from the body at death, and it will be reunited with the body at the ﬁnal Resurrection.’’ This is not just a Catholic view either, as some of the ethical positions discussed above may be. Nearly every Christian denomination, perhaps all of them, includes the view that persons are partly or wholly composed of a nonphysical, spiritual entity that survives the death of the body and lives forever.

66

Chapter 2

There are many competing views about the nature of the mind in contemporary philosophy—that the mental supervenes on the physical, that mental properties are emergent from physical properties, various identity theories, functionalism, eliminativism, and so on. One view that is almost universally rejected is substance dualism. Even philosophers like Frank Jackson or David Chalmers who are skeptical about the prospects for a completely satisfactory materialist explanation of the mental don’t think that our mental lives reside in an immortal soul. To be sure, there are some secular philosophers who are still substance dualists, but their ranks have rapidly shrunk since Descartes.12 And even they don’t argue for immoral souls destined for a future of reuniﬁcation with resurrected bodies. Jaegwon Kim remarks that ‘‘the idea of minds as souls or spirits, as entities or objects of a special kind, has never gained a foothold in a serious scientiﬁc study of the mind and has also gradually disappeared from philosophical discussions of mentality’’ (Kim 1996, p. 3). He states that there is an almost complete consensus among philosophers in rejecting the existence of an immortal spiritual soul (p. 4). The existence of souls may be an even more explicit example in support of the conﬂict thesis. The method of revelation supports a belief in p—each human being is at least partly composed of a spiritual, immortal soul— and the method of rational intuition (it is overwhelmingly accepted) supports a belief that not-p. In fact, I am going to go out on a bit of a limb and claim that the majority of philosophers who reject revelation as a legitimate beliefacquiring method, preferring to rely on rational intuition alone, also do not believe that there is a God. It is generally considered that Hume and Kant killed off the traditional (ontological, teleological, and cosmological) arguments for God’s existence, despite periodic attempts to disinter their skeletons and dress them in modern clothes. Philosophers like Plantinga, Alston, and Swinburne, who make many rationalist, intuitionbased arguments on behalf of God, also all accept some form of revelation (Christian Mystical Practice, Internal Instigation of the Holy Spirit, etc.) as an authentic epistemic means of gaining beliefs about the divine. Their arguments would be crippled without endorsing revelation, and they would be largely reduced to debating how a triune God is a logically coherent concept, or how it is logically possible that there could be a Hell and a perfectly just God. All these are idle academic speculations without some reason to think that there really is a God, or there really is a Hell, reasons that most philosophers who dismiss revelation believe are

A Problem for Philosophical Knowledge

67

lacking. So even if the method of intuition does not directly lead to atheism, it does seem to at least lead to agnosticism. I am offering an inductive argument, and am not laboring under the illusion that it is a knockout blow. Catholic defenders of revelation would no doubt reply that rational intuition alone, unaided by revelation, is bound to fail. John Paul II claims exactly that in Fides et ratio. We will not be able to correctly understand the nature of the self, devise a true theory of mentality, know how to behave in matters of sexual and medical ethics, or develop an accurate and comprehensive theory of justice without the aid of revelation. Revelation is needed to supply the data upon which reason can then operate. Trying to do philosophy without the input of God is a fool’s errand. This defense is a reiteration of the separate spheres view: there are nonempirical propositions that cannot be known through rationalist methods. However, pushing the separate spheres line is not an adequate rebuttal of the inductive argument I gave. No doubt adding the basic claims of Christian revelation into one’s set of beliefs would yield a different moral outlook, or a different theory of the mind, than what one might develop without Christian input. Whether such theories are closer or further from the truth is a separate matter. I am merely arguing that the method of rational intuition yields results inconsistent with the results of revelation (at least the Catholic version of it). The evidence is empirical: philosophers who rely on rational intuition as their principal method of acquiring moral beliefs or beliefs about the nature of mentality tend to agree with each other and disagree with the Catholic position on a wide range of issues. I am not arguing that secular philosophers are right. That’s a different issue altogether. Defenders of revelation might also reply that the fact that secular philosophers largely disagree with Catholic moral teachings or Christian doctrines about an immortal soul does not show that there is a fundamental conﬂict between rational intuition and revelation. The philosophy of the future will no doubt vindicate what has already been expounded ex cathedra. Any conﬂict between Athens and Jerusalem (to use the imagery of Tertullian) is a temporary chimera. Plantinga supports a view along these lines, writing, ‘‘if there were a demonstration or a powerful argument from other sources against Christian belief . . . then . . . this would be a genuine example of a clash between faith and reason. No such demonstration or argument, however, has so far reared its ugly head’’ (Plantinga 2000, p. 259, n. 34). Plantinga is a ﬁrst-rate, well-read

68

Chapter 2

philosopher, so one wonders how he has failed to notice that, in Blanshard’s words, ‘‘revelation and church authority have frequently said one thing while secular reﬂection has demanded another’’ (Blanshard 1975, p. 323). Presumably Plantinga assumes that he and his fellow travelers will be able to defeat all comers, past and present, who use arguments grounded in rational intuition to reject Christian belief. One gathers that he thinks it a small matter to refute the near-consensus among secular philosophers relying on rational intuition that the doctrines of original sin and the existence of immortal spiritual souls are irreparably mistaken. This is of course a possibility. But it is hard to ﬁnd sympathy for such a condescending, magisterial ‘‘when you learn to reason well, you’ll come to see that I was right all along’’ approach like this one. It is even harder to see how one might go about refuting it in the form stated above, as it appears unfalsiﬁable. Perhaps it is possible that the majority of contemporary philosophers are wrong about souls, sin, euthanasia, masturbation, premarital sex, and the rest, and that in the future rational intuition and reasoning will lead everyone to the Christian positions on these topics. However, this bare possibility does not refute the conﬂict thesis. Compare: it is logically possible that the earth science, paleontology, and biology of the future will reject evolution through natural selection and abandon the position that the earth is four-and-a-half billion years old. True; but this possibility does not show a compatibility between modern science and fundamentalist creationism. There are of course Christian philosophers devoted to giving a priori defenses of traditional religious claims, trying to reconcile the deliverances of rational intuition with Christian dogma. A common strategy among them is to give an intuition-based theory of knowledge and then discover (surprise!) that plenty of mainstream theological propositions— ones also supported by revelation—turn out to be warranted given the theory. I trust no examples are needed. Sometimes reading these writers, and their heroic attempts to square contemporary philosophy and scientiﬁc knowledge with the millennial teachings of an ancient Roman cult, one is reminded of Nietzsche’s comment that the faith of Pascal ‘‘resembles in a gruesome manner a continual suicide of reason—a tough, longlived, wormlike reason that cannot be killed all at once and with a single stroke.’’13 But is the situation much better for the secularists? It may be that practitioners of the traditional philosophical enterprise of reﬂection upon and analysis of basic intuitions are not terribly worried about the

A Problem for Philosophical Knowledge

69

conﬂict with revelation. Either they dismiss revelation as outmoded superstition, a shadow of the dead God that remains to be vanquished, or they take revelation seriously and hold out a candle, a` la John Paul II, Plantinga, Alston, and Swinburne, that everything will turn out all right in the end. I do not think that either of these responses is adequate. Philosophers in the West have for the most part gotten a substantial exposure to the beliefs and methodology of Christianity and have found some way of coming to terms with them. But there are other methods of coming to form beliefs about philosophical propositions that people in other cultures have taken seriously that when viewed through the eyes of analytic philosophy seem exotic indeed. I will discuss one of these in detail. The problem facing the method of rational intuition isn’t merely intuition vs. revelation, a familiar matchup that may seem capable of resolution, but that intuition is but one method of acquiring philosophical beliefs in a great sea of belief-acquiring methods. There are many competitors that must be shown to be less reliable than rational intuition if we are to retain the latter. Another Epistemic Method: Ritualistic Consumption of Hallucinogens The belief-acquiring method I will now turn to is the ritualized consumption of psychoactive drugs. In his history of drugs, Antonio Escohotado writes: Before the supernatural became concentrated into written dogmas, when priestly classes interpreted the will of a sole and omnipotent god, what was perceived in altered states was the core of innumerable cults, precisely under the heading of revealed knowledge. The ﬁrst hosts or holy sacraments were psychoactive substances, such as peyote, wine, or certain fungi. (Escohotado 1999, p. 2)

There is a complex interconnection between the use of psychoactive substances and religion. In ancient Greece, the use of wine was associated with the god Bacchus, and was seen as a gift of the god. Not all uses— even socially sanctioned uses—of wine were part of Bacchus-worship, however. In the Laws, Plato praises wine, noting that it was a gift from Bacchus to be used in invoking the gods in banquets. Yet Plato also speciﬁcally discusses the usefulness of wine as a solace for the elderly, one that allows them to feel young again and lift the burdens of old age (Laws, §666). It is worth recalling that the ancient Greeks routinely added psychotropic herbs to wine, in such quantities that wine nearly always required considerable dilution by water before it was safe to

70

Chapter 2

drink.14 Plato was an old man when he wrote the Laws, his ﬁnal work; no doubt he knew ﬁrsthand the salviﬁc virtues of wine. Wine was not the only psychoactive substance the Greeks used in worship. Perhaps the most important of the ancient Greek religions was the worship of the fertility goddess Demeter though the Eleusinian Mysteries. The Eleusinian Mysteries attracted thousands of celebrants from pre-Homeric Greece to height of Imperial Rome. It was a religion that lasted two thousand years, and was treated with great respect; one hierophant (the high priest or priestess of the cult) boasted that she had crowned both Marcus Aurelius and Commodus (Mylonas 1961, p. 231). The central teachings of the Mysteries are lost to our time. They were an oral tradition, passed down from hierophant to hierophant, and never written down. Furthermore, there were severe civil penalties if initiates into the religion ever spoke about or revealed what they witnessed at the Mysteries. There are several examples, but here are two: Alcibiades, while drunk, dared to imitate some of the secret ceremonies; despite his popularity he was promptly condemned in absentia and had all his property conﬁscated. In another case, Aeschylus was nearly lynched by an angry mob of theater-goers when they thought that he had revealed some of the secrets of the Mysteries in one of his plays.15 That said, there is a good bit known about the religion. One is that an important part of the initiation ceremony was the drinking of the kykeo´n, a potion composed of water, a grain (probably barley), and mint. This ritual was connected with the myth of Demeter and her breaking of a fast with the kykeo´n. Several recent writers have argued that the kykeo´n was a critical component of the Mysteries because it contained a grain parasite, the ergot mushroom (Calviceps purpurea), which is a hallucinogenic fungus that grows wild near Athens. Escohotado writes that this was quite psychoactive; to obtain its effects one needs only to pass the cereal sheaves through water and then discard the cereal, because the lysergic acid amides are soluble in water while the poisonous components are not. Considering that water was the medium utilized by the administrators of the sanctuary, we can therefore explain—without resorting to miracles or to the simple credulity of the devotees—the deep and infallible effect of the initiation. (1999, p. 17)16

Indeed, the use of hallucinogenic substances in religious and quasireligious ceremonies is as worldwide as religion itself. The Rig-Veda, the sacred text of the ancient Aryans of Northern India, makes numerous references to the intoxicating plant divinity Soma. It is now widely con-

A Problem for Philosophical Knowledge

71

sidered that Soma is none other than the ﬂy agaric mushroom (Amanita muscaria), a tall red mushroom with white spots whose physical properties and psychological effects closely match the descriptions of Soma in the Rig-Veda.17 In the ‘‘new’’ world, numerous hallucinogenic plants are at the center of indigenous religions, including the mushroom Psilocybe mexicana, and its kin, known to the Indians of Mexico as teonanacatl, a Nahuatl word meaning ‘‘ﬂesh of the gods.’’18 An Aztec religion involving the consumption of teonanacatl is nearly three thousand years old, despite the fervent attempts of the Spanish conquistadors to stamp it out. In northern Mexico, there are religions where the ingestion of the peyote cactus (Lophophora williamsii) is a crucial component. Peyote ceremonies have migrated into the United States and today are part of the Native American Church. In the Upper Amazon basin the vine Banisteriopsis caapi is boiled to produce a hallucinogenic drink. The resulting ‘‘vine drunkenness,’’ known to Peruvians as ayahuasca, is considered the portal to the spirit world (Harner 1973a). Common among nonliterate or preliterate societies are shamans, persons who are in direct contact with a spirit world through a trance state, often induced through chemical means, and who may have one or more spirits at his or her command to do the shaman’s bidding for good or evil (Harner 1973, p. xi). The value of intercourse with the supernatural in shamanic societies can vary greatly. The Jı´varo of the Ecuadorian Amazon hold that normal waking life is an illusion and that the real forces behind daily experience are supernatural ones, with witchcraft the cause of most illnesses and nonviolent deaths. These supernatural events can be known and manipulated only through the use of hallucinogenic vegetables (Harner 1973a, p. 16). For the Jı´varo, then, shamans are of vital importance—they cast spells on enemies and effect cures of those afﬂicted by magical curses. At the other end of the spectrum are the Apaches of the Mescalero reservation, where reliance on peyote-consuming shamans has become marginalized as the result of Westernization.19 Throughout the anthropological literature on the ritualistic use of hallucinogens are repeated discussions of shamans entering a spirit world after ingesting psychoactive compounds. Shamans go on ‘‘spirit-ﬂights,’’ journeys, and quests, do battle with or enlist the aid of supernatural beings, and attempt to bring back knowledge from the spirit world to help their tribe or community. This is quite similar to Christianity, whose adherents have interactions with good and bad spirits (angels and

72

Chapter 2

demons) and try to gain knowledge through commerce with a supernatural spirit world. Both shamans and Christians engage the spirit world through a great deal of stylized ritual, thick with symbolism and ceremonies for rites of passage and puriﬁcation. Shamans are typically called on to be healers who cure patients of magical ailments brought about by evil sorcerers or malicious spirits. Christian ministers are also called on to heal followers of supernatural malady (e.g., sin or demonic possession) through rituals of prayer, confession, and exorcism. In addition, both shamanic religions and Christianity include psychoactive sacraments; in fact some authors have argued that Christianity’s acceptance in Rome was partly the result of its familiar rituals. The consumption of wine and bread during the Eucharist would have reminded Romans of the worship of Bacchus (with its wine) and the Eleusianian Mysteries (with its grain-based kykeo´n). Thus the Romans would have been predisposed to view the Eucharist as an authentic religious rite.20 More generally, ancient religions established their bona ﬁdes among the general public by claiming roots in deep antiquity; there was widespread suspicion of anything new or recent in religion. This is one of the reasons the version of Christianity that claimed continuity with Judaism was the historical victor over the Marcionites, who rejected all things Jewish and insisted that Jesus established a religion de nuovo.21 The chief difference between Christianity and shamanism that I want to emphasize here is that in the case of Christianity knowledge is supposedly acquired from the supernatural world through revelation, whereas shamans probe the mysteries of the supernatural through chemical means. Revelation is essentially a passive affair in which God shares information with a person. No one can force God to give up his secrets; even Calvin, who thinks that each person has a capacity to sense the divine, doesn’t hold that a person could just up and have a revelation of their own volition. Shamans, on the other hand, intentionally ingest psychoactive plants precisely in order to engage the spirit world. It would be a serious mistake to revise or assimilate these indigenous religions so that, as Calvin thought, they are just primitive, crude attempts to discover the truth of the Christian God. Psychoactive Drugs Are a Method of Acquiring Beliefs about Philosophical Propositions There is no real historical tradition of Western logic and philosophical reasoning interacting with shamanic hallucinogen religions in the same way that one ﬁnds an Anselm or an Aquinas attempting to wed philoso-

A Problem for Philosophical Knowledge

73

phy and the Christian religion. The result is that the language one ﬁnds in the ethnographic reports of anthropologists is not as sensitive to commonplace philosophical distinctions as that of virtually any contemporary Christian theologian. So it is somewhat more difﬁcult to make explicit precisely which propositions are believed as the result of hallucinogenic experiences. It is clear, however, that hallucinogens are used in many cultures not for mere recreational purposes, or even as a component of a nondoxastic ritual analogous to, say, Christian genuﬂection, but have a deﬁnite epistemic aspect. As in the case of revelation, the ritual use of hallucinogens generates beliefs about the nature of the mind. The anthropologist Michael Harner writes, ‘‘there can be little doubt that the use of the more powerful hallucinogens tends to strongly reinforce a belief in the reality of a supernatural world and in the existence of a disembodied soul or souls’’ (Harner 1973b, p. xiv). For example, the Jı´varo believe that the effect of ayahuasca is that part of the soul leaves the body. The Conibo-Shipibo of the Ucayali region of eastern Peru report that ayahuasca can cause the souls of shamans to leave their bodies and assume the form of predatory birds. The Amahuaca, neighbors of the Conibo, also report that souls leave bodies when one drinks ayahuasca, and many of the Desana branch of the Tukano in eastern Columbia hold similar beliefs.22 This is quite easy to understand, as hallucinations caused by ayahuasca frequently include the feeling of ﬂying and soaring over great distances. Beliefs about the existence and nature of the divine are also produced by hallucinogen rituals. The Tukano of the territory of Vaupe´s in the northwest Amazon region of Columbia drink ayahuasca, or yaje´. The purpose of doing so is, among other things, to see the tribal divinities. Upon waking from the yaje´ trance, the Tukano are convinced of the truth of their religious system, having experienced ﬁrsthand, courtesy of the Banisteriopsis caapi vine, their gods and the seminal events of the tribe’s mythology (Reichel-Dolmatoff 1972, p. 102). Experience of the gods is also related in the Rig-Veda: We have drunk the Soma, we are become Immortals, We are arrived at the Light, we have found the Gods. What now can hostility do to us, what malice of mortal, O Immortal Soma!23

Here the Vedic poet relates the direct experience of the divine through the ingestion of the sacred hallucinogen. The connection is quite explicit: drink the Soma and ﬁnd the Gods. In the Bwiti religion of sub-Saharan

74

Chapter 2

Africa, the Eboka plant is used as a sacramental hallucinogen. Takers of the drug do so for a wide variety of reasons, but one is in order to know God, and to know things of the dead and the land beyond (Fernandez 1972, p. 251). In fact, there is a quite extensive literature on the historical use of entheogenic, or god-spawning, substances and their role in the formation of religions. Some scholars argue that entheogens are the root of all shamanic ecstasy—the mysterium tremendum—and ultimately the ancient basis of all religion.24 These more extravagant claims aside, there can be little doubt that for many religions, the ritual use of psychoactive substances has played and continues to play an important role in forming beliefs about the nature of the divine. Finally, hallucinogens have been used to acquire beliefs about moral propositions. In De legibus (II.14.36), Cicero writes the following concerning the Eleusinian Mysteries: For among the many excellent and indeed divine institutions which . . . Athens has brought forth and contributed to human life, none, in my opinion, is better than those mysteries. For by their means we have been brought out of our barbarous and savage mode of life and educated and reﬁned to a state of civilization; and as the rites are called ‘‘initiations,’’ so in very truth we have learned from them the beginnings of life, and have gained the power not only to live happily, but also to die with a better hope. (Cicero 45 b.c.e. [1928])

This is clearly moral language. Cicero is not praising the Mysteries because of the practical knowledge that they communicate, or the metaphysics they endorse, but because they taught the Romans a moral lesson. Cicero claims that it is the sense of community and the knowledge of how to live a good life as conveyed by the Mysteries that civilized the Romans. If Wasson, Escohotado, Schultes, and others are right that the key to the Eleusinian Mysteries was the consumption of the hallucinogenic barley ergot, then the initiate Cicero is proposing that the ritualized use of psychoactive agents is a means of gaining moral knowledge.25 In the New World, the Huichol of Mexico have a complex ritual life centered around the hunt for peyote.26 Many Huichol participate in the traditional (indeed pre-Columbian) peyote pilgrimage, in which they make an arduous journey to collect and consume the cactus. The point of this pilgrimage is to ‘‘learn what it means to be Huichol’’ and ‘‘to ﬁnd their life.’’ Again, this ritual has a speciﬁcally moral component. Through the consumption of the peyote hallucinogen, the Huichol gain beliefs about how they are to be good members of their tribe and how it is that they ought to conduct their lives.

A Problem for Philosophical Knowledge

75

Hallucinogen-generated Beliefs Are Foundational and Basic Alston argues that there can be direct perceptual experiences of God, experiences that give rise to epistemologically basic beliefs that God is doing something (comforting, strengthening, guiding, communicating) or that God has some property (goodness, power, lovingness). While Alston is quick to acknowledge that some ostensible experiences of God may be mistaken, just as some ordinary perceptual beliefs may be the result of illusions, he does think that there are genuine God-perceptions. He goes to great lengths to show that mystical experiences ought to be construed as a kind of perception. The kind of mystical experience he is willing to countenance is illustrated well here: I remember the night, and almost the very spot on the hilltop, where my soul opened out, as it were, into the Inﬁnite, and there was a rushing together of the two worlds, the inner and the outer. It was deep calling unto deep—the deep that my own struggle had opened up within being answered by the unfathomable deep without, reaching beyond the stars. I stood alone with Him who had made me, and all the beauty of the world, and love, and sorrow, and even temptation. I did not seek Him, but felt the perfect unison of my spirit with His. The ordinary sense of things around me faded. For the moment nothing but an ineffable joy and exultation remained. It is impossible fully to describe the experience. It was like the effect of some great orchestra when all the separate notes have melted into one swelling harmony that leaves the listener conscious of nothing save that his soul is being wafted upwards, and almost bursting with its own emotion. The perfect stillness of the night was thrilled by a more solemn silence. (Alston 1991, p. 22)

Compare the preceding apprehension of the divine with the following, drug-induced experience: Suddenly I burst into a vast, new, indescribably wonderful universe. Although I am writing this over a year later, the thrill of the surprise and amazement, the awesomeness of the revelation, the engulfment in an overwhelming feeling-wave of gratitude and blessed wonderment, are as fresh, and the memory of the experience is as vivid, as if it had happened ﬁve minutes ago. And yet to concoct anything by way of description that would even hint at the magnitude, the sense of ultimate reality . . . this seems such an impossible task. The knowledge which has infused and affected every aspect of my life came instantaneously and with such complete force of certainty that it was impossible, then or since, to doubt its validity. (Smith 1964, p. 522)

It would take tortuous reasoning indeed to show that the former is an acceptable candidate experience for a perception of the divine and the latter experience is not. Notice that in the latter example there is an

76

Chapter 2

even more explicit claim to have learned something from the episode; the drug experience is supposed to provide knowledge. When W. T. Stace, author of Mysticism and Philosophy, was asked whether the drug experience is similar to the mystical experience, he answered, ‘‘It’s not a question of being similar to mystical experience; it is mystical experience.’’27 It is striking that Alston, who gives lengthy arguments that mystical experiences should be understood as perceptual ones, and indeed as putative perceptions of God, completely fails to engage the extensive literature on entheogens. In any case, if Alston’s mystical experiences give rise to epistemically basic, noninferential beliefs, so do those caused by hallucinogenic substances. Reason Is Used to Develop Foundational Beliefs into a System It is worth reminding ourselves that Himalayan shamans, preliterate indigenous Peruvians, and Eleusinian adepts are not going to express themselves in the terms of twenty-ﬁrst-century philosophy. It would be a grievous error to declare that they have not developed the sophisticated apologetics and eschatology of, say, Alvin Plantinga and so should not be taken as seriously. Civilizations as great as the Aztecs and the Greece of Pericles incorporated the use of psychoactive drugs as legitimate belief-acquiring methods into their societies. This is not to say that no one ever dissented from the common practice, just as there are atheists today; but these were familiar ways that people used to gain (what they took to be) knowledge. The experiences induced by peyote, ﬂy agaric, rye ergot, and the rest of the garden did not on their own give rise to a coherent system of thought. It took the teachings of the hierophants, the shamans, and the tribal elders to make sense of the hallucinogenic visions and give them a context in which they could be understood. It is one thing to have a sense of ﬂying, seeing bird-headed people, and having conversations with dragons who declare themselves to be gods. It is another to have an opinion about what that all meant, and how it was relevant to everyday life. In the so-called Miracle of Marsh Chapel of 1962, twenty Christian theology students participated in a placebo-controlled double-blind psilocybin experiment at Boston University. Walter Pahnke, a Harvard graduate student experimenter, administered 30 mg of psilocybin to ten students. A control group of ten students received an active placebo of 200 mg of nicotinic acid, a vasodilator which causes no psychic effects, only warmth and tingling of the skin. Nine out of the ten who had received psilocybin reported having what they considered to be an au-

A Problem for Philosophical Knowledge

77

thentic religious experience. In a chapel, on Good Friday, with theology students as test subjects, psilocybin led to an experience of the Christian God. In 1991, a researcher who did a long-term follow-up of Pahnke’s original test subjects reported ‘‘all psilocybin subjects participating in the long-term follow-up, but none of the controls, still considered their original experience to have had genuinely mystical elements and to have made a uniquely valuable contribution to their spiritual lives’’ (Doblin 1991). Many of these subjects are now Christian ministers and have, like shamans, integrated their psilocybin experience into their religious belief systems. Their hallucinations were not rejected as perceptual outliers but taken, Alston-like, as authentic perceptions of the divine. The communities discussed above did not dismiss the experiences caused by psychoactive drugs as recreational cinema, but incorporated them into the ongoing life of the society. That is, the beliefs produced by the hallucinogens didn’t stand on their own, but had to be ﬁtted into a coherent body of beliefs. As in the case of Christian revelation or philosophical intuition, reason must be employed to take the foundational propositions about morality, the mind, and the divine, and systematize them into at least a minimally coherent way of understanding those topics. The Conﬂict Thesis Harner argues that the beliefs generated by the use of hallucinogens are ‘‘sooner or later bound to come into conﬂict with the orthodox dogma fundamental to the ideological structure of state religions’’ (1973a, p. 7). He takes this fact to explain why the use of hallucinogens tends to eventually be forbidden in Western societies. Even under the paternalistic obsession with safety currently prevalent in the United States, it is very difﬁcult to make out an argument for legally prohibiting the personal use of psychotropic plants on the basis of harm to self. So Harner may well have a point. In any case, it does seem that many hallucinogenproduced beliefs are inconsistent with the beliefs of a majority of philosophers who rely on rational intuition. This inconsistency is most clearly seen in the philosophy of mind. An overwhelming number of individuals believe that they possess (at least one) spiritual, nonphysical soul as the result of ingesting psychotropic drugs. As discussed previously, practically no contemporary philosopher accepts substance dualism. Furthermore, beliefs in the tribal gods are also a common result of the ritual use of hallucinogens. I claimed earlier that it is a minority of philosophers who rely solely on rational intuition to gain beliefs about philosophical

78

Chapter 2

propositions who believe that there is a supernatural realm of spirits and gods. In the case of ethics, it is much more difﬁcult to codify the moral propositions believed by the Huichol or the Jı´varo than it is to examine the moral claims of the Catholic Church. Ergo it is that much harder to show how their hallucinogen-based moral beliefs are directly in conﬂict with what most philosophers think follows from rational intuition. However, I will suggest this: it would be astonishing if further ethnography showed the moral beliefs of a Huichol shaman to be isomorphic with those of most rationalist philosophers. Which Method Is Best? I have presented three different methods of acquiring beliefs about philosophical propositions: rational intuition, Christian revelation, and ritualistic consumption of hallucinogens. I have argued that all three methods produce noninferential, basic beliefs, that those who use these methods proceed to take the basic beliefs and develop them into more comprehensive belief systems through the use of reason, and that one method may yield a belief that p while another produces a belief that not-p. The question then arises: which of these methods is the most reliable way of acquiring beliefs about philosophical propositions? We plainly can’t use all three methods without running into inconsistency. One answer, that of the naturalists, is that all three methods are equally reliable. That is, none of them is reliable at all, and so the question of which method is to be preferred is like this one: ‘‘which is a better predictor of future experience, reading tea leaves, divining the entrails of a chicken, or shaking the Magic 8-Ball?’’ The naturalists’ challenge will be dealt with more fully in a later chapter. Clearly, the naturalists’ answer is rejected by rationalist philosophers. But even if (a big, as yet unargued if) naturalism is to be rejected, why think rational intuition is any better at gaining the truth than revelation or peyote? Here are four arguments to the conclusion that rational intuition is epistemically better than revelation or the ritual use of psychoactive drugs when it comes to gaining beliefs about philosophical propositions. I will argue that each of the following arguments does not adequately show the relative superiority of intuition. Four Briefs on Behalf of Rational Intuition Argument 1: The proof of the pudding is in the eating All three methods rely on the post hoc evaluation of the foundational beliefs in order to

A Problem for Philosophical Knowledge

79

achieve reﬂective equilibrium. That is, analytic philosophers, Christian theologians, and native shamans all seem to agree that the set of our beliefs about philosophical propositions should exemplify certain epistemic virtues. There is general agreement that we need to integrate our basic beliefs about the mind, the gods, and the good life into a systematic, cohesive way of understanding the world and our place in it. As Kai Nielsen suggests, the virtue of wide reﬂective equilibrium is that it offers a procedure for selecting one network of beliefs over another: the more stable, cohesive, and comprehensive system of beliefs is the one to be retained (Nielsen 1993, p. 327). Now, suppose that the set of beliefs produced by rational intuition is more epistemically virtuous than belief-sets generated by psychoactive drugs or revelation. Beliefs based on rational intuition turn out to be more comprehensive, uniﬁed, consistent, and explanatorily cohesive than those generated by other methods. This superior result is a good reason to prefer intuition as a basic method—and the perspective of rationalist philosophy as more credible—than the alternatives we have been considering. There are two main problems with this approach. The ﬁrst is the contention that rationalist philosophy has already beaten out the competition when it comes to philosophical beliefs in wide reﬂective equilibrium is mostly ill-informed hubris. Consider, for example, the intricacy and level of detail one ﬁnds in Catholic theology. The Catholic worldview is as meticulous and all-encompassing as the most ambitious philosophical system, indicating that belief-sets resulting from revelation fare quite well by the standards of comprehensiveness, consistency, explanatoriness and similar criteria. Furthermore, just as analytic philosophy has recognized the need to incorporate scientiﬁc discovery, so has the Catholic Church, which now concedes that both Galileo and Darwin were right and has found ways to accommodate this opinion in their theology. One might claim that traditional rationalism is at least better off than tribal religions founded upon the inspirations of yaje´, but even there further argument is needed, with anthropologically informed evidence and not merely complacent philosophical optimism. The second problem with the ‘‘proof of the pudding’’ argument is more systemic. The overall strategy of the argument is very similar to traditional coherentist theories of justiﬁcation. For such theories, a belief is justiﬁed just in case it is an element of a set of beliefs that are mutually explanatory, consistent, and the like. The better off a set of beliefs is on these measures, the more justiﬁed its elements. Here, a belief-acquiring method is vindicated if it produces belief-sets that are eminently virtuous by similar criteria. A classic difﬁculty for coherence theories of

80

Chapter 2

justiﬁcation is the plurality problem: it is possible for there to be two sets of beliefs that are (1) equally coherent; (2) no other set is more coherent than they are; and (3) such that in one set p is believed, but in the other not-p is believed (imagine detailed ﬁctional stories, or paranoid fantasies). Thus both p and not-p are equally justiﬁed. The coherence theory of justiﬁcation has failed to guarantee the very thing that we want to get out of our epistemic methods: an ultimate convergence on the truth. Persons contemplating p are left in the position of Buridan’s ass—to the left is p, and a fully coherent set of which it is a member, and to the right is not-p, an element of a distinct but equally coherent set that one could adopt. There is no principled way to choose. The fundamental hurdle for the proof of the pudding argument is an analogue of the plurality problem; there are many possible belief-sets that are equally coherent, equally capable of being held in wide reﬂective equilibrium, but inconsistent with each other. Indeed, I argued above that there actually are such belief-sets. I confess that I am not sure how to forestall every argument a defender of rational intuition might invent to show that the totality of rationalism is immune to the plurality problem. Let us rest content with enumerating these challenges: showing that the beliefs produced by intuition are eminently virtuous requires (1) a substantial empirical argument demonstrating that the belief-sets that rely on revelation or the ritual use of hallucinogens as basic beliefacquiring methods are not as comprehensive, consistent, and explanatorily cohesive as sets of beliefs that eschew those methods in favor of rational intuition; and (2) a philosophical argument that demonstrates how rationalism can avoid the plurality problem and shows that intuition must have more systematizable results than revelation or the shamanistic consumption of yaje´. Perhaps these concerns could be answered, but so far it looks as if the proof of the pudding argument is chieﬂy wishful thinking. Argument 2: Common doxastic practice George Bealer writes, ‘‘It is standard justiﬁcatory practice to use intuitions evidentially. Unless and until a reason for departing from this standard practice is produced, we are entitled—indeed, obligated—to continue using intuitions as evidence.’’28 In other words, if everybody else trusts intuition, we should too. It would be irrational of us to discard intuition when recognized experts in our peer group rely on it and we don’t have a powerful argument against them. Analogously, it is standard justiﬁcatory practice to use perceptions and sense experience evidentially. Powerful arguments

A Problem for Philosophical Knowledge

81

are needed to overthrow this method. Bealer’s conclusion is that, pending excellent skeptical reasoning, we are justiﬁed in continuing to employ the method of rational intuition to acquire beliefs about philosophical propositions. Unfortunately, the common doxastic practice argument is not going to privilege rational intuition over the two competitors we are considering. In fact, Plantinga and Alston use exactly the same argument, mutatis mutandis, to defend the use of Christian revelation. Here are representative passages from Alston: My main thesis in this . . . whole book is that Christian Mystical Practice is rationally engaged in since it is a socially established doxastic practice that is not demonstrably unreliable or otherwise disqualiﬁed for rational acceptance. (1991, p. 194) It is prima facie rational to engage in Christian Mystical Practice . . . because it is a socially established doxastic practice; and it is unqualiﬁedly rational to engage in it . . . because we lack sufﬁcient reason for regarding it as unreliable or otherwise disqualiﬁed for rational participation. (1991, p. 223)29

Furthermore, the common doxastic practice argument can be equally well employed on behalf of the ritual use of hallucinogens. Chewing peyote buttons to acquire moral beliefs is a socially established doxastic practice among the Huichol that is not demonstrably unreliable or otherwise disqualiﬁed for rational acceptance. In Alston’s words, ‘‘it is a reasonable supposition that a practice would not have persisted over large segments of the population unless it was putting people into effective touch with some aspect(s) of reality and proving itself as such by its fruits’’ (1991, p. 170). Now, we might try to disqualify the method on the grounds that the moral beliefs arrived at through peyote are inconsistent with those derived from rational intuition. Therefore trusting the teachings of Huichol shamans is an unreliable way of gaining moral truths. However, this begs the question in favor of intuition—we have not yet demonstrated that intuition is more reliable than peyote. The same goes for Christian revelation. In fact, both Plantinga and Alston want to salvage revelation from a similar threat. They argue that revelation is a basic belief-acquiring method, and that it is illegitimate to use one basic method (they are particularly concerned with sense perception) to evaluate another. They argue that it is arbitrary to assume in advance that if revelation is a source of warranted belief, then it must be likely to be reliable by the lights of ordinary sense perception. It would be equally unjustiﬁed to insist that sense perception turn out reliable

82

Chapter 2

with respect to some group of epistemic powers that doesn’t include sense perception (like revelation). Christian revelation and sense perception are autonomous, independent methods of acquiring beliefs. They are separate but equal.30 So eating ﬂy agaric to ﬁnd the gods may look valueless from the point of view of rational intuition, rational intuition may look incomplete from the vantage of Christian revelation, and sober revelation may look bizarre from the points of view of both traditional shamans and rationalist philosophers. If Plantinga and Alston are right and we can’t use one basic method to assess another, then everyone can use the common doxastic practice argument; all we need to do is ﬁnd the right peer group. Therefore it is no help in adjudicating among our three belief-acquiring methods. (Incidentally, I am quite sympathetic to the notion that what we ought to believe about philosophical issues is deeply connected to systematic doxastic practices. Under the heading of ‘‘perspectives,’’ this idea will play a vital role in the subsequent chapter.) Argument 3: The mint of nature

Thomas Reid writes:

The skeptic asks me, Why do you believe the existence of the external object which you perceive? This belief, sir, is none of my manufacture; it came from the mint of Nature; it bears her image and superscription; and, if it is not right, the fault is not mine: I even took it upon trust, and without suspicion. Reason, says the skeptic, is the only judge of truth, and you ought to throw off every opinion and every belief that is not grounded on reason. Why, sir, should I believe the faculty of reason more than that of perception?—they came both out of the same shop, and were made by the same artist; and if he puts one piece of false ware into my hands, what should hinder him from putting another? (1764 [1975], pp. 84–85)

Richard Foley and Paul Tidman interpret Reid’s remarks as treating our epistemic faculties as equally trustworthy, and that reliance on reason or intuition is just as reasonable as trust in our senses. Tidman, for example, writes, ‘‘nothing further [than intuition] is needed to justify appeals to intuitions . . . because these beliefs are produced by a basic belief-forming mechanism we have no reason to question. . . . Each of our faculties is innocent until proven guilty.’’31 Given that there are many different methods of forming beliefs about philosophical propositions that we might choose, why should we trust the belief-forming method of intuition? Reid’s response is that the mint of nature has beneficently issued to us innate tendencies to use certain belief-acquiring methods and the ability to use these methods so that they reliably generate

A Problem for Philosophical Knowledge

83

true beliefs. Nature gave us sense perception, and that was pretty good. Nature gave us intuition as well, so we should go ahead and trust it too, unless confronted with an excellent reason not to. The cheery optimism of Reid and his followers is open to several criticisms, and which of these is especially telling depends partly on exactly how we interpret the preceding argument. If the mint of nature argument rests on the idea that we should trust any natural (whatever this means exactly) procedure of gaining beliefs so long as we don’t have a reason not to, as Tidman intimates, then it incorporates a fallacious appeal to ignorance. If its central premise is if one natural method of belief-acquisition is reliable, then we are justiﬁed in believing that any natural method of belief-acquisition is reliable, then the argument is clearly guilty of hasty generalization. Moreover, even if this charge could be answered, the mint of nature argument won’t uniquely pick out rational intuition as the best method of gaining beliefs about necessary propositions. In fact, Plantinga and Alston cite exactly the same passage from Reid to defend their trust in revelation. Here’s Alston: ‘‘we will follow the lead of Thomas Reid in taking all our established doxastic practices to be acceptable as such, as innocent until proven guilty.’’32 And, of course, the established doxastic practice Alston is really interested in is Christian mystical practice, or revelation. We can’t give short shrift to the ritualistic use of psychoactive drugs either; this too is an established doxastic practice, just one not practiced by Alston. The mint of nature may have coined perception and reason, but it also issued Banisteriopsis caapi, Lophophora williamsii, Amanita muscaria, and Psilocybe mexicana. ‘‘Why, sir, should I believe the faculty of rational intuition more than the eating of hallucinogenic vegetables?—they came both out of the same shop, and were made by the same artist,’’ and so on. If Reid’s main point is that the feature of ‘‘having been generated by nature’’ certiﬁes a belief-acquiring method’s trustworthiness, then the argument is open to myriad criticisms by the naturalists. Stephen Stich, for example, has argued that nature—evolution by natural selection— has no interest in producing organisms whose beliefs are largely true, nor do we have a compelling reason to maintain that any of our cognitive processes are innate.33 Stich argues that an organism with a very cautious, risk-averse inferential strategy may rack up a disproportionate number of false beliefs about the probability of danger. Nevertheless, overcaution in forming beliefs about danger may be a better reproductive strategy than a moderate caution that yields more reliable beliefs about danger. As Stich puts it, ‘‘natural selection does not care about truth; it

84

Chapter 2

only cares about reproductive success. And from the point of view of reproductive success, it is often better to be safe (and wrong) than sorry’’ (1990, p. 62). In addition, there may be high costs, in terms of time, effort, and cognitive hardware, associated with optimizing the accuracy of our belief-acquiring methods. For instance, students will stop studying for an exam and go to a party once they have satisﬁed themselves that they have achieved adequate mastery of the material. They will do this even when—as is usually the case—further studying would provide greater comprehension and more true beliefs. At some point, they decide that attaining additional knowledge of the subject matter is not worth the cost of missing the party. Similarly, if there is a declining marginal utility for true beliefs, then natural selection might well select a less accurate (and less costly) cognitive system over a more accurate one. Even if the Reidians are unpersuaded by Stich and insist that we can trust nature for some methods, for example, sense perception, memory, or induction, the naturalists will argue that extending this conﬁdence to rational intuition is unjustiﬁed. Danger is routinely piggybacked upon our adaptive traits. For example, people with just one gene (heterozygotes) that codes for a deviant hemoglobin structure are protected against malaria, so there is a selective advantage to be had by preserving the gene. However, people who have both the father’s and the mother’s gene (homozygotes) for this deviant hemoglobin structure develop sickle-cell anemia. Despite the fact that those with sickle-cell anemia are less likely to replicate their genes than those without the disease, evolution has not eradicated sickle-cell because of the antimalarial beneﬁts for heterozygotes. Sickle-cell anemia is an evolutionary spandrel, an accidental by-product of the forces that made deviant hemoglobin cells in heterozygotes an adaptive advantage.34 Analogously, it may be nothing more than an accidental by-product of evolution that we ﬁnd some propositions to be intuitively true and not others, an accident that has only a random connection to the truth. Or it could be that reliance on intuition is positively damaging to the quest for truth. If the Reidians are to place their faith in natural sources of belief-formation, they need to demonstrate the selective advantage of believing reliable intuition of necessary propositions. Otherwise there is no reason to think that rational intuition is anything other than a spandrel. Either that, or the Reidians must provide some other interpretation of ‘‘natural’’ in epistemic contexts. I have no intention of settling the debate over naturalism here. Rather, I want to make two points: ﬁrst, the optimistic conﬁdence that

A Problem for Philosophical Knowledge

85

if nature gave us one good method of belief-acquisition, then we should prima facie trust any other natural method, is naive and misplaced in light of the naturalists’ arguments that natural selection is relatively indifferent to the production of true beliefs and that all three of the belief-acquiring methods we have been considering may be evolutionary spandrels, with their output only accidentally related to the acquisition of true beliefs. Even if in the end we reject naturalism, it has enough going for it to undermine prima facie trust in nonnaturalistic beliefacquiring methods. I am not arguing that we are never prima facie justiﬁed, absent further evidence, in believing the output of a belief-forming faculty. I am arguing that given the naturalists’ arguments above, we are not prima facie justiﬁed in believing that any of the three methods we have been considering are trustworthy producers of true beliefs. More argument is needed. The second point to be made is that Reid’s mint of nature argument is not going to adjudicate among rational intuition, revelation, and hallucinogen use. Defenders of each of these methods can claim their chosen methodology to be a natural faculty whose deliverances we are prima facie justiﬁed in accepting. Thus all protagonists can use the Reidian argument to their advantage and we are left with the same problem of conﬂicting output. The mint of nature argument therefore provides no means of showing that rational intuition is an epistemically superior means of gaining philosophical beliefs. Argument 4: The priority of reason The ﬁnal argument on behalf of the authority of rational intuition over its competitors is as follows. All parties agree that we need to use reason to operate upon our basic beliefs and develop epistemically virtuous belief-sets. Reason is used to uncover and resolve inconsistencies, eradicate fallacious inferences, promote explanatory unity, and so on. In other words, rational reﬂection is the basic method by which we come to have justiﬁed beliefs about philosophical propositions. Other methods, like revelation and drinking yaje´, are addons. Rational intuition is the basic model and the others are optional features. So if push comes to shove, and there are inconsistencies among the basic beliefs provided by intuition and those provided by revelation or psychoactive plants, we’d better stick with the basic package. Since everyone agrees that reason is essential and not all agree that we need these ‘‘bonus’’ methods, in the case of trouble, the bonus methods should be the ﬁrst to go. Rational intuition remains the best way of acquiring beliefs about philosophical propositions.

86

Chapter 2

Unlike the proof of the pudding, common doxastic practice, and mint of nature arguments, the priority of reason response is not one that can be used by all protagonists. I have argued that Christians theologians, hallucinogen-consuming shamans, and traditional intuition-driven philosophers rely on reason to systematize their beliefs and attain reﬂective equilibrium. However, there are three problems with the priority of reason argument. First, there is no reason to suppose that all three groups recognize the same set of properties as being epistemically virtuous. There may be a wide discrepancy as to what qualities they want their belief-sets to exemplify. Even if both the Jı´varo and Christian theologians count certain beliefs as good inferences from logically prior beliefs, this does not imply that they are using the same concept of ‘‘good inference.’’ Even among philosophers, some regard justiﬁcation to be a normative concept and some think it is causal-nomological. Some philosophers defend an internalist notion of justiﬁcation, and some an externalist one. We should be cautious not to overstate agreement about exactly how reason is to be used in developing inferential beliefs. The priority of reason argument takes cross-party agreement about the value of reason as an inferential procedure and uses that fact to encourage acceptance of rational intuition as the best way to gain basic beliefs about philosophical propositions. Recognizing that there may be a good bit of diversity concerning exactly how reason is to be used in achieving reﬂective equilibrium undermines the central claim of cross-party agreement. Second, there is a difference between using reason to evaluate basic beliefs and draw inferences (inductively or deductively) from these basic beliefs, and using rational intuition to generate the basic beliefs themselves. That is, once one has a set of basic beliefs, arrived at by whatever method, reason is then employed to operate upon this set and produce inferential beliefs. This procedure is a different activity from the use of a priori philosophical intuition to arrive at basic beliefs. The use of reason as an inferential procedure is logically separate from how one acquires noninferential beliefs. An independent argument is required to demonstrate that intuition is the best way to get noninferential beliefs; the value of reason in making good inferences from the foundations tells us nothing about how to get the right foundations. There is an acronym in computer programming: GIGO—garbage in, garbage out. A computer is an extensional logic machine, but its output is only as good as the data it is fed. The present concern is not over how to process the data, but how to get the data in the ﬁrst place. The priority of reason argument is that

A Problem for Philosophical Knowledge

87

since reason is a good inference procedure, rational intuition is a good way of gaining noninferential beliefs. This is like arguing that since a computer’s inferences are all logical deductions, its input must be logical truths—a manifest non sequitur. The ﬁnal difﬁculty with the priority of reason defense is brought out clearly by analogy. In 1610 Galileo published an account of his recent astronomical discoveries in Sidereus nuncius, revealing among other things that the moon is mountainous, craggy, and not perfectly spherical. Galileo faced some legitimate criticisms, for example that his telescopes were still fairly crude, imprecise, and not entirely to be trusted. Purely rationalist arguments were offered as well; the Aristotelians of the day opposed Galileo on the grounds that empirical science relies on the use of reason to adjudicate competing hypotheses, evaluate data, draw correct inductive inferences, and so on. Thus reason, rational reﬂection, a priori intuition is the primary method by which we gain empirical knowledge. Should there be any discrepancies between what pure reason tells us about the nature of the physical world and the observations of the experimentalists, we must discard experimentation and rely solely on reason, our fundamental method. Since reason dictates that the universe conform to the a priori beauty of the spheres, argued the pedants, we are entirely justiﬁed in discarding Signore Galileo’s telescopic ﬁndings to the contrary. In fact, Giulio Libri, one of the foremost philosophers at Pisa, refused even to look through Galileo’s telescope.35 Obviously, this mulishness is precisely the sort of scholastic hubris that fomented resistance to experimental science in the Renaissance. It is clear in the case of empirical knowledge that reason alone is inadequate. We must also use the method of sense perception to acquire basic beliefs. Simply because reason (in the form of science) operates upon these basic beliefs to produce intricate theories is no evidence that sense perception is not also needed as a method of belief-acquisition. Just as reason operates upon the beliefs produced by sense perception to develop a systematic scientiﬁc understanding of the world, so too reason operates upon the beliefs produced by Christian revelation to develop a comprehensive theological outlook, and utilizes the beliefs generated by psychoactive drugs to develop a coherent view about various philosophical propositions. In short, the priority of reason argument is precisely the same argument that Aristotelian dogmatists used to dismiss Galileo’s experimental method of acquiring empirical beliefs. It is employed in the present context to dismiss revelation and the ritual use of hallucinogens as

88

Chapter 2

methods of acquiring beliefs about philosophical propositions. If the priority of reason argument wasn’t a good argument then, it isn’t a good one now. Galileo didn’t deny that pure reason had its place; he denied that it alone could discover the nature of the heavens. Likewise, John Paul II doesn’t deny that some knowledge can be gained through rational intuition. Rather, what he denies is that it alone can discover the nature of heaven. Reason alone isn’t enough for science. So far, we lack an adequate argument to think that is enough for philosophy. At this point I have considered four different arguments designed to show the relative superiority of intuition: the proof of the pudding, common doxastic practice, mint of nature, and priority of reason arguments. All four turn out to be inadequate—they lend no more credibility to intuition than they do to Christian revelation and the ritual consumption of hallucinogens. Our quandary remains: the methods of intuition, revelation, and peyote produce inconsistent beliefs, and we can’t seem to show that one is better than the others. What now? A Trilemma for Philosophical Knowledge One might think that externalists have an easy response here. Externalists maintain that true belief is converted to knowledge just in case some objective relation, that the believer may or may not be aware of, holds between the truth of the belief and the belief itself. Probably the best known version of externalism is Goldman’s process reliabilism. Ignoring the subtle nuances, reliabilism maintains that one knows that p if and only if p is true and one’s belief that p was formed by a generally reliable process. One of the nice things about externalism is that it has a quick and easy response to traditional skeptical arguments. ‘‘You can’t know anything about an empirical world outside of your mind,’’ charges the skeptic, ‘‘because you can’t show that you’re not dreaming, or a brain in a vat, or that any other skeptical hypotheses aren’t true. Your evidence won’t eliminate them. Therefore even if your empirical beliefs are largely true, they don’t amount to knowledge.’’ ‘‘I don’t need to eliminate such possibilities,’’ replies the externalist; ‘‘my view is that so long as my beliefs are true and formed by a process reliable over relevant alternatives, they constitute knowledge. I don’t need to show that my method of gaining beliefs is in fact reliable; it only needs to be reliable. It is the internalists who think that one needs to provide cognitively accessible reasons for belief. I merely hold that it is the cause of one’s belief that matters for knowledge, not one’s rational-

A Problem for Philosophical Knowledge

89

ization for it.’’ It seems that an externalist could make a similar reply with respect to philosophical knowledge. It is a mistake to worry about showing that rational intuition is a better epistemic method than revelation or hallucinogen use; so long as our philosophical beliefs are true ones, and intuition is objectively a reliable method, then we have philosophical knowledge. If peyote and revelation are objectively unreliable methods of gaining true beliefs about philosophical propositions, then they do not produce knowledge. End of story. Unfortunately, even if externalism is true, this is not an adequate response. Being an externalist does not mean that one’s doxastic methods never need to be examined, or that questions about those methods can’t be raised. Otherwise externalism is nothing more than a sleight-ofhand that allows believers to remain complacent about any means they use to gain beliefs. Sophisticated externalists are sensitive to this issue; Goldman, for example, thinks that cognitive science will enable us to identify basic belief-forming processes which can then be evaluated in terms of reliability, power, and speed (Goldman 1986, p. 181). But how exactly is this procedure to work? Consider a ﬁrst-order belief-acquiring method M. If M is generally reliable, then its use will produce knowledge whether the believer is aware of its reliability or not. Suppose though that a curmudgeonly philosopher comes along and suggests that one ought to use belief-acquiring method N instead of M. Both methods yield beliefs about the same subject matter, and furthermore, yield mutually inconsistent beliefs. Since the results of M and N are inconsistent with each other, at best one of these methods is producing a true belief. But which one? It is sheer dogmatism for an externalist to insist on preferring M, even granted the externalist’s conditional if M is generally reliable, then it produces knowledge. What’s being called into question is the antecedent of this conditional. The externalist may then appeal to a secondorder belief-acquiring method M to adjudicate between M and N. If M is a generally reliable means of assessing ﬁrst-order belief-acquiring methods, then the use of M will yield knowledge of which method is preferable: M or N. However, the curmudgeonly philosopher will rejoin that this only pushes the problem back a step, offering N as an alternative second-order belief-acquiring method. What means shall we use to evaluate ﬁrst-order belief-acquiring methods? M or N ? Obviously any appeal to a third-order adjudication method to select between M and N will only result in an inﬁnite regress. Bonjour takes this sort of argument to be a skeptical reductio of externalism and a good reason to be an internalist (Bonjour 2002, pp.

90

Chapter 2

236–237). My conclusion is more modest: even if externalism is correct, it doesn’t provide an adequate solution to the problem facing philosophical knowledge. This is not to say that internalism fares any better; I have just argued that there do not seem to be any good reasons to prefer rational intuition as a basic belief-acquiring method to revelation or the ritual use of hallucinogens. Given an inability to show the relative superiority of rational intuition over the two other methods we have been discussing, there are three possible responses. The ﬁrst is purely epistemic: skepticism. Since we don’t know which of various competing methods is the best one to use to gain justiﬁed beliefs about philosophical propositions, if we pick the best method, it is merely a matter of luck. Therefore we have no knowledge of philosophical propositions. The second two responses are metaphysical: nihilism and relativism. Perhaps our failure to vindicate rational intuition over the competition is evidence that there are no properly philosophical propositions to be known at all. It is our attempts to acquire justiﬁed beliefs about the nonexistent that is the problem. The ﬁnal alternative, relativism, is the idea that there are knowable philosophical propositions, but which ones are true is somehow dependent on method. Given the methodology of the Jı´varo, there are nonphysical, spiritual souls, but given the methodology of rationalist, analytic philosophy, there aren’t. All three of these alternatives are disturbing ones. I will examine each of these in turn. Skepticism I have argued that the belief-acquiring methods of rational intuition, Christian revelation, and the ritual use of psychoactive drugs yield inconsistent beliefs. I have further argued that attempts to show the relative superiority of the method of intuition have all failed. One natural response at this point is to view these results as a demonstration of skepticism concerning philosophical propositions. The literature on the correct way to construe skepticism is, of course, voluminous, and I am loath to get bogged down in those thickets. However, I will hazard that skeptical arguments are generally based on the notion that S doesn’t know P because S’s true belief that P is improperly dependent on good luck. Sara doesn’t know that it is 3:30 because it is no more than luck that she glanced at the stopped clock at exactly 3:30. Smith doesn’t know that either Jones owns a Ford or Brown is in Barcelona because he is just fortunate that logical addition yielded a true disjunction from the justiﬁed yet false premise that Jones owns a Ford. Henry doesn’t know that he is seeing a barn, even though it is a barn and his true belief

A Problem for Philosophical Knowledge

91

was caused by the state of affairs that it is a barn, because it is simply luck that he is not fooled by one of the many papier-maˆche´ barns in that area. More global skepticism proceeds from the contention that we can’t perceptually discriminate among veridical sense perception and dreams, evil demons, or the neural inputs of alien scientists. Since we can’t tell the difference, if we have true beliefs based on sense perception then it is the sheerest fortune that we are not dreaming, deceived, etc. Thus our true empirical beliefs never rise up to knowledge. This is meant as a light gloss on familiar skeptical arguments; I don’t wish to be committed to any particular theory or treatment of skepticism. Perhaps the best analysis of skepticism won’t be in terms of epistemic luck, but I do think that accidental fortune in believing the truth is a convenient way to express the skeptic’s concerns. It looks as if a skeptical argument similar to those above can be deployed against knowledge of putative philosophical truths. We have no defensible reason to prefer one basic method of acquiring beliefs about philosophical propositions over another basic method that gives different results. Any true beliefs we have about philosophical propositions are accidental—it is just good fortune if we pick the right method. If my intuition-based belief that ‘‘necessarily, it is wrong to use persons merely as means and never as ends-in-themselves’’ is true, then it is no more than good luck that I chose intuition as my method instead of one of the competitors. I could have consumed Psilocybe mexicana, gone on a spirit quest, and after interpretative discussions with my tribal elders, come to believe that it is wrong to use members of our society merely as means, but that outsiders can be slain and their ﬂesh consumed merely as a means of propitiating the god Huizilopoctli. Since my choice of beliefacquiring method is ultimately arbitrary and not based in reason, I don’t know the categorical imperative or mutatis mutandis any other philosophical proposition to be true as the result of using rational intuition. In short, our failure to vindicate intuition means that we don’t know any philosophical truths. There may be truths about morality, the divine, metaphysics, and other philosophical propositions, but we’ll never know them. Intuition-driven philosophy is a waste of time; as far as we know, we’re just as well off eating teonanacatl. The skeptical position looks strong—too strong, as it turns out. The skeptical alternative appears self-defeating, and so skepticism is not a viable response to the problems so far raised. I have suggested above that accidentally true belief is not knowledge, and that some similar idea is behind most familiar skeptical arguments. Yet the proposition

92

Chapter 2

that accidentally true belief is not knowledge is justiﬁably believed only on the basis of intuitions about the sort of cases discussed two paragraphs back. The conclusion of the skeptical argument is that we can’t know any philosophical proposition as the result of rational intuition. If that is correct, then we can’t know that accidentally true belief is not knowledge as the result of intuition. Therefore we lack a good reason to be skeptics about the knowledge of philosophical propositions. The argument is easily schematizable: Argument 1 1. If skepticism about philosophical propositions is true, then we can’t know the truth of any philosophical proposition. (Deﬁnition of skepticism) 2. Skepticism is a philosophical proposition. (Premise) 3. Therefore, p: if skepticism about philosophical propositions is true, we can’t know it. (From 1, 2) Notice that it does not matter how we analyze skepticism, whether we understand it to involve accidentally true belief as presented above or in some other way. The key point is that skepticism about philosophical propositions is itself a philosophical proposition and thus itself subject to skepticism. Now, a skeptic might dispute premise two, denying that skepticism is either philosophical or propositional. However, this approach does not seem very plausible. Epistemological propositions are surely at the very heart of the rationalist enterprise; what we know and don’t know, what we are justiﬁed in believing and what we aren’t are precisely the sorts of things upon which our intuitions are brought to bear. Analyses of knowledge and justiﬁcation are defended or rebutted on the basis of possible cases involving outlandish scenarios, not on the basis of scientiﬁc ﬁndings. Pyrrhonian and Cartesian skeptics do not offer empirical reasons to doubt the deliverances of our senses; instead they present thought experiments about evil geniuses, alien scientists, and the like. Even Quine acknowledges that epistemology is philosophical; he just wants to replace it with empirical psychology. So skepticism, however it is is construed, is philosophical. Is it propositional? One might argue that skeptics don’t propose any thesis, or defend any proposition. Instead a skeptic might just be a sort of a critic who says, ‘‘I’m not claiming that I know anything, or that any particular philosophical proposition is true. You are the one who is claiming that this or that philosophical proposition might be known. All I am claiming is that by your own standards, it follows that skepticism is

A Problem for Philosophical Knowledge

93

true. Again, I am not asserting anything, I am merely showing what follows from your premises when we assume they are true.’’ The problem with this approach is that if the skeptic is truly asserting nothing, then there is nothing to rebut, literally nothing to worry about. If skepticism is not a claim or a thesis of any sort but simply a deﬁant ‘‘oh yeah?,’’ the appropriate response is ‘‘yeah!’’ Argument is not needed. But skepticism is a position taken. For one thing, the skeptical rejoinder considered above makes an entailment claim about what follows from what. For another, what exactly is supposed to follow from the putative standards? Presumably some skeptical consequence about what is not known. It is propositions that are the logical consequences of other propositions, so the skeptical conclusion is a proposition, and one that the skeptic is asserting, at least as conditional upon the premises. Put another way, here is the proposition a skeptic must be asserting: ‘‘given epistemic standard/principle/proposition A concerning knowledge of items in domain B, it follows that no one has any knowledge of items in B.’’ The skeptic needn’t accept standard A, but the skeptic is still defending a thesis, one about knowledge, one that is philosophical. In fact, things get worse. The conclusion of argument 1 is an epistemological contention, not an empirical one. That is to say, proposition p is also philosophical in nature and falls under the skeptic’s purview. Therefore, the conclusion of the previous argument must, according to the skeptic, be unknowable. Even though argument 1 is valid with all true premises, its conclusion is unknowable. To wit: Argument 2 1. If skepticism about philosophical propositions is true, then we can’t know the truth of any philosophical proposition. (Deﬁnition of skepticism) 2. p is a philosophical proposition. (Premise) 3. Therefore, if skepticism about philosophical propositions is true, then we can’t know that p. (From 1, 2) Argument 1 proves that, granted skepticism about philosophical propositions, skepticism is unknowable. Argument 2 demonstrates that the conclusion of argument 1 is itself unknowable, despite its having been proven. It now seems that the assumption of skepticism about philosophical propositions leads to a version of the knower paradox. Compare: The paradox of the knower (G) Sentence G is not known to be true.

94

Chapter 2

Is G true or false? For the purposes of reductio, let’s suppose that G is false. 1. sG (Premise) 2. sG ! s(sentence G is not known to be true). (Deﬁnition of G) 3. sG ! sentence G is known to be true. (2, double negation) 4. Sentence G is known to be true. (1, 3, modus ponens) 5. Necessarily, if a sentence is known, then it is true. (Deﬁnition of knowledge) 6. G (4, 5, modus ponens) 7. G & sG (1, 6, conjunction) Thus it must be the case that G is true. Assumption of its falsehood leads to a contradiction. So why is this a problem? Because no one can know that G is true. Yet we’ve proven that it is true. We’ve proven the truth of a perfectly comprehensible sentence that not even God could know. This is pretty mysterious; just as in the skeptical arguments above, we seem able to know the truth of the premises of an obviously valid argument, and yet the conclusion is unknowable. In the ﬁrst argument we proved that p: if skepticism about philosophical propositions is true, we can’t know it. In the second argument we proved that if skepticism about philosophical propositions is true, then we can’t know that p. Thus, under the assumption of skepticism concerning philosophical propositions, we not only can’t know that skepticism is true, but we can’t know that we can’t know it! Skepticism about philosophical propositions is not merely self-destructive; it leads to the knower paradox. Since skepticism is not only unknowable but results in a paradox, this is an excellent reason to reject the skeptical response and look for another solution. Nihilism Another possibility is that there are no philosophical propositions. As naturalists will be eager to argue, the fact that we are unable to show that rationalist philosophy is a more reliable means of acquiring beliefs than Christian revelation or the ritual use of hallucinogens constitutes a reductio ad absurdum of the very idea of nonnaturalistic philosophical knowledge. Our mistake all along was to think that there are any properly philosophical truths to be had. Once we give that up, along with the suspicious, magical-sounding belief-acquiring methods we have been considering, we will be able to see that the only knowledge to be had is empirical knowledge. Traditional intuition-driven philosophy is empty foolishness; it is the ethereal pursuit of the nonexistent.

A Problem for Philosophical Knowledge

95

There are two main ways to understand the nihilist solution. One is that propositions that are putatively philosophical and tractable only by nonscientiﬁc means are in fact secretly empirical. Concepts like ‘‘God,’’ ‘‘knowledge,’’ ‘‘object,’’ ‘‘free will,’’ ‘‘duty,’’ and so on secure their meaning through some causal series of events, ﬁnally grounded in the empirical world. Scientiﬁc discovery of these causal mechanisms will enable us to establish the truth-values of ‘‘philosophical’’ claims involving these concepts without needing to invoke rational intuition or (heaven forbid!) divine revelation or psychoactive drugs. So ‘‘philosophical’’ propositions are real propositions with determinate, unique truth-values; they just aren’t really philosophical. In other words, all truths are fundamentally scientiﬁc, and Michael Devitt is right when he claims that ‘‘[we should] reject a priori knowledge and embrace ‘naturalism,’ the view that there is only one way of knowing, the empirical way that is the basis of science’’ (1999, p. 96). The second nihilistic approach is to hold that philosophical propositions are indeed philosophical, but they aren’t really propositions. Instead sentences about the divine, morality, epistemology, and the rest are pseudo-propositions, which is to say that such sentences don’t express propositions at all, but are instead empty verbiage. At best philosophical language expresses approbation or disapprobation, or is a kind of poetry. Cicero’s praise of the Eleusinian Mysteries is best understood not as appreciation for the moral lessons learned from the drinking of the kykeo´n, but as an aesthetic appreciation of performance art. Carnap is thus right when he writes that ‘‘metaphysics is a substitute, albeit an inadequate one, for art’’ (1959, p. 80). So far I have argued that we have found no reason to prefer the method of rational intuition to either revelation or the ritual consumption of psychoactive plants as a means of gaining beliefs about philosophical propositions. Furthermore, these three methods produce conﬂicting results about what we ought to believe. The skeptical response to this problem, namely that it shows that we cannot have philosophical knowledge, leads to the knower paradox and for that reason should be rejected as unsatisfactory. The nihilist-cum-naturalist solution will be addressed more extensively in a subsequent chapter. For now it looks as if we must either ﬁnd a third alternative, or reject philosophical propositions as either unphilosophical or not really propositions. Fortunately, there is a third alternative: relativism. According to the relativist, when it comes to philosophical propositions, there is no way to decide rationally among basic belief-acquiring

96

Chapter 2

methods. Dogmatic faith in rational intuition is no better than faith that peyote will reveal the right way to live, or faith in . . . well . . . faith. If we select rational intuition as our method, certain propositions will come out necessarily true. If we choose Christian revelation, different propositions will come out true. If we pick shamanic hallucinogen consumption, still a different set of propositions will be deemed true. According to the relativist, what propositions are true is therefore dependent on, and relative to, method. There is more than one true philosophical story to be told about morality, epistemology, metaphysics, and the divine. Relativism has a bad reputation among most analytic philosophers, and defenders of Christian revelation aren’t too keen about it either.36 Yet unless a hard-core naturalism is true, relativism is the only choice we have left.

3

The Relativist Solution

It has been argued so far that there is a problem with the acquisition of philosophical knowledge, namely, there are various ways that we might come to gain philosophical beliefs, and it seems that we are unable to prefer, rationally and noncircularly, any one of these ways to the others.1 It has also been argued that skepticism about knowing philosophical propositions is not a satisfactory solution, not because such skepticism would put philosophers out of work, but because such a deep skepticism is self-applicable and ﬁnally results in the knower paradox. The other two solutions that seem available are metaphysical: either there are no philosophical propositions (another answer destined to swell the ranks of the unemployed), or relativism about philosophical claims is true. In this chapter I will defend the relativist solution as an internally consistent, coherent, and plausible answer to the problem of philosophical knowledge. Relativism has the odd property of simultaneously being a darling and a pariah among philosophical positions. Perhaps the best way to put matters is this: you either love relativism or you hate it. In contemporary scholarship, relativism is chieﬂy associated with the work of ‘‘continental’’ writers like Derrida, literary/social critics like Richard Rorty, and anthropologists desperate to avoid the taint of imperialism. There are, though, a few analytic philosophers who have come out of the closet and admitted sympathy for some form of relativism. Among these is Robert Nozick, who in his ﬁnal book before his death, wrote that ‘‘philosophically very interesting . . . versions of relativism can be formulated that are plausible. That I have reached this position surprises me’’ (Nozick 2001, p. 17). Stich, red-faced, also concedes his relativism, writing, ‘‘the account of cognitive virtue I have come to defend is . . . relativistic. . . . for a long time I was rather embarrassed to hold such a view, since it aligns me with a small minority among philosophers and

98

Chapter 3

an even smaller minority among philosophers whose work I most respect’’ (Stich 1990, p. 14). In his ﬁne book devoted to defending the coherence of relativism in metaphysics, Michael P. Lynch calls his position ‘‘metaphysical pluralism,’’ seemingly nervous about embracing the epithet of relativism (Lynch 1998).2 Relativism among analytic philosophers seems a secret vice, like a taste for trashy television shows, admitted only in one’s cups and then among friends. No doubt skeptical readers of this book are in vigorous agreement. Many think that relativism is so straightforwardly self-refuting that it deserves little further comment. Thus Putnam freely writes, ‘‘although we all know that cultural relativism is inconsistent (or say we do) I want to take the time to say again that it is inconsistent’’ (Putnam 1983, p. 236). Relativism is such an easy and tempting target, why not knock it down on the way to more important matters—rather like Nietzsche’s comment about free will: the charm of free will lies in the fact that every philosopher seems compelled to refute it.3 Nevertheless, like free will, relativism is one of those ideas that just won’t go away. As Alasdair MacIntyre (1985) notes, it’s been refuted just a little too often, a sure sign that the doctrine must embody some not-to-be-neglected truth. Not only is relativism one of the most tenacious theories about truth, but its pedigree is as old as philosophy itself. Equally venerable is the chief criticism of relativism, namely the charge that the theory is self-refuting. The Greeks knew the self-refutation objection. In Theaetetus, Plato considers the relativism of Protagoras, who reportedly began his book Truth with the words ‘‘man is the measure of all things.’’ Socrates and his interlocutor Theaetetus evaluate this statement, and Socrates argues that, taken to its logical conclusion, Protagorean relativism means that no one is any wiser than anyone else. Protagoras himself is no more knowledgeable than a baboon or a pig and yet he is as wise as any of the gods (Theaetetus 161–162). Since all who disagree with Protagoras are as equally in the right as he is, those who deny his relativism must also be in possession of the truth. This, of course, implies that Protagoras is wrong about man being the measure of all things. Yet the doctrine is that Protagoras, along with everyone else, must be right in his opinions; thus for Protagoras relativism is true. The apparent contradiction Socrates calls ‘‘a really exquisite conclusion’’ (Theaetetus 171a).4 Modern commentators can’t resist this argument. Apart from Putnam (cited above), Joseph Margolis, James Harris, Thomas Nagel, and Richard Rorty have all endorsed versions of the Socratic criticism.5 Boghossian insightfully comes close to the central arguments

The Relativist Solution

99

of the present book; he recognizes the deep conundrum of justifying basic epistemic principles and entertains the relativist solution. However, in the end he rejects relativism on the basis of a version of the selfrefutation argument (Boghossian 2001, pp. 15–19). It cannot be denied that one of the main problems that plagues relativists is that they try to bite off more than they can chew. I will attempt to curb my appetite by arguing only that a form of relativism is true for philosophical propositions, namely, that such propositions are true in some perspectives and false in others. As a prolegomena to that substantive conclusion, I will demonstrate the consistency of global relativism. Since the main objection to relativism tends to be that the global version is incoherent and self-refuting, it is worthwhile to put that objection to bed and formulate the most comprehensive sort of relativism one can consistently maintain. It will not be argued that global relativism, the contention that every proposition is true in some perspectives and false in others, is true. The chief thesis of the present book is that absolutism is false for a speciﬁc class of propositions, namely, philosophical ones. Keep in mind that it will not be argued that absolutism is false for empirical propositions, much less for all propositions. Perhaps more global forms of relativism like these are correct, but these are beyond the scope of the current enterprise. Exactly How Global Relativism Is Self-Refuting The celebrated charge of self-refutation goes roughly like this: the relativist thesis is that everything is relative (nothing is absolute). Well, what about the claim itself, that everything is relative? It must just be relative too—relative to a perspective, conceptual scheme, viewpoint, or what have you. In other words, there are perspectives in which the relativism thesis is true, and there are those in which it is untrue. After all, its truth is relative. Hence there is a perspective in which absolutism is true and so relativism is false. Yet absolutism resists conﬁnement to perspectives; it insists on extending its reach without bound. To get at the heart of this argument, let us compare the claim that everything is relative with the claim that everything is possible. Suppose that everything is possible. That is, for all F, UF. Allow F to be ‘‘it is necessarily not true that everything is possible.’’ Then the following turns out to be true: possibly, it is necessarily not true that everything is possible. A well-known theorem in modal system S5 tells us that whatever is possibly necessary is necessary. We can thereby conclude

100

Chapter 3

that it is necessarily not true that everything is possible. Thus, by reductio, it cannot be the case that everything is possible. So what should we do? Should we abandon all talk of modality, give up possibility and necessity, and purge ourselves of possible worlds? Of course not (not as the result of this problem anyway). No one is seriously prepared to claim that everything is possible. Yet everyone is prepared to afﬁrm this thesis: everything that is true is possibly true. There are two important things to note about the latter proposal. The ﬁrst is that it does not entail that nothing is necessarily true. That is, to say that everything true is possibly true does not mean that everything true is merely, or only, possibly true. Something possibly true could also be necessarily true. The second thing to note is that the thesis that everything true is possibly true does not entail that possible truth is a cheap version of real, actual truth. It may be that possible truth is weaker than actual truth, or somehow is not as good, but the matter is hardly straightforward. After all, modal realists hold precisely that possible truth is every bit as good as ‘‘actual’’ truth, or that there is no salient difference between the two. Consider then the problem faced by the relativist. Suppose that the relativist claims that everything is relative (nothing is absolute). That is, for all F, F is true relatively. Let Y be an operator that takes sentences and indexes them to perspectives, so that YF is to be read as ‘‘it is relatively true (true in some perspective) that F.’’ The claim that everything is relative is thus: for all F, YF. Let us allow ‘‘it is absolutely not true that everything is relative’’ to be a substitution instance of F. Then the following turns out to be true: relatively, it is absolutely not true that everything is relative. If we accept an S5-like theorem that whatever is relatively absolute is absolute, then it will follow straightaway that it is absolutely not true that everything is relative. And, by reductio, the relativist thesis is false. Of course, for this argument to work, we need some motivation to think that whatever is relatively absolute is absolute. Let us introduce o as an ‘‘absolute’’ operator so that oF is to be read as ‘‘it is absolutely true (true in all perspectives) that F.’’ The principle under consideration—the one that permits the reductio on everything is relative—is thus formalized: YoF ) oF. Let us call this P. Here are some semantical reﬂections on absolutism and relativism that are meant simultaneously to sharpen the standard charge of selfrefutation and show that the root intuition behind this charge lies in the acceptance of P. The relativism thesis is everything is relative. Absolutism can be characterized as a denial of this, or not everything is relative. By ‘‘everything is relative,’’ let us understand the claim that every propo-

The Relativist Solution

101

sition is true in some perspective and not true in another. Thus absolutism is: there is at least one proposition that has the same truth-value in all perspectives. Clearly, either the thesis of relativism is true absolutely (true in all perspectives) or just relatively (true in some, but not all perspectives). Suppose that relativism is true in all perspectives. If so, then there is a proposition that has the same truth-value in all perspectives— namely, the thesis of relativism itself. Yet, if there is some proposition that has the same truth-value in all perspectives, then absolutism is true. Thus, if relativism is true in all perspectives, absolutism is true; equivalently, if relativism is true in all perspectives, then by reductio relativism is not true. Suppose then that relativism is merely relatively true, that is, true in some perspectives and not true in others. Consider the latter case, the perspective in which relativism is not true. In such a perspective, call it p, not-relativism—that is, absolutism—is true. Now, absolutism is true only if there is some proposition that has the same truth-value in all perspectives. That is, in p there is some F such that oF. However, it does not seem that p could contain such a proposition. F could not be the thesis of absolutism itself, since ex hypothesi there are perspectives in which it is not true and relativism is true. Nor could F be the thesis of relativism, since ex hypothesi there are perspectives in which it is not true. Nor do any other candidates for F look promising, since—given the assumption that there are perspectives in which relativism is true— we are guaranteed that the truth-value of every proposition F will vary across perspectives. Hence, there is no proposition that is true in all perspectives; that is, for every proposition there are perspectives in which it is true and perspectives in which it is not true. Then relativism is true in all perspectives, and this, I have already shown, entails that relativism is not true. Thus it seems that relativism can be neither absolutely nor relatively true. The claim that everything is relative must be false. Let us look more closely at the situation that developed in the last paragraph. There we considered the option of relativism being relatively not true. Therefore, in some perspective there was a proposition F that was absolutely true. Formally: YoF. Yet it turned out that there could not be such a proposition since the assumption of relativism prevented any proposition from being true in all perspectives. In other words, there could not be a F such that oF. This is why YoF could not be true. The form underlying this argument is modus tollens. The conditional relied on is none other than the S5-like principle P: YoF ) oF. The preceding argument does not constitute a formal proof that P is true; rather it

102

Chapter 3

is a set of semantical considerations designed to uncover the intuition that P.6 It is a tacit acceptance of P that I suspect under girds many rejections of ‘‘relativism is absolutely false’’ as being true merely relatively. As Bennigson writes, ‘‘the phrase ‘relative to the absolutist framework’ drops out’’ (Bennigson 1999, p. 216). ‘‘Relatively’’ drops out because once a sentence is relatively absolutely true, it is absolutely true. Here are two interesting features of this analysis. First, the perennially popular attempts to save relativism from self-refutation by declaring the relativism thesis itself to be true merely relatively are completely misguided.7 If relativism is true merely relatively, then there is a perspective in which absolutism is true, and on principle P it follows that absolutism is true and thus that relativism is false. Second, it is often thought that the self-reference problem faced by relativism is a liar-like paradox.8 The preceding arguments show that it is not—it is straightforward selfcontradiction. A truly paradoxical sentence cannot have a stable truthvalue. ‘‘Everything is relative’’ does have a stable truth-value—namely, false. The Solution to Self-Refutation and Its Advantages The solution to the problem of self-refutation is as follows. Recall that it cannot be the case that everything is possible because the claim runs afoul of the principle that whatever is possibly necessary is necessary. As we have seen, it cannot be the case that everything is relative either, because the claim runs afoul of the intuitive principle that whatever is relatively absolute is absolute. However, just as we can safely assert that whatever is true is possibly true (and concomitantly whatever is untrue is possibly untrue), so too the relativist can claim that whatever is true is relatively true (and that whatever is not true is relatively not true). The relativist is free to abandon the odoriferous view that everything is relative, and instead promote the more modest view just stated. There is nothing self-contradictory or paradoxical about the claim that everything true is relatively true, just as there is no puzzle engendered by the claim that whatever is true is possibly true. As in the case of alethic modality, it is entirely consistent for the new-and-improved relativist to hold that some propositions are absolutely true, and that perspectival truth is every bit as decent and upstanding as ‘‘real’’ truth. Indeed, ‘‘real’’ truth is just truth in this perspective, just as actual truth is truth in this world. Absolute truth turns out to be truth in all perspectives, just as necessary truth is truth in all worlds. For the relativist it will be

The Relativist Solution

103

nonsense to talk about truth outside of the structure of perspectives— that is, nonperspectival or extraperspectival truth. However, formally this stricture should be no scarier than forbidding talk of truth outside of the structure of worlds once we have accepted possible world semantics. We do need arguments for thinking that truth is or ought to be indexed to perspectives (as it is to languages and worlds), but we should not be dissuaded from relativism because of worries about it being logically incoherent or self-refuting. Relativists will be loath to accept principle P since its acceptance yields that their own position is self-refuting. Nevertheless, it is not in their best interest to reject it. Honest relativists worry about the selfrefutation problem, and any adequate account of relativism should be able to not only defuse the puzzle, but also explain its power. The analysis above locates the problem at the conjunction of ‘‘everything is relative’’ and P. The revised relativism of ‘‘everything true is relatively true’’ avoids the self-refuting conﬂict with P. A relativist who rejects P in order to keep ‘‘everything is relative’’ is once again saddled with the task of explaining and solving the self-refutation problem, and thus pays a hefty price for goods of dubious merit. There are relativists who seem reluctant to give any ground at all here, but it is difﬁcult to resist the speculation that such resistance has more to do with a psychological compulsion than a philosophical one. Still, relativists will no doubt be of mixed minds about the proposed revision. They will be disgruntled that (1) the view is consistent with all truths being absolutely so, and that (2) principle P (and the other axioms and theorems of the relativist logic detailed in the appendix) is absolutely true. On the other hand, relativists should be pleased that the view proposed is also consistent with most truths being merely relatively (and not absolutely) true. It is then up to the relativists to argue that most truths are merely relatively true. They must earn through honest toil what a logic that entailed that some truths are merely relatively true would obtain by theft. I will engage in some of this toil shortly by arguing that philosophical truths are merely relatively true, and not also absolutely true. The second concern of the relativists that P is absolutely true is a bullet they must bite—under the present logical system, of course; under an S4-style relativist logic other theorems would be absolutely true. This is the compromise relativists must make if they hope to achieve any kind of rapprochement with absolutists. A relativist who is wholly unconcerned with either logical consistency or taking the claims of absolutists

104

Chapter 3

seriously will be unpersuaded by much of this chapter. There is not much that can be done about that. It is a beneﬁt of my proposal that the truth of a robust relativism does not fall out of or is not embedded in the logic of relativism. Relativists need to argue for their strong claims, not simply write up a logic in which their claims are true. Absolutists and relativists are in many ways at the same impasse that Spinozists and anti-Spinozists were at years ago. Spinoza held that all truths are so necessarily, whereas most other philosophers have demurred that at least some truths are contingent. A real advance in this debate was the development of a logic that explained possible truth, necessary truth, contingency, and the relations among these in a way that did not clearly tip the balance in favor of either Spinoza or his adversaries. This logic is of course our now-familiar alethic modal logic. Spinoza would ﬁnd ‘‘possible worlds talk’’ pointless, unnecessary, and perhaps even empty. So do philosophers at the other end of the spectrum from Spinoza, who holds that all truths are contingent (Quine in some moods). Yet these are the limiting positions on a logic that is widely accepted as providing the common ground required for meaningful dialogue about which, if any, truths are contingent and which, if any, are necessary. We would undoubtedly consider it sophistry if an anti-Spinozist merely devised a modal logic in which some truths turned out to be contingent. This would hardly persuade someone who thought all truths are necessarily true. It is a beneﬁt of the analysis of relativism offered here that absolutists as well as relativists can accept the formal system. That way all disputants can quit arguing about the selfrefutation problem, or talking past each other, and consider reasons for and against strong relativist claims on equal footing. Absolutists will also be of mixed minds. They will be aggrieved that (1) the view is consistent with most truths being merely relative (and not also absolutely true), and that (2) the semantics of this relativist logic are committed to perspectives and the indexing of truth to them. On the other hand, absolutists should be pleased that that the view proposed is formally consistent and offers a logic with absolutely true axioms and theorems. It is up to absolutists to argue that all truths are absolutely true and not merely relatively true. They too must earn through honest toil what a logic that entailed that all truths are absolutely so would obtain by theft. The second concern of the absolutists, regarding perspectives, is the bullet they must bite. It should be emphasized that this solution forges a strict and important distinction between a proposition’s being true independently of any

The Relativist Solution

105

perspective and a proposition’s being true in all perspectives. The latter characterizes absolute truth, and carves out logical space for absolute truth even while permitting a robust relativist program. The former is an antirelativism in that it rejects the idea of propositions being essentially indexed to perspectives. So, there are two ways an absolutist might go: deny that there are such things as perspectives and that truth is indexed to them, or admit that there are perspectives but maintain that the truth-value (whatever it is) of any proposition is the same in every perspective. Absolutists are likely to be tempted by the former position, and deny the existence of perspectives at all. Lynch has suggested that this is precisely what an absolutist is likely to do, since they see no point in claiming that any fact is relative to a scheme (whether that be one or all). Facts, by their very nature, don’t require relativization to anything like conceptual schemes, forms of life, perspectives, and so on. They are independent of such things. Lynch calls truth in all perspectives ‘‘virtual absolutism.’’ The real deal, he thinks, rejects indexing to perspectives tout court. This is legitimate if the denial is the result of some argument, such as Davidson’s argument against diverse conceptual schemes, but it is illegitimate as a preliminary move. That is, a haughty sneering in the direction of perspectives as an opening strategy is no more than a refusal by the absolutist to even come to the negotiating table. The relativist logic presented here is an attempt at rapprochement, a compromise that takes seriously the absolutist’s demands for rigor and consistency. By comparison, it is hard to see how one might even meaningfully discuss contingency with a Spinozist who refuses possible worlds talk from the get-go. One is tempted not to try. The logic of relativism and absolutism may be one of the things up for grabs in the fray, but there are good ecumenical reasons for preferring the logical scheme detailed here. Lynch-style absolutists who ﬁnd perspectives—even if all truths are true in all perspectives—distasteful need to argue against them directly. So, treating relativism as a kind of modality provides a simple explanation of the criticism of self-refutation and a rigorous formulation of a relativist thesis that avoids the charge. It is worth noting that this analysis also yields a simple and compelling refutation of an objection leveled by W. Newton-Smith. He writes that [One] might say . . . that it is propositions . . . that vary in truth-value [across perspectives]. But this is to take the short road to incoherence. For propositions are individuated in terms of truth-conditions. It is just incoherent to suppose that the same proposition could be true in C and false in F. (Newton-Smith 1982, pp. 107–108)

106

Chapter 3

Newton-Smith seems to be conﬂating two distinct relativist theses: relativism of content and relativism of truth.9 P. M. S. Hacker holds ﬁxed the notion of truth in order to defend a relativism of content, writing, ‘‘It is not truth that is relative to conceptual schemes, but—pleonastically— concepts. Differences between conceptual schemes result not in relative truth but in incommensurable truth—which is, I take it, what Kuhn, among others, was driving at’’ (Hacker 1996, p. 303). So for Hacker it is concepts and meanings that vary across perspectives, but once these are established, the truth-values of propositions are nonrelative. On the other hand, the relativism I have developed above holds the truth-bearers, propositions, as cross-perspectivally constant, whereas what can ﬂuctuate are their truth-values. Newton-Smith seems to be assuming that propositional content is relative to perspectives and infers from that that relativism about truth is incoherent. Since his argument is grounded in an equivocation, it is easy to see that Newton-Smith’s conclusion does not follow. It is no more incoherent to relativize the truth of propositions to perspectives given a perspectivist semantics than it is to relativize the truth of propositions to possible worlds given a possible worlds semantics or to relativize truth to languages given an array of languages. Commensurability I have argued that relativism can be made consistent if truth is relativized to perspectives in the same way that it is relativized to possible worlds in ordinary modal logic. To have a full logic of relativism, though, we will need another semantic notion, one analogous to the accessibility relation. To see why, let us ﬁrst back up and ask why modal logic needs such a relation. The answer is that it is needed in order to provide ﬂexibility about the kinds of necessity and possibility the formalism is able to represent. If we are concerned only with the widest sort of logical possibility, we can get along ﬁne without an accessibility relation.10 Yet it is useful to allow the machinery of modal logic to do more than simply model broadly logical possibility. Other sorts of possibility are interesting too—physical possibility, for example, or even possible chess moves at various stages in a game. Physical possibility differs from broadly logical possibility in that fewer worlds are accessible from the actual world for the former than for the latter. The worlds additionally excluded when modeling physical possibility are those that are inaccessible from the actual world (i.e., that are physically impossible) given the physical laws of the actual world.

The Relativist Solution

107

For the class of the physically possible and impossible to change, the laws of the actual world would have to change. Put another way, were a person to become a member of an inaccessible world (i.e., a different possible world with physical laws different from those in this world), the things that are physically possible and impossible would change for that person. The prospects for packing up and moving to another possible world are not very good, however. Much the same is true of relativism. A complete logic of relativism (such as that offered in the appendix to this chapter) must be able to express various kinds of relativistic theses. One of the things that will distinguish relativisms of differing strengths is which perspectives are commensurable ones from a given perspective, and which are not. It is the commensurability relation that ﬁlls the role of accessibility for relativism. This may seem strange—relativists have traditionally talked about different perspectives being incommensurable to each other. For example, in detailing his conception of scientiﬁc relativism, Thomas Kuhn (1970) focuses on which scientiﬁc perspectives are incommensurable to which. Kuhn maintains that the perspective in which the caloric theory of heat is true is incommensurable to the perspective in which the kinetic theory is true, the perspective of Ptolemaic astronomy is incommensurable to the Copernican one, and Newtonian physics is incommensurable to Einsteinian. However, thinking that there are differing incommensurable scientiﬁc perspectives should in no way lead us to think that there are not differing commensurable ones. For example, that Ptolemaic astronomy is incommensurable to Copernican astronomy does not preclude the possibility that Ptolemaic astronomy is commensurable to the perspective of Aristotelian physics. The set of perspectives commensurable to the perspective of Ptolemaic astronomy is simply the complement of the set of perspectives incommensurable to Ptolemaic astronomy. On Kuhn’s view, if a person were to enter into a scientiﬁc perspective incommensurable to his or her previous perspective, the class of available theories, solutions, and puzzles (and truths, presumably, although Kuhn shies away from this ﬁnal step) would change for that person. He thinks that, unlike the case of possible worlds, the prospects for packing up and moving to a previously incommensurable perspective are rather decent. This is where his discussion of crises and Gestalt switches comes into play. However, none of these features of Kuhn’s theory should lead us to think that his scientiﬁc relativism cannot be captured or modeled by the logic of relativism developed here.

108

Chapter 3

Suppose we tinker a bit with the range of the commensurability relation and compare the result with Kuhn’s scientiﬁc relativism. Let’s say that all scientiﬁc perspectives are commensurable with each other— the perspectives of Newtonian physics, Einsteinian physics, Copernican astronomy, Ptolemaic astronomy, and so on are all commensurable. However, scientiﬁc perspectives are incommensurable to religious perspectives. Thus modern cosmology is incommensurable to Old Testament cosmology, Darwinian biology is incommensurable to Christian creationism, and so on. To change from being a fundamentalist Christian to being a thoroughgoing proponent of contemporary science requires a Gestalt-like paradigm shift. The same is true of the reverse: it takes a Gestalt switch to change from an atheistic scientiﬁc perspective to a deeply religious one. Now, one might argue that the sort of relativism sketched out here is not plausible, or not as plausible as Kuhn’s, but it is plainly a coherent and at least somewhat intuitive alternative. It does not matter. The point is that just as changing the range of the accessibility relation changes the type of possibility being represented, so too it is through changing the range of the commensurability relation that the type of relativism being modeled changes. What are commensurability and incommensurability anyway? Are they really connected with language and translatability as many have thought? There may well be a valuable notion of commensurability that has to do with languages and translatability. However, any relativist adopting the relativist logic developed here need not be concerned with this sort. In fact, there is no reason to be concerned with any kind of ‘‘analysis’’ of commensurability. The reason is that there is no deep fact about commensurability. The relation is invented to service the needs of the logic. To say that a perspective p 0 is commensurable to a perspective p is to say, roughly, that p 0 is a consistent or compatible perspective with p given certain facts about p. What these facts are will depend on the type of relativism that is being modeled. Compare: there is no deep fact about the accessibility relation that one ought to worry about discovering. That is, to say that a world w 0 is accessible from a world w only means that w 0 is a possible world relative to certain facts about w. What these facts are will depend on the type of possibility that is being modeled. For example, if the sort of possibility being represented is ‘‘possible moves of my rook in this chess game,’’ the salient facts will be the rules of chess, the position of the pieces in this game, and whose move it is. The point is that accessibility was invented to explain or account for this type of relative possibility. It is just a concept that is designed to play

The Relativist Solution

109

a role in the logic of possibility and necessity. It is certainly not a concept ‘‘out in the world,’’ one that is part of our ordinary stock of concepts, or one that merits conceptual investigation. So too with commensurability. It is a concept that exists by design, whose sole nature is consumed by the function or role that it plays in the logic of relativism. Perspectives What about the other semantic notion a relativist logic requires, namely the idea of a perspective, or a conceptual scheme? I have little to add to the remarks of others in this literature by way of an intuitive presentation. Despite the supposed unclarity of the idea of a perspective, most philosophers seem to have a grasp of it. Michael Krausz and Jack W. Meiland write: [The notion of a perspective] is so pervasive in our intellectual life that it is a major element in the thinking of philosophers who are in other respects radically different from one another. For example, the neopositivist Rudolf Carnap speaks of ‘‘linguistic frameworks’’; the neo-Kantian C. I. Lewis speaks of ‘‘networks of categories’’; and the later Wittgenstein talks about ‘‘forms of life.’’ But they are all talking about conceptual schemes in a broader or narrower sense of that expression. (Krausz and Meiland 1982, pp. 7–8)

Many others, including Rescher, Goodman, Kuhn, Whorf, and Nietzsche, might be added to this list. Indeed, the idea that there is a feminist perspective or an African-American one is the commonest currency of contemporary philosophy. Yet perhaps all these writers are mistaken in thinking that the concept of a perspective is a cogent one. Let us consider for a moment Davidson’s famous criticism of the intelligibility of diverse conceptual schemes. His argument goes something like this: 1. Commensurability is to be understood in terms of translatability; incommensurability is to be understood in terms of untranslatability.11 2. Conceptual schemes are to be individuated in terms of incommensurability (but not commensurability). That is, only incommensurability divides conceptual schemes. There cannot be two diverse yet commensurable perspectives.12 3. There is no such thing as untranslatability. 4. Thus there is no such thing as incommensurability. 5. Thus nothing separates conceptual schemes. 6. With no way of individuating perspectives, there is no reason to accept their existence.

110

Chapter 3

I am happy to agree that his conclusions 4 through 6 follow from his premises. Premise 3 is the one Davidson spends the bulk of his essay defending, and I grant it for the sake of argument. It is premises 1 and 2 that must be rejected. I have already argued that the commensurability relation (and hence incommensurability) is best understood as a stipulative relation that simply plays a functional role in the logic of relativism. To be sure, Davidson has ﬁne historical grounds for interpreting commensurability the way he does. Nevertheless, relativists are not obliged to buy into varieties of commensurability that involve translatability or other sorts of heavier theoretical baggage. Premise 2 claims that without incommensurability there is no way to make decent sense out of the notion of a perspective. Again, historical reasons aside, there is little motivation to accept this. Compare: the concept of a possible world is perfectly intelligible without (and was intelligible long before the advent of) an accessibility relation. Likewise it is hard to see why Davidson will only allow incommensurability between perspectives. Just as we are willing to agree that there are possible worlds that are both nonidentical with the actual world and yet accessible from the actual world, so too we should admit perspectives commensurable to (yet nonidentical with) one’s actual perspective. The upshot is that Davidson’s insistence on hitching the viability of perspectives to translatability issues is unnecessary and misleading. Once that move is rejected, his argument may be safely set aside. Other relativists have also rejected Davidson’s argument, albeit for different reasons. For the most part, however, they remain attached to a view of perspectives as inextricably linked with concept possession. For example, Lynch defends a view of conceptual schemes as networks of concepts used in propositions, networks that differ from each other to the extent that they differ in basic concepts. Conceptual schemes are internally consistent, but the component concepts can be vague and the propositions in which they ﬁgure can be indeterminate in truth-value. So far, this is a familiar story—perspectives involve concepts and conceptual constructs that meet some sort of minimal consistency constraints. For Lynch, conceptual schemes are a component part of a larger category, that of worldviews. Worldviews include additional factors, including ‘‘beliefs and the concepts we employ in forming our beliefs . . . the interests we have that help explain why we have those concepts, the values that guide those interests, and the underlying practices and capacities that limit and deﬁne our cognitive production and intake’’ (Lynch 1998, p. 51). Here Lynch goes well beyond the usual metaphysical/

The Relativist Solution

111

linguistic discussion of concepts to an epistemic account of worldviews. Essential to Lynch’s notion of a worldview are, in the terminology I have been using, basic belief-acquiring methods—how people come to believe the things that they do, which beliefs are noninferential, and how they give rise to an integrated, comprehensive system of understanding the world. These epistemic factors are, I think, more of a key to understanding perspectives than concepts. I suspect that the ‘‘concept’’ view of perspectives is rooted in an understanding of incommensurability as involving untranslatable concepts. I have already argued that commensurability is a purely formal relation, and that we needn’t understand it in terms of translatability or alien concepts. So I think some of the motivation for treating perspectives in terms of concepts falls away. As we saw in the previous chapter, different noninferential belief-acquiring methods are apt to produce mutually inconsistent results. Thus the method of the ritual consumption of yaje´ by Tukano shamans tends either to generate a basic belief that there are noncorporeal souls, or produce basic beliefs that, once systematized and brought into reﬂective equilibrium, have as an inferential consequence that there are noncorporeal souls. In the case of rationalist philosophy, the method of rational intuition, coupled with subsequent critical reﬂection designed to achieve reﬂective equilibrium, produces the belief that there is no such thing as a disembodied soul. There is an incommensurability here between the pro-soul beliefs of the Tukano and the anti-soul beliefs of most rationalist philosophers— not in the sense of untranslatable concepts, but in the sense that there is no possibility of convergence. Philosophers who eschew the drinking of yaje´ as a legitimate means of coming to acquire beliefs about the nature of the mind are not going to arrive at the same set of beliefs as a Tukano shaman. Their respective basic methods yield incompatible results, and, as I have previously argued, there is no good argument for privileging rational intuition over competing noninferential methods like ritualistic yaje´ consumption. Tukano shamans accept a mutually supporting set of beliefs that is different from the philosophical systems of intuition-driven philosophers. Moreover, the elements of a Tukano shaman’s set of philosophical beliefs are largely inconsistent with those of a rationalist philosopher. Shamanistic hallucinogen methods thereby produce a different worldview (at least with respect to philosophical propositions) from that of rationalist philosophers. Kuhn was on the right track when he emphasized the idea that different conceptual schemes give rise to different kinds of questions and

112

Chapter 3

different possible answers. Clearly, a secular philosopher who relies solely on the method of rational intuition to acquire basic beliefs about philosophical issues is not going to be troubled about which is the correct doctrine of atonement, or how we should best understand original sin, or whether salvation is achieved through good works or faith alone. In one sense, these issues are sensible to a philosopher who rejects revelation as an authentic belief-acquiring method. Listening to people discussing these topics is not the same as listening to someone speak in an unknown language, or hearing barking dogs. Yet there is another sense in which these issues seem empty, a kind of stylized cultural wordplay reminiscent of the glass bead game in Hesse’s Magister Ludi. Perhaps this sensation might be more vivid if readers of this book try to take seriously—that is, as a genuine contribution to the correct understanding of the world—the peyote-induced ethical debates of the Huichol, or the ﬁne points of the theology of the Rig-Veda (‘‘The swollen men piss the ﬂowing Soma. The lords, with full bladders, piss Soma quick with movement’’ [Mandala IX 74]). The question ‘‘how can I get into Heaven?’’ has a meaning within the Christian perspective, a meaning that even an atheist can recognize. Likewise, the range of possible answers to the question includes ones that are comprehensible to those not adopting a Christian perspective. The answers ‘‘salvation is preordained by God, there is nothing you can do’’ and ‘‘personally accepting Jesus as Lord is your ticket to paradise’’ are perfectly understandable answers to an atheist, just not ones that could be accepted as true description of reality. It is rather like the case of a chess player who can understand the question ‘‘How can I get crowned?’’ as perfectly sensible within the perspective of a game of checkers. Likewise the answer ‘‘Land your piece on my back row to get crowned’’ isn’t nonsense or gibberish to a chess player, yet is totally inapplicable within the conﬁnes of chess. The truth of ‘‘To get crowned, land your piece on my back row’’ is relative to speciﬁc games; it is true in checkers and false in chess. In the case of games, they are deﬁned by their rules. In the case of comprehensive philosophical belief systems, they are bounded by the mechanisms by which basic beliefs are generated. Another analogy is that of ﬁction. A sentence like ‘‘beam me up Scotty, there’s no intelligent life here’’ is perfectly comprehensible when seen on a bumper sticker. It would be silly to argue that ‘‘to beam’’ is not a proper English verb and so the bumper sticker means nothing in English, nor is it likely to be translatable into English. Within the

The Relativist Solution

113

conﬁnes of the ﬁctional Star Trek universe, the sentence has a reality, a seriousness to it that it lacks in the actual world. In the actual world it is a light-hearted joke. Likewise, within the perspective of initiates into the Eleusinian Mysteries, the story of Persephone and Hades in the Hymn to Demeter is a description of reality. WWDD—What Would Demeter Do? Perhaps an authentic question of moral guidance for Cicero, although not for John Paul II. From the Christian perspective, the story of Persephone’s escape from Hades is a pagan myth, a ﬁction suitable for amusement but not ediﬁcation, whereas the story of Jesus’ escape from Hell is to be shelved with nonﬁction, along with scientiﬁc reports. Perspectives themselves I propose we treat in the same manner as possible worlds, namely, as abstract intensional objects. They are ways of knowing. This characterization does not beg the question in favor of relativism, as one might consistently maintain that all propositions have the same truth-values in all perspectives. Thus the existence of perspectives as ways of knowing is perfectly compatible with global absolutism. Furthermore, not every belief-acquiring method marks off a perspective. Forming weather beliefs based on the Weather Channel forecast instead of the forecast in the local paper is a method of gaining weather-related beliefs, but it alone is not a perspective. Suppose one adds that Weather Channel–based beliefs are not accepted dispositively, but must be brought into accord with other related beliefs and perceptions. For example, to be believed, the Weather Channel report must be consistent with one’s current perception of the weather, the report must be generally congruent with possible weather (forecasts of 10,000 C temperatures over the next 24 hours are extremely unlikely to be accurate), the weather report must conform with known physical law, and so on. In other words, reason must operate to take and evaluate the data provided by the Weather Channel and integrate it with one’s other empirical beliefs in an effort to achieve reﬂective equilibrium. At a certain point, consulting the Weather Channel is no longer an isolated, free-ﬂoating method of gaining weather beliefs, but becomes a component method of a larger, more comprehensive perspective concerned with gaining empirical knowledge. Belief-acquiring methods are similar to data points. Having one or two is not enough to plot a curve, but as more are amassed, the pattern begins to emerge. Patterns themselves are abstract objects, as are perspectives. Writers like Alston and Bealer are sensitive to the need for our belief-acquiring methods to be integrated into more comprehensive

114

Chapter 3

perspectives, although they do not use this language. Recall that Alston stresses that an isolated instance of a belief formed by revelation is not enough for that belief to be retained or for others to inferentially accept the putative revelation. The revelation must ﬁt into Christian doxastic practice. A doxastic practice is ‘‘the exercise of a system or constellation of belief-forming habits or mechanisms, each realizing a function that yields beliefs with a certain kind of content form inputs of a certain type’’ (Alston 1991, p. 155). Doxastic practices for Alston are ultimately abstract; concrete speciﬁc ones realize abstract functions. What makes a particular doxastic practice a Christian one is that it ‘‘takes the Bible, the ecumenical councils of the undivided church, Christian experience through the ages, Christian thought, and more generally the Christian tradition as normative sources of its overrider system’’ (Alston 1991, p. 193). It is interesting that Alston, who lays so much stress on the irreducible plurality of doxastic practices, eschews the relativist implications.13 We will see shortly that this is easier said than done. Merely trotting out the Quinean chestnut about intensions being creatures of darkness (Quine 1976c) is not enough to dismiss this approach. So far we have seen that there is much to be gained by treating relativism as a kind of modality analogous to possibility and necessity. Perspectives as abstracta is the natural continuation of this analogy. Opinions vary tremendously as to what possible worlds are—sets of propositions, states of affairs, properties, and so on.14 Still, this deliberation does not mean that there are no possible worlds, or that we should be skeptics about possible world semantics. Similarly we should not take a lack of nice and neat individuation criteria for perspectives as sufﬁcient grounds for skepticism. As Chris Swoyer notes, we are happy to accept many concepts (games and tables, for example) for which reductive analyses are not easily found (Swoyer 1982, p. 88). Alternatively, one might argue that, given the similarities between perspectives and possible worlds, perhaps there is no difference between the two. Maybe relativism is just another manifestation of ordinary alethic modality. This objection raises a good question (viz., what are the ontological differences between possible worlds and perspectives?) but gives a bad answer (viz., there are none). The considerations that motivate the belief in possible worlds are completely distinct from those that motivate the belief in perspectives. A skeptic about perspectives will need to do much more work to show that these two concepts are ‘‘really the same’’ or that one is reducible to the other.

The Relativist Solution

115

Initial Objections Shogenji Tomoji Shogenji has argued that the proof I gave of the self-refuting character of ‘‘everything is relative’’ fails if one uses an S4 relativist logic instead of S5 (Shogenji 1997). This in itself should come as no surprise; the choice of logical system is crucial for all proofs. Even modus ponens fails if we abandon traditional logic for certain relevance logics. Two questions immediately arise: (1) should an S4 relativist logic be employed instead of S5?, and (2) can ‘‘everything is relative’’ be proven false in an S4 relativist logic? Shogenji argues that the answer to the ﬁrst question is yes. His argument proceeds by questioning the symmetry of the commensurability relation, a requirement of the formal logic of relativism presented in the appendix to this chapter. To my mind, however, the rejection of symmetry is too hasty, and implicitly relies on a dubious and undefended concept of a perspective. No matter. The answer to question (2) is also yes, as Shogenji admits in a footnote (Shogenji 1997, p. 476, n. 3). Since ‘‘everything is relative’’ is false in both an S5 relativist logic (like the one presented here) and an S4 one, it is pretty hard to see the great advantage the defender of ‘‘everything is relative’’ gains by going the S4 route. His central thesis is false in both. In fact, Shogenji argues that ‘‘everything is relative’’ turns out false no matter which modal system is supplemented with ‘‘absolutely’’ and ‘‘relatively’’ operators. What more is needed to persuade us that ‘‘everything is relative’’ is false and to look about instead for which logical system will permit the strongest relativistic claims? Strangely, Shogenji is still not convinced. Instead, he concludes that ‘‘this only means that a global relativist should not impose a single logic on all perspectives as a universal tool of truth evaluation’’ (Shogenji 1997, p. 476, n. 3). I am not quite sure what this means, but it sounds like a champion of ‘‘everything is relative’’ is advised cagily to avoid being pinned down as to choice of logical system, in the knowledge that as soon as she settles on one, her thesis is false. I leave to the reader an evaluation of the philosophical plausibility of this approach. Ko¨lbel Max Ko¨lbel (1999) has also raised some objections to the present account of relativism. Ko¨lbel has claimed that the argument previously

116

Chapter 3

presented to demonstrate the self-refuting nature of ‘‘everything is relative’’ is fallacious (Ko¨lbel 1999, p. 94). The argument I gave was a dilemma: either ‘‘everything is relative’’ is absolutely true or it is relatively true. If it is absolutely true, then ‘‘everything is relative’’ is false. If it is relatively true, then ‘‘everything is relative’’ is false. Ko¨lbel agrees with the ﬁrst horn of the dilemma, but rejects the second. His argument is that the second horn of the dilemma goes through only if one accepts principle P (the S5-like principle that whatever is relatively absolutely true is absolutely true), which is thus a perfectly good reason to reject it. Ko¨lbel’s rebuttal is a poor argument for several reasons. First, I explicitly argue that ‘‘everything is relative’’ can be proven to be selfrefuting on the assumption that whatever is relatively absolutely true is absolutely true. That is, I list P as one of the premises of the argument. Ko¨lbel thinks that this makes the dilemma argument fallacious. This is certainly not true—in a fallacious argument the conclusion does not follow from the premises due to equivocation, amphiboly, afﬁrming the consequent, or something similar. Ko¨lbel is merely arguing that relativists should reject one of my premises, precisely because the premises really do entail the conclusion that ‘‘everything is relative’’ is false! In short, Ko¨lbel concedes that the dilemma argument is valid, but for some reason thinks that since it relies on a premise that he dislikes, the argument is fallacious. By similar reasoning, one assumes that Ko¨lbel ﬁnds all uncongenial arguments to be fallacies, as they would contain premises that he disputes. Why does Ko¨lbel reject principle P? He thinks that the acceptance of this principle is undermotivated and that relativists ought to reject it since it entails that their own position is self-refuting. I have already conceded that it is impossible to present a formal proof of principle P without making certain other assumptions, such as taking commensurability as an equivalence relation. I did present a intuitive semantic case for P— it seems to be at the root of the widespread sense that ‘‘everything is relative’’ is self-refuting, so by identifying P as the culprit, we can explain in a plausible and rigorous way why it is a mistake to maintain ‘‘everything is relative.’’ A relativist could of course deny principle P, but they would need an argument against it. The only argument that Ko¨lbel seems to provide against the principle is question-begging, namely a relativist insistence that since ‘‘everything is relative’’ is true, by modus tollens P must be false. Apart from assuming the very thing that needs to be proven, a relativist who wishes to keep ‘‘everything is relative’’ while rejecting P is

The Relativist Solution

117

again faced with the challenges of explaining and defusing the appeal of the Socratic objection, and being able to show the consistency of his own view in a logically rigorous fashion. I am not, of course, claiming that a defender of ‘‘everything is relative’’ must show why his own view is selfdefeating. Rather, the Socratic objection is similar to the theological problem of evil. The problem of evil is an equally ancient and compelling argument, and theologians concerned with argument and consistency (those in the Tertullian and Barthian tradition presumably are not) need to show ﬁrst why the problem of evil has such prima facie appeal and second how their theodicy can escape it. A facile rebuff of the problem of evil, like saying that since there is a morally perfect, omniscient, omnipotent god, therefore worldly evil must all be part of some greater plan, will not persuade a religious skeptic. Likewise a too-easy dismissal of the self-refutation objection (such as a simple denial of principle P) is going to create more skeptics of relativism than converts. These are serious challenges that Ko¨lbel has not met. Percival Philip Percival has argued that there is a conceptual difﬁculty with relativism other than the usual self-refuting kind.15 It is worthwhile to show how his concern is not telling against the relativist logic offered here. Percival’s main argument is that there is no ‘‘clear statement of [relativism’s] consequences for the evaluation of utterances,’’ and thus ‘‘it is empty and worthless’’ (1994, p. 208). In explaining this charge, Percival makes it plain that he is concerned with how to resolve a conﬂict over the truth of a proposition between two parties who (apparently) disagree. If a person A believes that proposition p is true, whereas B believes it to be false, Percival claims that B’s belief obliges her ‘‘to insist that A withdraw his utterance’’ (1994, p. 209). So far so good. The problem, thinks Percival, arises if A holds p to be true relative to some perspective Y and B rejects p relative to some different perspective Z. He writes, ‘‘how can I believe both that the aims given A, for him, by the language he employs were successfully pursued, and that I have every right to force him to withdraw his utterance?’’ (ibid.). The answer is that he cannot. Percival’s problem only arises if one presupposes an absolutist conception of truth. His mistake is in thinking that under relativism, B is really obliged to insist that A change his mind about p. If p is merely relatively true, then A’s assertion of p can, given his perspective, be a successful speech act, and B’s assertion of not-p can, given her perspective, also be a successful speech act. The

118

Chapter 3

conﬂict between A and B is merely illusory; they are in fact debating at cross-purposes. There is genuine disagreement only if both A and B adopt the same perspective with regard to p. It is clearly an option of the relativism I present that two people could adopt the same perspective, although this is a possibility Percival oddly fails to recognize. As far as the matter of the evaluation of a truth claim under relativism, such a claim can be properly evaluated only once the perspective in which it is made is established. It is true that the issue of what exactly a perspective is and how one is identiﬁed arises at this point. I have already stated that explicating this is a key task faced by relativists. Yet the fact that there is some further question regarding to perspectives hardly makes relativism ‘‘empty and worthless,’’ as Percival thinks. All of philosophy would be empty and worthless if every concept with a bit of mystery were thus rejected. Moreover, there is some explanation at hand with the analogy to possible worlds. Propositions are to be evaluated with respect to perspectives in the same way, formally, as they are to possible worlds. Percival also ﬁnds the matter of possible worlds obscure (1994, pp. 210–211), but if relativism is no worse off than possible worlds semantics, then much has been accomplished. So far I have not presented a defense of relativism. To sum up the chapter to this point, I have argued that logically consistent versions of global relativism can be formulated. The discussion I believe has borne much fruit: we can clearly and rigorously understand truth as relative and truth as absolute and see that they are not mutually exclusive; the commensurability relation is not problematic and mysterious, but just an explanatory tool on a par with accessibility; and we have a way to avoid Davidson’s criticisms of the very idea of a conceptual scheme. Yet these are mostly pleasant side effects. Most importantly, the theory developed resolves the 2500-year-old charge that global relativism is selfrefuting. It is entirely consistent for a relativist to assert that whatever is true is relatively true and that whatever is false is relatively false. This is a kind of compromise position, and no one likes a compromise. Relativists must accept that the strongest logically consistent relativism will include some absolute truths (e.g., the axioms and theorems of RL), and absolutists must countenance an ontology that includes perspectives. Hopefully the disputants will prefer de´tente to dogmatism, move past the ancient debate about self-refutation, and proceed to the consideration of whether relativism is the correct theory of truth. I will argue that a speciﬁc local relativism is true. I leave to the more ambitious a defense of global relativism about all propositions.

The Relativist Solution

119

A Defense of Relativism It has been demonstrated that it is possible to formulate very sweeping forms of relativism that are logically consistent. The present question to consider is whether any particular form of relativism is true, or at least whether there are good reasons to ﬁnd some form of relativism likely. Perhaps the most widely discussed kind of relativism (after global relativism) is that empirical propositions are merely relatively true, true in some perspectives but not in all. This is the relativism defended by Kuhn and Feyerabend and so commonly pilloried—even parodied—by philosophers of science.16 I have no plans to defend empirical relativism; I don’t see how to mount an original defense of it, and am doubtful that it is true anyway. However, I do think that there are good reasons to accept relativism about philosophical propositions. Suppose that all philosophical propositions are unknowable, that is, skepticism about philosophical propositions is true. I argued in the previous chapter that since skepticism is itself a philosophical claim, it leads directly to the knower paradox and that this result is an excellent reason to reject skepticism about philosophical propositions. Thus there are knowable philosophical propositions. Now, it is a logical truth that either all knowable philosophical propositions are absolutely true, or some are merely relatively true. Let’s consider the ﬁrst disjunct—all knowable philosophical propositions are absolutely true. This means that all knowable propositions about morality, the nature of the mind, the existence and properties of the divine, the scope and limits of human knowledge, the existence of abstract objects, and all the rest have the same truth-value in every perspective. Perspectives, recall, are ways of knowing; they are abstract systems or patterns of belief-acquiring procedures and mechanisms. Part of what individuates perspectives is which methods of gaining basic, noninferential beliefs are constituents of the perspective. As we have seen, from the perspective of analytic rationalist philosophy, rational intuition is a means of gaining noninferential beliefs about philosophical propositions. From a Christian perspective, this method is supplemented, or perhaps replaced (there may be more than one Christian perspective) by the method of revelation. In the case of various shamanistic perspectives, ritual use of hallucinogens is a way to gain basic beliefs about philosophical propositions. It was argued in the previous chapter that these different basic belief-acquiring methods, each components of distinct perspectives, produce inconsistent results. Since they are basic methods, there is no

120

Chapter 3

neutral ground, no independent objective source to which one might defer. Ritual use of hallucinogens may not, by the lights of rational intuition, seem like a good way of gaining moral beliefs, but this is irrelevant unless it can be demonstrated that intuition has the upper hand, that it is somehow the court of ﬁnal appeal. Yet it was argued in the previous chapter that there is no justiﬁable way to privilege rational intuition over the competition. So there is just as much reason to think that revelation or yaje´ yield knowledge as there is to think that rational intuition does. To put the argument another way, either at least one of these methods yields knowledge of philosophical propositions or none do. If none do, then skepticism about philosophical propositions is true. Since we have already seen that this sort of skepticism is untenable, by disjunctive syllogism we can conclude that at least one method of gaining beliefs about philosophical matters yields knowledge. Further suppose that at most one of the three belief-acquiring methods that we have been considering yields knowledge. Let’s call this uniquely successful method M. Thus the other methods for gaining philosophical beliefs do not produce knowledge—just false (or true) beliefs that are not appropriately related to the truth. Now, if no rational grounds can be provided for trusting and relying on M over the competition, then again skepticism is the result. This was defended in the previous chapter; a hallmark of skeptical arguments is the contention that a true belief is the result of luck, that fortuitous circumstances play an indispensable role in a belief-acquiring method’s success. It is this element of epistemic luck that undermines a belief’s pretensions to being knowledge. So if no reasons can be provided for preferring M, but nevertheless we choose to go with M and it really does lead to true beliefs about philosophical matters, this is a matter of luck. Picking belief-acquiring methods by eeny-meeny-miny-moe is not a path to knowledge. It is a route to skepticism. Yet skepticism about philosophical propositions is false; it is selfundermining and leads to the knower paradox. Therefore, by modus tollens, it is the case that rational grounds can be provided for trusting and relying on method M (whatever it is) instead of competitor methods. However, it has been argued at length that neither rational intuition, nor Christian revelation, nor the shamanistic use of hallucinogens, nor anything else can be shown to be the uniquely preferable method M. Thus we have arrived at a contradiction: there are rational grounds for preferring one method of gaining philosophical beliefs to the others, and also there are no such grounds. This contradiction was derived from the as-

The Relativist Solution

121

sumption that at most one of the three belief-acquiring methods that we have been discussing produces knowledge of philosophical propositions. Therefore, by reductio, it is not the case that at most one of the three methods produces knowledge. Since it was earlier established that there is at least one method of gaining knowledge about philosophical issues (on pain of the now-discredited skepticism), we can conjoin the two results. To wit: there is at least one method of acquiring philosophical knowledge and there is more than one method. There is one further step to the relativist conclusion. It has just been shown that there is more than one method of gaining philosophical knowledge. However, it has also been argued that different methods of acquiring philosophical beliefs are in competition with each other and produce inconsistent results. Thus if methods A and B both produce philosophical knowledge, and they have inconsistent results, it follows that method A can generate knowledge that p, and method B can yield knowledge that not-p. If skepticism isn’t an option (and it has been argued that it isn’t) and there really are philosophical propositions, relativism is the only way to ameliorate this result. Proposition p is true in one perspective, and can be known by a method A indigenous to that perspective, and not-p is true in a different perspective, and is knowable by method B, a component method of that perspective. Further Objections Objection 1 A defender of absolutism might respond this way. Grant the conclusion above that there is at least one method of acquiring philosophical knowledge and that there is more than one method. Something similar is clearly true in the case of empirical knowledge. There is more than one method by which one might come have empirical knowledge. Moreover, these methods can yield inconsistent results. Suppose one hears someone calling his name, but looks around and sees no one there. Thus the method of hearing leads to a belief that p: someone is calling my name, whereas the method of sight leads to a belief that not-p: no one is calling my name. Yet the resulting inconsistency is no reason to adopt relativism, and think that there is no absolute fact about whether there is a person calling one’s name. There are different methods of gaining empirical knowledge and they may produce inconsistent results, but this hardly leads to relativism about empirical propositions, and it shouldn’t lead to relativism about philosophical propositions either.

122

Chapter 3

In response, let’s consider how an absolutist about empirical propositions would resolve the prima facie inconsistency between the hearingproduced belief that someone is calling one’s name and the sightproduced belief that no one is calling one’s name. One thing that might be done is to use further methods to acquiring more information. For example, one might shout out ‘‘Halloo’’ and wait for a reply. If none is forthcoming, then that is evidence that one’s ears are playing tricks and no one is there. Or if there is a reply, then that is evidence that one has not looked around carefully enough to spot the caller. One could also take into account mitigating circumstances, such as the background noise level, how crowded the area is, whether it is dark or well-lit, and so on. In short, the typical procedure by which one tries to eradicate an inconsistency in one’s set of beliefs is to suspend judgment about the proposition at hand and search about for further evidence that makes either p or not-p more probable. One gathers more evidence and adjusts one’s beliefs accordingly in an effort to attain reﬂective equilibrium. Notice that the kinds of acceptable evidence and belief-acquiring procedures in the example above are of the same kind. They are empirical methods based in perception and observation. Within the perspective of empirical science there are acceptable procedures and belief-acquiring methods for gaining beliefs about the physical world. It is not scientifically respectable, in order to determine whether someone is really calling one’s name, to ingest peyote, go on a spirit quest and inquire as to whether one’s dead ancestors are attempting communication. Similarly, practitioners of traditional rationalist philosophy, when confronted with conﬂicting intuitions, proceed to devise new thought experiments, propose new counterexamples, and come up with original theoretical ways of accommodating and explaining the various data of rational intuition. Conﬂicts, inconsistencies within the perspective of rationalist philosophy, are largely tractable. There is a well-established and time-honored procedure for addressing and resolving them. The method of example and counterexample is an acceptable way to winnow out proposals in rationalist philosophy, but rolling dice, for example, is not. In short, it is only within a given perspective, with its network of belief-acquiring methods and evaluative strategies, that there is a way to address problems and conﬂicts among the data. Pointing to the fact that intraperspective there are procedures for resolving prima facie inconsistencies among the results of different belief-acquiring methods does not show that interperspective that there is a uniform way of resolving inconsistencies. It is the problem of interperspective conﬂict that seems

The Relativist Solution

123

soluble only by relativism. There is no metaperspective, or absolute cross-platform epistemic standard, to which we might appeal in order to determine whether Christians are right that there are nonphysical immortal souls, or whether the majority of rationalist philosophers are right and there are no such things. Each belief seems justiﬁed and supported by the basic belief-acquiring methods and subsequent ratiocinating intrinsic to their respective perspective. Objection 2 I have argued that it is philosophical propositions that are merely relatively true. The nature of philosophical propositions themselves I have left a bit vague; I have argued that when true they are necessarily true and when false necessarily false, although there are necessities that are not philosophical. What bounds this subset as being properly philosophical is terribly hard to deﬁne, and I have not tried, settling for ostensively indicating propositions from ethics, aesthetics, the philosophy of mind, and metaphysics as exemplars. However, one might object that the boundaries between the properly philosophical and the empirical have been historically negotiable. Consider a proposition about the best course of action to cure a case of pneumonia. Rationalist philosophers will consider such a proposition to be an empirical, scientiﬁc matter and not one to be addressed by reﬂecting on a priori intuitions. Yet it may be that there are shamanistic societies in which hallucinogens are to be consumed to discover the best way of propitiating the evil spirits that cause pneumonia and thus effecting a cure. Likewise, Christians have often relied on revelation to form beliefs about what rationalist philosophers would consider empirical matters—the origin of species, for example, or end-of-the-world prophesies. But then the problem becomes pressing: how are we to tell which propositions are properly philosophical and thus subject to relativism and which ones are not, and instead absolutely true or false? I agree with all the points of this objection. However, I do not think it is a great difﬁculty for the defense of relativism that I am providing. True, it would be a problem if there were no properly philosophical propositions at all, and everything were empirical. This is the claim of the naturalists and will be extensively addressed in the subsequent chapter. I think it is a virtue of the relativist thesis currently being defended that it is ﬂexible enough to allow for debate and argument about precisely which propositions are merely relatively true and which are absolutely true. Even among rationalists there is disagreement about which

124

Chapter 3

questions belong to the empiricists and which remain to the philosophers. How much of the theory of content should be construed as empirical, for example? It seems to be an open question. Yet as long as there are any propositions that remain in the philosophical side of the house, it is these that are subject to relativism. The ones that remain are normative propositions, propositions about the divine, the nature of the mind, epistemology, and ontology. These seem to be the meat-and-potatoes of rationalist inquiry. I freely concede that I do not know how to draw a bright line between the philosophical and the empirical, but I don’t need to. If you think that any philosophical propositions are nonempirical (draw up your own list) then relativism comes into play. The hypothetical objector might rejoin at this juncture that relativism about philosophical propositions might well lead to relativism about empirical propositions after all, thus obviating any need for a distinction between the two. Even more, this would provide a good reason to reject the account of relativism offered here. If philosophical relativism entails empirical relativism and empirical relativism is false, then by modus tollens so is philosophical relativism. Why think the entailment holds? Well, it will partly depend on one’s philosophy of language, but compare the views of, say, an idealist who accepts a bundle theory of objects and conceptualism about properties with those of a direct realist. One would naturally think that a sentence like ‘‘there is a chair in front of me’’ expresses an empirical proposition. However, the idealist would consider such a proposition to be false and the realist would think it is true. If philosophical relativism is true, then we can’t adjudicate between idealism and realism, and so the empirical proposition ‘‘there is a chair in front of me’’ comes out both true and false. Hence empirical relativism. My response to this line of thought is to deny the ﬁnal conditional above. Both idealists and direct realists employ the same methodology to defend their views and to criticize each other, namely, the method of rational intuition and appeal to possible cases. Both parties adopt the perspective of rationalism. I have argued that relativism arises when there are different, irreconcilable methods of acquiring basic beliefs about philosophical propositions. Idealists and direct realists do not use different basic belief-acquiring methods; their disputes are internecine, like a debate between Catholics and Lutherans over transubstantiation. A common response to relativistic arguments based on intractable differences is that the calm and continued use of reason will, in the end, settle upon the truth and that present differences are therefore not intractable at all. Within a given perspective, this is a plausible view—a failure to

The Relativist Solution

125

ﬁnd a solution is no proof that no agreed-on solution is to be found. So realists and idealists may one day settle their differences. I have argued, however, that there is no way in principle for peyote-consuming shamans and rationalist philosophers to settle their differences. Their disagreement is far more fundamental: they disagree about how to acquire the noninferential beliefs upon which their systems are based. An unlimited amount of fair-minded reasoning will not produce convergence, since the disputing parties cannot agree about the data upon which reason is to operate, and there is no supraperspectival way to set the data. Perhaps relativism is ﬁner grained than I have argued, as Mackie and others think. It may be that even within a perspective like rationalism there are irreconcilable competing views and this could provide grounds for relativism. And one might thence try to derive empirical relativism as the objector does above. However, this is neither something my account of relativism is committed to nor something that it necessarily precludes. I am doubtful about the prospects for such a relativism, for the reasons indicated: given my account of perspectives as individuated by basic belief-acquiring procedures, the argument that continued application of those procedures will (or at least may) produce eventual intraperspective agreement seems to carry some weight. But in any event empirical relativism is not a logical consequence of the relativism defended here. Objection 3 Kai Nielsen has argued that relativism is inconsistent with the pursuit of wide reﬂective equilibrium. I have defended the virtue of wide reﬂective equilibrium as a component of not only the most promising kind of rationalist philosophy but also of the Christian revelation-based perspective and shamanistic hallucinogen-based doxastic practices. Therefore, one might think that either wide reﬂective equilibrium must be abandoned as a goal of these perspectives, or relativism must go. He writes: WRE is impossible if conceptual relativism is true. With conceptual relativism there could be various narrow conceptual-scheme-dependent reﬂective equilibria, mutually incommensurable; but with conceptual relativism there could be no WRE. This is so because there could be in such a circumstance no comparison across incommensurable conceptual schemes to determine which are the more adequate. Considered judgments, principles, theories, descriptions of the world where they belong to different conceptual schemes could not even in principle be compared, if conceptual relativism were true, and thus, because we cannot make such comparisons, we could not, even in principle, get our considered judgments into wide reﬂective equilibrium. (Nielsen 1993, pp. 328–329)

126

Chapter 3

First of all, Nielsen is stalking conceptual relativism, which is a different kind of animal than the relativism I am defending here. The present account of relativism relativizes the truth of propositions to perspectives, but not that of concepts or propositional content. Second, even absolutist defenders of wide reﬂective equilibrium do not think that in order to get our networks of beliefs (even restricted to philosophical beliefs) into reﬂective equilibrium that we must take seriously the results of each and every belief-producing method. For example, an absolutist proponent of wide reﬂective equilibrium might insist that in developing our philosophical views we accommodate the ﬁndings of our best scientiﬁc theories. So, in settling on views of the nature of the self and self-identity, we ought to integrate our beliefs with the latest ﬁndings of psychology and neurology. When Rawls writes that the development of an adequate moral theory requires that we take into account other ﬁelds of knowledge to attain wide reﬂective equilibrium, he has in mind economics and psychology. Wide reﬂective equilibrium does not demand that we accommodate the latest horoscope, or the wisdom of the leading Eleusinian heirophant, or the last papal bull in developing our philosophical theories. Some perspectives and belief-acquiring methods are seen as beyond the pale; they are incommensurable perspectives to one’s own. Which beliefs and belief-acquiring methods are viewed as the lunatic fringe and so properly ignored varies: what a shaman would dismiss out of hand is not the same as what an analytic philosopher would. This variability does not imply that neither can achieve wide reﬂective equilibrium in their beliefs. Thus Nielsen’s view of wide reﬂective equilibrium is uncritically and unmanageably wide. Objection 4 An absolutist might object that I have failed to show anything about the relativity of truth. To be sure, different belief-acquiring procedures might well lead to a diversity of beliefs about philosophical propositions. Furthermore, suppose we grant that the various methods of gaining beliefs are inevitably components of a broader doxastic system or perspective with integrated, systematically connected belief-acquiring methods and strategies for adjusting one’s beliefs to achieve reﬂective equilibrium; and suppose we even grant that the doxastic programs of different perspectives may irreconcilably lead in different directions and sanction beliefs that are inconsistent between perspectives as justiﬁed or warranted. None of this, claims the absolutist, will give the metaphysical conclusion that truth itself, even for the restricted class of philosophical

The Relativist Solution

127

propositions, is merely relative to perspective. All that can be derived is an epistemic conclusion that justiﬁcation (broadly construed) is perspective dependent. So the proposition that ‘‘there are immortal spiritual souls’’ is justiﬁably believed relative to Christian perspectives and is perhaps not justiﬁably believed relative to the perspective of pure rationalist philosophy. None of this is proof that there is no absolutely true, true in all perspectives, fact that there are souls (if there are) or there are not souls (if there aren’t). Likewise, if God exists, then it is true that he exists in every perspective. It’s not that ‘‘God exists’’ is true in some perspectives and not in others, even if ‘‘it is justiﬁable or warranted to believe that God exists’’ is true in some perspectives and not in others. This objection will not work. Suppose that some basic beliefacquiring methods justify the proposition that ‘‘there are immortal spiritual souls’’ and other basic belief-acquiring methods justify the proposition that ‘‘there are no immortal spiritual souls.’’ An absolutist will maintain that (1) nothing has been shown about which proposition is more justiﬁed, and that (2) nothing has been shown about which is true. However, as I have previously argued, in the case of fully ﬂeshed out doxastic perspectives like rationalism, Christian revelation, and the ritual use of hallucinogens, no perspective has a greater epistemic standing. I considered several arguments at the end of chapter 2 that attempt to vindicate the superiority of rationalism and concluded that they all fail. So in fact the absolutist’s contention that nothing has been shown about which proposition is more justiﬁed is false; ‘‘there are immortal spiritual souls’’ and ‘‘there are no immortal spiritual souls’’ are equally justiﬁed. That is, revelation justiﬁes ‘‘there are immortal spiritual souls’’ just as well as intuition justiﬁes ‘‘there are no immortal spiritual souls.’’ Let us consider the second point, that nothing has been demonstrated about truth. Previously I argued that we have good grounds for rejecting skepticism about philosophical knowledge. In other words, at least some philosophical propositions are knowable. The only way that they can be known is through belief-acquiring methods which themselves are integral to systematic perspectives. Without simply repeating the argument of a few pages back, a lemma of the argument for relativism was that there is at least one method of acquiring philosophical knowledge and that there is more than one method. Note that this claim is not that there is more than one method of gaining justiﬁed or warranted beliefs about philosophical propositions. There is more than one way of gaining knowledge, which of course adds to the conditions of belief and some appropriate relation that belief bears to truth, truth itself. If the present

128

Chapter 3

objection were correct, and it were merely the justiﬁcation or warrant of philosophical beliefs that is irreconcilably relative to perspectives, then there would be perspectives in which (say) ‘‘there are immortal spiritual souls’’ is justiﬁably believed and perspectives in which ‘‘there are no souls’’ is equally justiﬁably believed. Furthermore, the relativist contention would be that there is no hope of reconciliation or agreement between the two perspectives—they have incompatible mechanisms of gaining noninferential beliefs. If there were nevertheless an absolute truth about souls, one true in all perspectives, then it would be unknowable; skepticism would result. Why? Because there is no reason to think that the perspective of rationalist philosophy, with its method of rational intuition, is more likely to have a hold of the truth than the perspective of Roman Catholicism, with its method of revelation. By their own lights, internal to each perspective, the belief that ‘‘there are souls’’ is not justiﬁed, or is justiﬁed, respectively. There is no ‘‘metaperspective’’ from which we could select among ﬁrst-order perspectives. In other words, it is fundamentally arbitrary which perspective and attendant network of belief-acquiring methods one uses to gain philosophical beliefs. By hypothesis, there could be no rational preference for one over another; they generate different results, and yet the absolutist insists that philosophical truth is absolute and ﬁxed across perspectives. Thus by making justiﬁcation perspective-dependent, instead of truth, one arrives at skepticism about philosophical propositions. It is simply good fortune if we adopt a perspective that really gets us to the truth and ill fortune if we do not; there is no reason, from the fully reﬂective stance, to think the one is more capable of getting at the truth than the other. While one’s philosophical beliefs might be justiﬁed within a perspective, choice of perspective is ultimately capricious. Caprice does not add up to knowledge. In short, if (1) the justiﬁcation or warrant for philosophical propositions is perspective-dependent, and (2) there is no justiﬁable way to select one perspective over another, and (3) truth is absolute, then no one has knowledge of philosophical propositions. Since I have argued against skepticism, either (1), (2), or (3) must go. I have defended (1) and (2), and so it is (3) that must be rejected. Epistemic variability in forming noninferential beliefs leads to more than mere perspectivedependence for justiﬁcation. It makes truth relative to perspectives too. Objection 5 It might be objected that my defense of relativism for philosophical propositions is still open to a sort of self-refutation problem, similar to the

The Relativist Solution

129

one that plagued skepticism about philosophical propositions. That is, I have argued that philosophical claims are merely relatively true—true in some perspectives but not in all—while at the same time I have contended that the axioms and theorems of the logic of relativism (RL) are absolute truths and are true in all perspectives. If the axioms of RL are not absolute truths, then we can derive the familiar self-refutation of relativism. Yet, runs the objection, the axioms of RL are philosophical claims about how we should understand relativism. Relativism is a philosophical theory and its logic also consists of philosophical propositions. Therefore the axioms (and concomitantly the theorems) of RL must be merely relatively true, which is to say that they are not true in all perspectives but are merely true in some of them. This is a contradiction: the axioms of RL are both absolutely true and not absolutely true. Rejection of the ﬁrst conjunct lands us back with the original self-refutation problem, and rejection of the second undermines the central thesis of this book that philosophical propositions are merely relatively true. This is a powerful objection. There are two possible replies to it. The ﬁrst is that logical axioms are not philosophical propositions. I argued in the ﬁrst chapter that not all propositions accessed through a priori means are philosophical ones. Mathematical propositions are not a species of philosophical propositions, even though some mathematical beliefs are apparently acquired through a priori means. Peano’s axioms for arithmetic or Euclid’s axioms for geometry are not philosophical in nature, even if one wishes to argue that it is through rational intuition that one comes to accept those axioms. Thus the axioms for the logic of relativism are true in all perspectives, but they are not philosophical in nature and so not open to the preceding criticism. The second possible reply is to concede that the axioms of RL are absolutely true and that they are philosophical in nature. All that needs to be done to make the theory consistent is to give up the universal claim that all philosophical propositions are merely relatively true and instead argue that many are. Thus propositions about ethics, the nature of the mind, the existence and nature of the divine, and so on are deeply relative to belief-acquiring method and so are true in some perspectives but not in all. That ‘‘Jesus loves me, this I know’’ is true relative to Christian revelation, providing a series of basic beliefs about Jesus which are then systematized and codiﬁed and which believers then accept on the basis of reasoned inference from the credible testimony of those who experienced the revelation (i.e., because the Bible tells me so). The proposition that Jesus loves me is false relative to other perspectives and their belief-

130

Chapter 3

acquiring methods. But the formal logic that uniﬁes and explains this relativism has axioms and theorems that must be true in all perspectives. The result is still a substantial and provocative relativistic stance. I confess that I am not quite sure which of these alternatives to embrace. The account I gave in chapter 1 of the nature of philosophical propositions is largely ostensive, and a critic may argue that the exclusion of logical axioms is insufﬁciently motivated. On the other hand it does seem clear that the axioms of other formal systems like mathematics are not philosophical in nature and it is difﬁcult to say why logic should belong in the philosophy side of the house instead of the mathematics side. With respect to the second response to the critic, giving in on the absolutely true nature of some philosophical propositions (at least the axioms and theorems of RL) opens the door to traditional defenders of the verities for other classes of philosophical propositions. Perhaps this is not a bad thing—it requires argument to show that moral propositions are true in some perspectives but not in all, and it would also require argument to show that they are absolutely true. I am going to split the difference. I have argued that the basic belief-acquiring methods of rational intuition, religious revelation, and the ritual use of hallucinogens generate mutually inconsistent beliefs about the divine, morality, and the nature of the mind. These irreconcilable differences constitute a good reason to believe that propositions about those topics are true in some perspectives but not in all. If logical axioms are not philosophical propositions, then there is good reason to think that all philosophical propositions in general are merely relatively true. Of course, the self-refutation objection dies hard. One might argue that my account is open to self-refutation from a different direction. I have argued that there are no good reasons for privileging rational intuition over competing noninferential methods, and that it is these competing sources of basic beliefs that give rise to different epistemic perspectives and ultimately to relativism. Yet one might argue that the warrant for accepting my story itself is wholly and uniquely dependent on rational intuition. Revelation, shamanistic use of psychotropic drugs, and whatever else might be a basic method of gaining philosophical beliefs may not lead to the same thesis as does intuition. Suppose that revelation leads to the acceptance (in reﬂective equilibrium) of absolutism for philosophical propositions. Now this is a problem: intuition yields the result that the truth of philosophical propositions is relative to doxastic perspective, in one doxastic perspective the truth of philosophical propositions is absolute. Given the principle, discussed previously,

The Relativist Solution

131

that whatever is relatively absolute is absolute, it follows that philosophical propositions are absolutely true after all. Thus under the assumption of relativism for philosophical propositions, absolutism comes out true. Insofar as my argument for relativism is grounded in intuition, it winds up undercutting the universality of intuition and is ultimately self-refuting. While the concern raised is legitimate, is it not a successful criticism. The partisans of doxastic perspectives other than rationalism also face the trilemma the confronts the rationalist perspective. Defenders of revelation, for example, must recognize that their methods are producing a different set of philosophical beliefs in reﬂective equilibrium than the beliefs generated by rational intuition alone. They must then ﬁnd a way to argue for the superiority of revelation as a noninferential beliefacquiring method, or failing that, are driven to the conclusion that either there is no philosophical knowledge, there are no philosophical propositions, or relativism for philosophical propositions is true. In other words, supporters of revelation face all the same problems that rationalists do. The elucidation of these problems does not depend on the use of rational intuition to gain basic beliefs—the trilemma for philosophical knowledge is a logical consequence of the conﬂict among the the outputs of the varying doxastic perspectives and is not the result of intuitions about concepts, hypothetical cases, ‘‘what we would say,’’ or anything else. This means that supporters of revelation (and mutatis mutandis for the use of hallucinogens and any other noninferential method of gaining philosophical beliefs) will not be able to consistently defend absolutism in wide reﬂective equilibrium. The argument for relativism is not grounded in the use of intuition, a grounding that would be rejected by priests and shamans. The argument is based on the logical conﬂict among the outputs of the various means of gaining basic beliefs about the philosophical. Thus I am not using intuition to undermine intuition, as the hypothetical objector maintains. The relativist solution is one that everyone, no matter what their noninferential means of gaining philosophical beliefs, is driven to. Objection 6 I have argued from epistemological considerations to a metaphysical conclusion about the nature of truth. Given this approach, one might argue that the account I have given of philosophical knowledge—that it is irremediably linked with the belief-acquiring methods of particular perspectives—entails an epistemic theory of truth. The epistemic theory

132

Chapter 3

of truth is wrong, goes the objection, and so by modus tollens so is relativism. While the objection is simple to state, the connection between relativism and the epistemic theory of truth is somewhat complex. First we will have to consider what exactly is an epistemic theory of truth. We ﬁnd a variety of different proposals. Consider these contentions of C. S. Peirce: The opinion which is fated to be ultimately agreed to by all who investigate is what we mean by truth. . . . Truth is that concordance of an abstract statement with the ideal limit towards which endless investigation would tend to bring scientiﬁc belief, which concordance the abstract statement may possess by virtue of the confession of its inaccuracy and one-sidedness, and this confession is an essential ingredient of truth. . . . My social theory of reality . . . [is] that the real is the idea in which the community ultimately settles down. (Cited in Kirkham 1992, pp. 81, 84)

The correct interpretation of Peirce is a thorny matter, and his views on truth are well-trampled ground.17 Frederick Schmitt probably oversimpliﬁes matters when he interprets Peirce as defending one thesis about truth, but just so we have an epistemic theory of truth on the table to examine, let’s accept Schmitt’s interpretation. He proposes that Peirce’s theory is that true propositions are those that, given the proper method of inquiry, would not be subject to later revision as a result of further use of that method. The ‘‘proper’’ method of inquiry is the one that best stabilizes beliefs. Schmitt offers this analysis: Peircean epistemic theory of truth: P is true just in case belief that P would result from the proper method of inquiry applied until all beliefs that result from the method are stable. (Schmitt 1995, p. 113)

More recently, Putnam is best known for offering an epistemic theory in support of his internal realism (later ‘‘pragmatic realism’’ with its clear endorsement of the Peircean patrimony). He writes, ‘‘Truth,’’ in an internalist view, is some sort of (idealized) rational acceptability— some sort of ideal coherence of our beliefs with each other and with our experiences as those experiences are themselves represented in our belief system—and not correspondence with mind-independent or discourse-independent ‘‘states of affairs.’’ There is no God’s Eye point of view that we can know or usefully imagine; there are only the various points of view of actual persons. (Putnam 1981, pp. 49–50)

A few pages later (p. 55) he writes, ‘‘truth is an idealization of rational acceptability. We speak as if there were such things as epistemically ideal

The Relativist Solution

133

conditions, and we call a statement ‘true’ if it would be justiﬁed under such conditions.’’ Crispin Wright interprets Putnam’s theory this way: Putnamian epistemic theory of truth: p is true iff p would be justiﬁed under ideal epistemic circumstances.18 Other epistemic theories have been developed, for example, Wright’s own superassertability view (see Wright 1992, pp. 44–70), but the Peircean and Putnamian proposals are good representatives of the genre. Now, why would one think that the theory of relativism developed above entails either a Peircean or a Putnamian or some similar epistemic treatment of truth? Presumably the argument would go as follows. It is truth that is relative to perspectives on the present account, not epistemic properties like justiﬁcation or warrant. Justiﬁcation is wholly intraperspective; with respect to philosophical propositions, one gains new beliefs as a result of the methods indigenous to a speciﬁc perspective, and does not employ or take seriously basic methods from other perspectives. Bertrand Russell and Pope Paul VI both came to have moral beliefs, beliefs that were justiﬁed on the basis of methods legitimated by different perspectives, but they saw no need to take each other’s method seriously as a way to gain knowledge. Russell was hardly going to inquire as to the results of revelation in regulating and adjusting his moral beliefs, and Paul VI was not going to take rational intuition as the fundamental method of acquiring basic beliefs about moral propositions, even if he, like Russell, subsequently used reason to systematize and make coherent his moral views. Suppose that at least some of Russell’s moral beliefs were perfectly justiﬁed by the methods of rationalist philosophy. Suppose, for example, that Russell’s view that some premarital sex is morally permissible19 is justiﬁed within the perspective of a priori philosophy. Its permissibility is grounded in basic deontological or consequentialist intuitions about duty, obligation, personal liberty and harm to others, and subsequent reasoning that takes these intuitions as data. Moreover, imagine that Russell’s view is ideally justiﬁed and that there are no compelling counterexamples one might devise, no possible thought experiments that ultimately undermine the view. Furthermore, suppose that the same is true, mutatis mutandis, of the view of Paul VI. His belief that premarital sex is wrong is ideally justiﬁed within the perspective of Catholicism—no future revelations will undermine the belief, and no future reasoning based on prior revelation will defeat it either.

134

Chapter 3

On Peirce’s epistemic theory of truth, the proper method of inquiry is the one that best leads to reﬂective equilibrium, or stability of beliefs. I have argued that the achievement of reﬂective equilibrium is perspective dependent, and that the philosophical beliefs of a Tukano shaman or a Catholic Pontiff are likely to be as stable and revision-resistant as those of a rationalist philosopher. So a Peircean ‘‘proper method of inquiry’’ is no more than a belief-acquiring method nested in a wider doxastic system that is a perspective. If Russell’s view about the morality of premarital sex is true within the perspective of rationalist, intuition-driven philosophy and Paul VI’s view about the immorality of premarital sex is true within the Catholic perspective, then these things are so because those beliefs have resulted from the (perspective-speciﬁc) proper method of inquiry applied until all beliefs that result from the method are stable. Different methods are nevertheless proper Peircean ones because discrete methods can and do lead to stable beliefs. Is Putnam’s epistemic theory compatible with relativism? Putnam emphasizes that truth has two features: ﬁrst, it is independent of the present state of justiﬁcation in the here-and-now, but truth is not independent of all justiﬁcation. To claim that a proposition is true is to claim that it could be justiﬁed and in fact will be in the limit of inquiry when all the data are in and ideal epistemic circumstances (all the evidence one could want, ﬂawless reasoning, etc.) have been achieved. Second, truth is stable, or convergent, which is to say that as the justiﬁcation of a proposition approaches the ideal limit, the justiﬁcation of its negation decreases. Putnam thinks it would make no sense to consider a proposition true if both it and its negation were justiﬁed under ideal circumstances (Putnam 1981, p. 56). Given Putnam’s convergence criterion, one might think it is short work to show that relativism will not lead to a Putnamian epistemic theory of truth. Under the assumption that Russell’s support of the permissibility of premarital sex is ideally justiﬁed and also that Paul VI’s belief that premarital sex is always immoral is also ideally justiﬁed, presumably Putnam would deny that the proposition ‘‘some premarital sex is morally permissible’’ has no truth-value at all. Both it and its negation would be, ex hypothesi, ideally justiﬁed under ideal epistemic circumstances. However, someone pressing the objection that relativism entails an epistemic theory of truth would no doubt note that under the relativism presented here, ‘‘epistemic justiﬁcation,’’ at least for philosophical propositions, is speciﬁc to the belief-acquiring methods of perspectives. That is, the way in which a Jı´varo shaman would justify his beliefs about

The Relativist Solution

135

the philosophy of mind is not the same way that, say, Jerry Fodor would do so. However, since we can’t show that one method of acquiring basic beliefs is superior to the other, from the point of view of justifying beliefs about the mind, they are separate but equal. So it is within a perspective that convergence is satisﬁed. Intraperspective a proposition p, but not its negation, would turn out ideally justiﬁed in the hypothetical ideal epistemic circumstances. ‘‘Ideal epistemic circumstances’’ themselves are method-and-perspective relative. In another perspective, not-p would turn out ideally justiﬁed according to the methods and standards intrinsic to that perspective. Thus, in that perspective, not-p would be true. A Putnamian epistemic theory of truth ﬁts nicely with relativism after all. There are myriad problems with epistemic theories of truth, continues the hypothetical objector, and insofar as the present account of relativism leads to some version of an epistemic truth theory, it ought to be rejected. To evaluate this contention, I will ﬁrst examine a few of the prominent criticisms of epistemic theories of truth and consider whether the view that philosophical propositions are merely relatively true is prey to those objections. Ernest Sosa and Alvin Plantinga have argued that Putnam’s epistemic theory is subject to a sort of self-refutation.20 Sosa begins the argument by suggesting that the following proposition is true, or at least might be true: (P1)

No one is in ideal epistemic conditions.

Proposition P1 is equivalent to (P2)

(P1) is true.

This in turn amounts to (P3) Anyone who were to consider (P1) in ideal epistemic conditions would accept it. (P3) is manifestly absurd and yet is supposed to say no more than (P1), which is not absurd and is probably true. Therefore, conclude Sosa and Plantinga, Putnam’s epistemic theory is mistaken. This is an ingenious argument, but it may not prove quite what its proponents think. One could reject the Putnamian epistemic theory of truth and still maintain that anyone who is in ideal epistemic circumstances would believe any true proposition that he or she considers. If Jane is an ideal reasoner and has all the relevant facts about a proposition, she will believe it if the proposition is true and deny it if it is false.

136

Chapter 3

It doesn’t matter what theory of truth is the right one for this to be true of Jane. If (P1) is true, then Jane, the hypothetical ideal reasoner, will believe it. But that’s a problem, since under the assumption that Jane is in ideal epistemic conditions, (P1) is false. So it looks like (P1) must be false, and consequently someone is in ideal epistemic conditions (a conclusion sure to delight the theist Plantinga). Putting the argument slightly differently, anyone in ideal epistemic conditions will believe all true propositions p that they consider. Let ‘‘no one is in ideal epistemic conditions’’ be a substitution instance for p. Therefore anyone in ideal epistemic conditions will believe that no one is in ideal epistemic conditions. Since this contradicts the very idea of being in ideal epistemic conditions, it must be that ‘‘no one is in ideal epistemic conditions’’ can’t be a substitution instance for p, presumably because it is not an instance of a true proposition. Thus we arrive at the conclusion that someone is in ideal epistemic circumstances. While Sosa and Plantinga take the above argument to be the death blow for a Putnamian epistemic theory of truth, Putnam’s theory is nowhere present among the premises or the conclusion of the argument. In fact, neither (P1), nor (P2), nor (P3) mention or presuppose any theory of truth at all. In other words, no particular theory of truth is assumed for this puzzle to arise. It is a more general problem for the concept of being in ideal epistemic conditions, a seemingly coherent concept no matter what one’s epistemology or truth theory. There is no reason to think that the Sosa–Plantinga argument is especially telling against a Putnamian or Peircean epistemic theory of truth any more than it is against a correspondence, disquotational, or any other sort of truth theory. As far as the preceding argument is concerned, the epistemic theory of truth is no worse off than its kindred theories of truth. Furthermore, the Sosa–Plantinga argument may incorporate a fallacy. Either (P1) is a contingent truth, true in some possible worlds and not in others, or it is a necessary truth, true in all possible worlds. Let us assume the ﬁrst disjunct; (P1) is a contingent truth, one that it is true in the actual world but not in all worlds. (P3) is a counterfactual, one that requires the consideration of other possible worlds for its evaluation. In all possible worlds where someone is in ideal epistemic conditions, (P1) is false. Thus in a possible world in which Jane is in ideal epistemic conditions, (P1) is false and Jane surely would not believe it. Therefore no self-refutation arises. It is no help to say that if someone were in ideal epistemic conditions in that actual world and were to consider (P1) that they would then believe (P1), since in such a case (P1) would be false in

The Relativist Solution

137

the actual world. Suppose then that (P1) is a necessary truth, one that is true in all possible worlds. In that case, no one is in ideal epistemic conditions in every possible world. The evaluation of (P3) requires that we therefore consider what would be the case in an impossible world, one in which there is someone in ideal epistemic conditions. What would be the case is that anything goes—if an impossibility is true, then anything and everything follows. This is a pretty good reason to think that (P3) is false and not a harmless rewrite of (P2). A defender of an epistemic theory of truth may be committed to holding that (P3) is an analysis of (P2), but he is certainly not committed to holding that it is impossible that someone be in ideal epistemic circumstances. I conclude that the Sosa–Plantinga argument is not uniquely targeted toward a Putnamian epistemic theory of truth, and it is a dud anyway. Another objection to epistemic theories is that they are viciously circular. This argument has been worked out in detail by Richard Kirkham (1992) and echoed by Schmitt (1995). Kirkham argues as follows. To claim that a proposition is justiﬁed is to claim that it is justiﬁed as having some property. If a law is justiﬁed, it is justiﬁed as fair and useful; if a business expense is justiﬁed, it is justiﬁed as likely to eventually lead to increased proﬁts. There is no such thing as justiﬁcation simpliciter, without the following (at least implicit) prepositional phrase. Now, what makes a proposition epistemically justiﬁed? The traditional answer is that an epistemically justiﬁed proposition is justiﬁed as true. In other words, epistemic justiﬁcation is deﬁned in terms of truth. However, epistemic theories of truth hold that truth is deﬁned in terms of epistemic justiﬁcation. This is hopelessly circular. Furthermore, Kirkham is quick to reject any relativization to perspectives as a way out. He writes, Nor does it help any to equate ‘‘true’’ with ‘‘justiﬁed within a system’’ or ‘‘justiﬁed within a conceptual scheme.’’ For again I ask, Justiﬁed as what within the system? On this view ‘‘s is true’’ (or ‘‘s is true with a system’’) is circularly explained as meaning ‘‘s is justiﬁed as true within a system,’’ or else it is equated with something unintelligible like ‘‘s is justiﬁed as justiﬁed within a system within a system.’’ (Kirkham 1992, p. 51)

One solution to the circularity problem, one that Kirkham himself acknowledges, is to deny that ‘‘justiﬁed’’ is shorthand for ‘‘justiﬁed as v’’ for some value v. Instead the rules of justiﬁcation, whatever they are, are simply conventional ones, traditionally accepted within a particular doxastic practice. Kirkham rejects this approach because it fails to ‘‘explain why our rules of justiﬁcation are so successful, why, for example, our technology keeps getting better’’ (Kirkham 1992, p. 53). If

138

Chapter 3

noninferential belief-acquiring methods that yield prima facie justiﬁed beliefs are merely conventional, idiosyncratic to irreducibly pluralistic perspectives, what explains their success? Whatever the merits of Kirkham’s argument against a general epistemic theory of truth—that is, one that is a truth theory for all propositions—it does not provide a reason to reject an epistemic treatment of the truth-values of philosophical propositions. How would one measure the successfulness of the rules of justiﬁcation for propositions about ethics, or the divine? Certainly not in terms of our technology getting better. One might argue that accepting, say, the ritual use of hallucinogens as a legitimate method of gaining basic beliefs about the nature of the mind will have various pragmatic beneﬁts, and that the promotion of this pragmatic value is a good reason to ingest hallucinogens. Of course, this move would be to throw in the towel with respect to the epistemic theory of truth, because then justiﬁcation is deﬁned in terms of some value that it pursues, in this case pragmatic beneﬁts. That issue aside, how could one show that the use of rational intuition is a pragmatically superior way of gaining justiﬁed beliefs than eating the ﬂy agaric mushroom anyway? It was argued in the last chapter that there were no good arguments that demonstrated the superiority of intuition over its competition. To sum up, Kirkham has a point that in the case of empirical propositions, the justiﬁcatory procedure of the scientiﬁc method is strikingly successful. This success can’t be explained by a defender of an epistemic theory of truth who denies that justiﬁcation aims at any particular value. However, the same is not true of philosophical propositions. The justiﬁcatory procedure of rational intuition coupled with subsequent reasoning to achieve reﬂective equilibrium is not clearly a greater success than Christian revelation coupled with subsequent ratiocinating. So it seems that one could noncircularly subscribe to an epistemic theory of truth for philosophical propositions. Lynch has presented the following dilemma argument against Putnam’s epistemic theory of truth (Lynch 1998, pp. 108–109). How should we understand ‘‘ideal epistemic conditions’’? We can give either a weak or a strong interpretation of that phrase. Under the weak interpretation, ideal epistemic conditions are those in which a cognitive agent much like ourselves would have all the relevant evidence that agent could possess for or against a given proposition. The problem then is that there are many propositions that would never be justiﬁed or unjustiﬁed for such an agent, for example, the propositions that ‘‘it rained on this spot

The Relativist Solution

139

15,000 years ago to the day’’ or ‘‘the number of stars in the universe at this moment is odd.’’ It doesn’t seem that we could ever have evidence to support either the afﬁrmation or the denial of such propositions. Yet surely there is a truth of the matter about whether the present number of stars is odd or even. Unfortunately, under a Putnamian theory, if a proposition remains neither justiﬁed nor unjustiﬁed under ideal epistemic conditions, that proposition is neither true nor false. Consider the other half of Lynch’s dilemma. A strong interpretation of ‘‘ideal epistemic conditions’’ is that such conditions are those in which evidence for every proposition or its negation is available. This detaches the notion of ideal epistemic circumstances from cognitive agents with our evidence-gathering abilities. Ideal justiﬁcation is beyond our epistemic grasp. What’s worse is that the strong interpretation seems to sneak a nonepistemic kind of truth back into the account by declaring that the ideal epistemic situation is one in which all facts (i.e., all truths) are available. Lynch concludes that no matter how one interprets ‘‘ideal epistemic conditions,’’ the Putnamian theory of truth can’t be right; truth isn’t ideal justiﬁcation. I have two responses to Lynch’s dilemma. First, it does not seem that his argument is applicable to the Peircean version of epistemic truth. Recall that Peirce holds that a proposition is true just in case it is the result of the proper method of inquiry applied until all beliefs that result from that method are stable. So, for example, in the case of traditional rationalist philosophy, a philosophical proposition p is true iff p remains a member of the set of philosophical beliefs even after the relentless use of rational intuition and philosophical argument. There is no appeal to ‘‘ideal epistemic circumstances’’ in this account, and it is clearly wedded to the cognitive capabilities and investigative methods of agents like ourselves. My second response to Lynch’s argument is that even in the case of a Putnamian treatment of epistemic truth, it is not so clear that the ﬁrst horn of Lynch’s dilemma skewers philosophical propositions. Lynch presents propositions that (1) intuitively have determinate truth-values and (2) are such that we can’t acquire evidence for or against them. His examples are empirical propositions. But what about philosophical ones? What kind of claim about ethics, the mind, metaphysics, or the divine is such that (1) we have a powerful intuition that it has a truth-value (and it is not just the result of some more elaborate theory about the nature of propositions or of truth) and yet (2) we are convinced that no amount of philosophical inquiry can justify either the proposition or its negation?

140

Chapter 3

One might think that Colin McGinn’s mysterianism about the mental ﬁts the bill here. McGinn (1991) holds that there is some property of the brain that accounts naturalistically for consciousness, but we are unable to grasp what that property is; we are cognitively closed to it. So there is a truth about how technicolor consciousness arises out of soggy gray matter, but even unlimited philosophizing won’t justify our belief in that truth. The psychophysical nexus is ineffable for human beings. What exactly is the philosophical proposition that is true yet unjustiﬁable? Certainly not mysterianism itself, a view McGinn tries to justify with traditional philosophical argumentation and reasoning. Presumably the proposition is whatever is the truth about consciousness. That is the thing we are unable to justify. However, I am unconvinced that even McGinn’s mysterianism produces an adequate example of a philosophical proposition that is true yet unjustiﬁable. McGinn maintains that the connection between the brain and consciousness is a physical, naturalistic one. If that’s so, then the truth about consciousness isn’t properly a philosophical proposition at all, but a scientiﬁc one—one that we may not be able to understand, or ﬁnd persuasive reasons to believe, but empirical nonetheless. Thus we still do not have an example of a properly philosophical proposition that is determinately true and unjustiﬁable. I’m unsure how much weight to put on this second response to Lynch’s criticism of the Putnamian epistemic theory of truth. The key point to be made here is that neither the Sosa–Plantinga, Kirkham, nor the Lynch criticisms of the epistemic theory of truth conclusively show that an epistemic treatment of philosophical truth is mistaken. Thus even if the theory of relativism defended here entails an epistemic theory of truth, this is not necessarily a reason to be an absolutist about philosophical propositions. That said, I do not think that relativism has to lead to an epistemic theory anyway. Earlier I considered how both Peirce’s and Putnam’s epistemic theories of truth ﬁt with relativizing truth to perspectives, and in particular with the defense of philosophical propositions as merely relatively true (true in some perspectives but not in all). However, the fact that epistemic theories can accommodate relativism in no way proves that relativism entails an epistemic theory. One can accept other more general theories about the nature of truth, such as a correspondence theory, or a disquotational one, and still be a relativist. Here is a typical statement of a correspondence theory from Schmitt (1995, p. 145): a proposition is true just in case it corresponds to the

The Relativist Solution

141

facts of the world. More generally, a correspondence theory states that there is a relation between truth-value-bearers (whether they are propositions, sentences, beliefs, statements, etc.) and some segment of reality (facts, states of affairs, etc.). According to the present account, the truth of philosophical propositions is indexed to perspectives, with the same proposition being true in one perspective and false in another. Suppose we take propositions as the correct truth-value-bearers and assume a correspondence theory of truth. This would mean that whatever truths correspond to, facts, states of affairs, or whatever are themselves relatively the case. That is to say, under a correspondence treatment, those facts or states of affairs would be actual in some perspectives and not in others. The state of affairs of all premarital sex being immoral is actual in the Catholic perspective but not in that of rationalist philosophy. Corresponding to this state of affairs is the proposition ‘‘all premarital sex is immoral,’’ which is true for Paul VI (whose basic belief-forming methods for philosophical beliefs are intrinsic to the Catholic perspective) but not true for Russell (whose basic belief-forming methods for philosophical beliefs are intrinsic to the perspective of intuition-driven rationalist philosophy). One might wonder what it means to say that a state of affairs is actual in one perspective but not in another. We tend to think that reality is the same, no matter what one’s epistemic point of view. Of course, that is the relativist lesson—reality is not the same for everyone, or, more precisely, a certain small slice of reality that consists of philosophical states of affairs is not the same for everyone. Analogously, states of affairs are actual in some possible worlds and not in others. The state of affairs that I am currently typing this sentence at a particular date and time is actual in this world but not in all. In other possible worlds I lazily didn’t get around to writing this paragraph until tomorrow. Likewise philosophical states of affairs are actual in some perspectives and not in others. For a correspondence theorist, this means that the philosophical propositions that correspond to those states of affairs are true in some perspectives and false in others. Thus one can accept both a correspondence theory of truth and still be a relativist. Consequently, an epistemic theory of truth is not entailed by relativism; indeed relativism is compatible with multiple more general truth theories. If there are properly philosophical truths at all, and if we can have knowledge of them, then they are merely relatively true. Our philosophical knowledge is chained to the methods by which we come to have it,

142

Chapter 3

methods that in part deliver basic, noninferential beliefs. Those methods themselves are components of perspectives, abstract ways of knowing not necessarily compatible with each other. I have not argued that empirical truths are also merely relatively true; indeed I am inclined to think that they are absolutely true, true in all perspectives (although I have not argued for, and am not committed to, the view that empirical truths are absolute; I am just mentioning my own hunch). The way to promote absolutism for philosophical propositions, then, is to turn them into empirical ones, in short: naturalism. In the next chapter I turn to an assessment of the prospects for naturalizing all of philosophy. Formal Appendix We can deﬁne the two perspectival operators rigorously, and add them to ordinary predicate modal logic as follows.21 Language RL I (1) (2) (3) (4) (5) (6) (7) (8) (9) II

Primitive Symbols An inﬁnite set SEN of sentence letters fp; q; r; . . .g An inﬁnite set VAR of variables fx; y; z; . . .g An inﬁnite set CON of constants fa; b; c; . . .g An inﬁnite set PRD of predicate letters fF; G; H; . . .g Grouping signs ( ) and , The connectives &, s, 4, $, ! The quantiﬁers b and E The modal operators j and U The perspectival operators o and Y Syntax for RL

(1) If p A SEN then p A WFFRL (where the set WFFRL contains all and only well-formed formulas of RL) (2) If F n A PRD (where the superscript n denotes the adicity of F) and (a; b; . . . A VAR or a; b; . . . A CON) then Fa; b; . . . A WFFRL (3) If F; C A WFFRL then sF, ðF4CÞ, ðF&CÞ, ðF ! CÞ, and ðF $ CÞ A WFFRL (4) If F A WFFRL and x A VAR then bxF; ExF A WFFRL (5) If F A WFFRL then UF; jF A WFFRL (6) If F A WFFRL then YF;oF A WFFRL

The Relativist Solution

143

(7) Every element of WFFRL is constructed in a ﬁnite number of steps using (1) through (6). III Semantics for RL A model M for language RL consists of the following six things. (1) (2) (3)

A nonempty set W of possible worlds A nonempty set P of perspectives An accessibility relation Rp on W

The subscript ‘‘ p’’ represents a perspectival parameter on the accessibility relation. It effectively means that the model allows the accessibility of worlds to each other to vary from perspective to perspective. Thus Rp ww 0 means that w 0 is accessible from w given perspective p. This indexing makes the alethic modalities relative to perspectives and thus permits the claim that necessity is perspectival. This parameter might be abandoned while still maintaining a robust relativism. To do so would mean that the same worlds are accessible at a world no matter what perspective is chosen. (4) A commensurability relation C on P (5) A domain function Dp that assigns a domain Dp; w to each world wAW Here the perspective parameter relativizes what there is to perspectives. In our world, for example, from the perspective of shamanism there are noncorporeal souls. From the perspective of analytic rationalism there are none. (6) (i) (ii)

An interpretation function Ip which assigns an entity Ip ðaÞ to each a A CON a subset Ip; w ðF n Þ of Dp; w n for each F n A PRD

That is, given a perspective, a world, and an n-adic predicate F, the function Ip; w ðF n Þ speciﬁes the extension for F at that world given perspective p. The perspective parameter will ensure, for example, that at our world the extension of the predicate ‘‘is the moral cause of suffering’’ will be one thing in the Catholic perspective and something else in the rationalistic perspective. Also, in (6)(ii), Dp; w is raised to power n for the following reason. The extensions of n-adic predicates are ordered sets of n-tuples. The elements of these sets will be provided by the Cartesian products of Dp; w . The extension of a dyadic predicate with respect to

144

Chapter 3

world w and perspective p will be a subset of Dp; w Dp; w ; the extension of a triadic predicate a subset of Dp; w Dp; w Dp; w ; etc. Let M be a model, F A WFFRL , w A W, p A P, and g be a variable assignment. Also let ½tM; p; w; g ¼ Ip ðtÞ if t is a constant ¼ gðtÞ if t is a variable We can then provide a truth deﬁnition as follows: VM; p; w; g ðFÞ (that is, the truth value of sentence F in world w, given perspective p and model M, and any arbitrary variable assignment g) is: (i)

. . . tn Þ ¼ 1 iff ½t1 M; p; w; g A Dp; w ; . . . ; ½tn M; p; w; g A Dp; w and h½t1 M; p; w; g ; . . . ; ½tn M; p; w; g i A Ip; w ðFÞ ¼ 0 iff ½t1 M; p; w; g A Dp; w ; . . . ; ½tn M; p; w; g A Dp; w and h½t1 M; p; w; g ; . . . ; ½tn M; p; w; g i B Ip; w ðFÞ (ii) VM; p; w; g ðsFÞ ¼ 1 iff VM; p; w; g ðFÞ ¼ 0 ¼ 0 iff VM; p; w; g ðFÞ ¼ 1 (iii) VM; p; w; g ðF ! CÞ ¼ 0 iff VM; p; w; g ðFÞ ¼ 1 and VM; p; w; g ðCÞ ¼ 0 ¼ 1 iff VM; p; w; g ðFÞ ¼ 1 and VM; p; w; g ðCÞ ¼ 1 or VM; p; w; g ðFÞ ¼ 0 and VM; p; w; g ðCÞ ¼ 1 or VM; p; w; g ðFÞ ¼ 0 and VM; p; w; g ðCÞ ¼ 0 (iv) VM; p; w; g ðExFÞ ¼ 1 iff for every d A Dp; w : VM; p; w; g½x=d ðFÞ ¼ 1 ¼ 0 iff there is a d A Dp; w such that VM; p; w; g½x=d ðFÞ ¼ 0 M; p; w; g ðFt1

The notation g½x=d refers to the speciﬁc variable assignment of the value d to the variable x. VM; p; w; g ðjFÞ ¼ 1 iff for every w 0 A W VM; p; w 0 ; g ðFÞ ¼ 1 ¼ 0 iff there is a w 0 A W VM; p; w 0 ; g ðFÞ ¼ 0 (vi) VM; p; w; g ðoFÞ ¼ 1 iff for every p 0 A P VM; p 0 ; w; g ðFÞ ¼ 1 ¼ 0 iff there is a p 0 A P VM; p 0 ; w; g ðFÞ ¼ 0 (vii) VM; p; w; g ðYFÞ ¼ 1 iff there is a p 0 A P VM; p 0 ; w; g ðFÞ ¼ 1 ¼ 0 iff for every p 0 A P VM; p 0 ; w; g ðFÞ ¼ 0 (v)

such that Rp ww 0 : such that Rp ww 0 : such

that

Cpp 0 :

such

that

Cpp 0 :

such

that

Cpp 0 :

such

that

Cpp 0 :

The Relativist Solution

145

The clauses for the connectives &, 4, and $ follow from the above clauses along with the usual deﬁnitions for those connectives in terms of s and !. The clauses for U and b likewise follow from the above clauses plus their usual deﬁnitions in terms of j and s, and E and s, respectively. The clauses in the truth deﬁnition for both of the perspectival operators are given explicitly. It should be noted that, as in the case of b and E, and U and j, the perspectival operators are truth-functional duals. That is, oF , sYsF, and also YF , sosF. Thus, saying that a sentence F is absolutely true is equivalent to saying that it cannot be true in some perspective that notF. This accords well with our intuitions. If it is the case that F is absolutely true in perspective p, this means that it is true in all perspectives commensurable to p. Thus there can be no perspective commensurable to p in which not-F is true. Compare Nietzsche’s remarks in The Will to Power (§522) that ‘‘rational thought is interpretation according to a scheme that we cannot throw off.’’ This reﬂects his view that humans cannot live without interpreting their experience in a logical way. That is, logic is essential to human life, and so the laws of logic are absolutely true for humans. Now, in nonhuman perspectives, whatever they may be, things could be different. So if we specify the commensurability relation so that the range of C is all human perspectives, Nietzsche’s position is that every p 0 A P such that Cpp 0 : VM; p 0 ; w; g (the laws of logic) ¼ 1. Plainly, then, it cannot be relatively true (keeping the range of C ﬁxed, of course) that the laws of logic fail to hold. Thus it follows that oF , sYsF. It is also important to show how principle P, viz., that YoF ) oF, relates to language RL. It is this principle, recall, that turned out to be at the root of the self-refutation problem for ‘‘everything is relative.’’ P is a theorem of RL as long as C is symmetrical and transitive. Here’s the proof. Suppose that YoF is true. This means that VM; p; w; g ðYoFÞ ¼ 1. According to clause (vii) in the semantics for RL, VM; p; w; g ðYoFÞ ¼ 1 iff there is a p 0 A P such that Cpp 0 : VM; p 0 ; w; g ðoFÞ ¼ 1. That is, loosely speaking, YoF is true at a perspective p just in case there is some perspective p 0 that is commensurable with p and such that oF is true at p 0 . Given the truth of YoF at p, we know that oF is true at p 0 . In turn, the fact that oF is true at p 0 means that F is true at every perspective commensurable to p 0 . Suppose that p is commensurable to p 0 . We already know that Cpp 0 . Suppose now that Cp 0 p—that the commensurability relation is symmetrical. This establishes

146

Chapter 3

that, given the truth of YoF at a perspective p, F is also true at p. But the task at hand is to show the truth of oF at p. To get this, we need to further suppose that the commensurability relation is transitive. This can be expressed as oF ! ooF.22 In other words, given the truth of oF in perspective p 0 , we can get the truth of oF in all perspectives commensurable to p 0 . On our prior assumption of symmetry, oF turns out to be true in our original perspective p. Thus YoF ) oF in RL under the assumptions that C is symmetrical and transitive.

4

The Naturalists’ Challenge

So far I have presented the following picture of rationalist philosophy: rationalism is a way of knowing, a perspective whose primary beliefacquiring method is that of rational intuition. Rational intuition is a means of gaining basic, noninferentially justiﬁed (yet defeasible) beliefs about philosophical propositions. The Socratic tradition to which we are heir is that, like perception, there is a commonality, a uniformity in the deliverances of intuition. Intuition is supposed to be the uncovering and clariﬁcation of common sense. Moreover, just as science attempts to unify disparate perceptions into one explanatory system, so too rationalist philosophy attempts to construct theories that unite and explain shared philosophical intuitions. I have picked out philosophical propositions themselves ostensively, although I have argued that the familiar method of testing philosophical proposals and analyses through logically possible counterexamples can work only if the propositions in question are ostensible metaphysical necessities. The beliefs produced by intuition are subjected to the methods of analysis and counterexamples in an attempt to achieve wide reﬂective equilibrium. That is, philosophers strive for explanatory cohesion, formal consistency, and systematic connections among their philosophical beliefs and try to integrate them with justiﬁed beliefs in other domains. The problem of this approach is that there are competitor perspectives and methods of acquiring basic beliefs about philosophical propositions other than rational intuition. And some of these methods produce beliefs that are inconsistent with the deliverances of rational intuition, beliefs that seem equally capable of acceptance in reﬂective equilibrium. I have argued that it is untenable to infer that an inability to prefer the approach of one perspective to another leads to epistemic skepticism about philosophical propositions. Instead the true upshot is that we are forced into accepting relativism for philosophical claims: the truth-values

148

Chapter 4

of philosophical propositions vary across perspectives, which are abstract ways of knowing with speciﬁc belief-acquiring methods composing them. Long held in abeyance is the approach of naturalism. I suspect that naturalists, if they have read this far, have been frustrated at the failure to consider this alternative: there are no philosophical propositions! Why can’t we show that rational intuition is any better at getting at the truth than ritualistic hallucinogen consumption or Christian revelation? Because there is no standard for testing the results. When it comes to empirical propositions, there is a way to test, namely through experience and perception. As Socrates commented to Euthyphro, a disagreement about numbers is quickly settled by calculation, a difference of opinion over length is ended by measurement, dispute over weight is resolved by using a set of scales, and so on. ‘‘What sort of thing, then, is it about which we differ, till, unable to arrive at a decision, we might get angry and be enemies to one another? . . . See if it is not the following—right and wrong, the noble and the base, and good and bad’’ (Euthyphro, 7d). It is philosophical questions that are unanswerable, and for the same sorts of reasons that questions like ‘‘what if Hamlet had a cheery disposition?’’ or ‘‘who would win in a ﬁght between the Hulk and Spiderman?’’ or ‘‘would Funes the Memorious have a normal memory if he had Alzheimer’s?’’ have no deﬁnitive answer. They are all hypothetical questions about ﬁctions, with an unlimited number of elaborate stories that might be told in response. Such answers have no connection to reality, and there is no concrete way to settle on one as superior. Naturalists recommend that we turn from such entertaining yet ultimately pointless pursuits and recognize that only the scientiﬁc method will produce knowledge. Sometimes one is struck by the sense that the truly ancient problems in philosophy get more subtle and reﬁned over time but never really go away entirely. This seems true in the debate between the empiricists and the rationalists. I certainly have no pretensions of offering a deﬁnitive solution to this debate here. However, I do want to consider some of the primary ways that naturalists have criticized the evidentiary force of rational intuition, revelation, and the ritual use of hallucinogens. If none of these methods has any epistemic value, then they cannot be useful components in any kind of epistemic perspective. Naturalists would press that the conclusion to be drawn isn’t skepticism, but instead a rejection of all nonnaturalistic ways of knowing.

The Naturalists’ Challenge

149

There are three main ways that one might understand naturalism: as a primary methodological assumption that is presumed without argument; as a global thesis for which one provides arguments; and as a series of local arguments against this or that nonnaturalistic approach. Naturalism as a Primary Methodological Assumption This is the way in which Richard Lewontin presents the doctrine: Our willingness to accept scientiﬁc claims that are against common sense is the key to an understanding of the real struggle between science and the supernatural. We take the side of science in spite of the patent absurdity of some of its constructs, in spite of its failure to fulﬁll many of its extravagant promises of health and life, in spite of the tolerance of the scientiﬁc community for unsubstantiated just-so stories, because we have a prior commitment, a commitment to materialism. It is not that the methods and institutions of science somehow compel us to accept a material explanation of the phenomenal world, but, on the contrary, that we are forced by our a priori adherence to material causes to create an apparatus of investigation and a set of concepts that produce material explanations, no matter how counterintuitive, no matter how mystifying to the uninitiated. Moreover, that materialism is absolute, for we cannot allow a Divine Foot in the door. (Lewontin 1997, p. 31)

For Lewontin, naturalism is a premise, a stricture upon our epistemology, not something that is the result of argument. The preceding passage has made a great splash among creationists and other pro-religionists, who have been quick to latch on to Lewontin’s claims as an example of scientiﬁc hubris. Worse than hubris, his popular critics take his claims as proof positive that science is just as much a matter of nonrational faith as their own views, which, they infer, are thereby vindicated.1 This is a bit unfair. As Chisholm insightfully argues in ‘‘The Problem of the Criterion’’ (Chisholm 1982, ch. 5), two fundamental questions confront epistemology: 1. What do we know? What is the extent of our knowledge? 2. How are we to decide whether we know? What are the criteria of knowledge? In order to answer (1) we must know what knowledge is, and how to determine when we have some of it. In other words, we must have an answer to (2). How are we to go about answering (2)? We will have to consider some competing criteria for knowledge. How can we determine whether some proffered criteria for knowledge are good ones? The test

150

Chapter 4

procedure is presumably comparing what a set of criteria says with the data of what we know and what we don’t know. In other words, we must have an answer to (1). Now we have a circle: an answer to (1) depends on a prior answer to (2), and an answer to (2) depends on a prior answer to (1). Chisholm claims that there are only three possible solutions: Skepticism We can’t answer (1) without ﬁrst answering (2), and we can’t answer (2) without ﬁrst answering (1). Thus we can’t answer either question. Particularism We should beg the question by assuming an answer to (1) and then come to an answer to (2). Methodism We should beg the question by assuming an answer to (2) and then come to an answer to (1). In fact Chisholm is mistaken in holding skepticism to be a viable third alternative. Skepticism seems to collapse into either particularism or methodism. Here is an argument for the former disjunct. The skeptic asserts (A): we can’t answer (1) without ﬁrst answering (2), and we can’t answer (2) without ﬁrst answering (1); thus we can’t answer either question. Thus the skeptic is claiming that (A) is epistemically preferable to the other positions. If the skeptic claims to know that (A) is better, then the skeptic is offering (A) as an item of knowledge. If the skeptic is offering (A) as an item of knowledge, then the skeptic is a particularist. Therefore, the skeptic is a particularist. Similarly, we can show that the skeptic is a methodist if she defends (A) by appeal to a general principle like this one: you can’t know anything by begging the question. This sets up nonquestion-beggingness as a criterion for knowledge, and is thereby a form of methodism. Chisholm’s own solution to the problem of the criterion was to concede that we are forced to beg the question and assume an answer to one of the questions. We can be methodists (like Locke and Hume) and assume an answer to (2) or we can be particularists (like Moore and Reid) and assume an answer to (1). Chisholm was a particularist. Lewontin is a methodist. To be sure, Lewontin begs the question with his naturalistic methodism, although if Chisholm is right he has no other choice than petitio principii. Rationalist philosophers, defenders of revelation, and the like cannot share Lewontin’s assumptions about which is the correct method of acquiring knowledge, but neither are they compelled to mount a critique of his position. This does not mean that the perspectives of John Calvin or an Eleusinian hierophant are somehow

The Naturalists’ Challenge

151

vindicated by Lewontin’s undefended materialist assumptions (certainly Lewontin would have a ﬁt at the suggestion) although it does mean that begging the question against him is equally fair. Global Naturalism A Lewontinian assumption that all truths are physical truths and all acceptable explanations are materialist ones does not provide a reason to eschew nonnaturalistic perspectives and belief-acquiring methods, so long as one rejects the assumption. However, there are other ways in which naturalism can be understood. The second main way that one might treat naturalism is as a global research program. Considered as a research program, naturalism is a strategy, an approach that tries to provide only scientiﬁc solutions to philosophical problems. This is one way to understand Quine’s proposal to replace traditional intuition-driven epistemology with naturalized epistemology, that is, empirical psychology.2 Recall this famous passage from ‘‘Epistemology Naturalized’’: The stimulation of his sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving at his picture of the world. Why not just see how this construction really proceeds? Why not settle for psychology? Such a surrender of the epistemological burden to psychology is a move that was disallowed in earlier times as circular reasoning. If the epistemologist’s goal is the validation of the grounds of empirical science, he defeats his purpose by using psychology or other empirical science in the validation. However, such scruples against circularity have little point once we have stopped dreaming of deducing science from observations. (Quine 1969, pp. 75–76)

There have been a great many criticisms and defenses of Quine’s proposal since it was ﬁrst published; one, brieﬂy discussed earlier, is Kim’s (1988) complaint that epistemology is essentially normative, and a nonnormative psychology is no more of a replacement program for epistemology than is geology, tennis, or painting. It is simply another pursuit in which one might engage. I want to draw attention to different issues. The ﬁrst is that if Quine is understood as suggesting that traditional epistemological enterprises have failed and so we ought to try something else, such as turning to an empirical study of how human beings actually operate as cognizers, then he is not presenting a good reason to give up nonnaturalistic belief-acquiring methods altogether. Compare: one might argue that the internal combustion engine is inefﬁcient, oil supplies are ﬁnite and dwindling, and so we should turn our research efforts toward developing sustainable hydrogen fuel cells. A perfectly reasonable

152

Chapter 4

proposal, but not one that shows that all research into improving the internal combustion engine should stop. Perhaps granting agencies and private companies ought to spend some research and development money on fuel cell technology, but it is dangerously idealistic to wholly abandon traditional lines of research. Similarly, Quine’s naturalistic approach may be worth looking into, but we shouldn’t bet the farm on it. We certainly shouldn’t conclude that nonnaturalistic belief-acquiring methods yield no knowledge at all. Later naturalizers have attempted to draw the stronger conclusion that all nonnaturalistic methods must go. For example, Devitt has presented a defense of naturalized epistemology, one that is a corollary to his arguments on behalf of metaphysical realism. I have interpreted Quine’s naturalism as a research proposal—one that may have merit, but not something that shows all other approaches to acquiring philosophical knowledge to have no value. Devitt is doing something a bit different. He takes his arguments to show that we have excellent reasons for giving up nonnaturalistic methods of belief acquisition. Naturalism isn’t a research strategy, but the consequence of argument. Devitt is not attacking rational intuition in particular, nor are revelation or the use of hallucinogens his targets; but if he is right, then all these ways of gaining beliefs should be eschewed in favor of a strict scientiﬁc method. Thus the trilemma problem I raised in chapter 2 for the acquisition of philosophical knowledge never arises, and so the relativist solution is not needed to start with. This consequence is sure to please Devitt, who considers relativism to be prima facie absurd and grounded in a series of simple fallacies (Devitt 1997, ch. 13). In all fairness, Devitt opposes versions of relativism that entail empirical propositions to be true merely relative to some conceptual scheme, culture, and so on, and his dismissal of the view is partly motivated by the unrigorous and pompous claims some relativists are prone to make. Perhaps Devitt would be more sympathetic to a precise statement of relativism that is restricted to philosophical propositions. Devitt is concerned to defend realism against its critics. He deﬁnes realism as the view that tokens of most commonsense, and scientiﬁc, physical types objectively exist independently of the mental. Thus he wants to champion not only jaguars, but quarks as well. Much of Devitt’s discussions of realism center around his desire to separate the realism issue from other topics he considers to be red herrings, such as the correspondence theory of truth and semantic theories of reference. Setting those issues aside, Devitt claims that there are two chief argu-

The Naturalists’ Challenge

153

ments against realism, one stemming from Descartes and one arising out of Kant. These will be presented in turn. Cartesian Argument 1. If traditional skeptical arguments are right, then we do not know that tokens of most commonsense, and scientiﬁc, physical types objectively exist independently of the mental. (Premise) 2. If we do not know that tokens of most commonsense, and scientiﬁc, physical types objectively exist independently of the mental, then we do not know that realism is right. (Def. of realism) 3. Traditional skeptical arguments are right. (Premise) 4. Thus, we do not know that realism is right. (From 1–3) The ﬁrst premise is simply the traditional skeptical claim, which, despite the voluminous literature on the skeptical problem, is still fairly popular (contextualists and those who deny that knowledge is closed under known implication aside). Devitt assumes that it is true. It is the third premise that takes most of the heat from antiskeptics. Devitt offers some assessment of antiskeptical positions, but ﬁnally concludes that the familiar attempts to disprove premise (3) are all failures (Devitt 1997, secs. 5.2, 5.3). The next step to be taken from the Cartesian argument is that since we do not, and presumably cannot, know that realism is correct we have no reason to believe it. This is not the same as possessing a reason to deny its truth, but nevertheless it is a powerful tool for antirealists. Devitt’s response to the Cartesian argument is largely the same as the response he makes to the Kantian Argument, which is as follows. Kantian Argument 1. If we impose innate, a priori concepts on our experience (e.g., causality, time, and spatial relations) in order to make it knowable, then we do not know how things are in themselves. In particular, we do not know that tokens of most commonsense, and scientiﬁc, physical types objectively exist independently of the mental. (Premise) 2. If we do not know that tokens of most commonsense, and scientiﬁc, physical types objectively exist independently of the mental, then we do not know that realism is right. (Def. of realism) 3. We do impose innate, a priori concepts on our experience in order to make it knowable. (Premise) 4. Thus, we do not know that Realism is right. (From 1–3) As with the Cartesian argument, the lynchpin premise for the Kantian argument is the third one. Devitt thinks that both the Cartesian and

154

Chapter 4

the Kantian arguments are unmitigated philosophical disasters. They lead ‘‘either to a lack of any worthwhile knowledge or to knowledge at the expense of a truly bizarre metaphysics’’ (Devitt 1999, p. 96). He proposes that this terrible result should force us to reexamine our methodology and to try to diagnose exactly how we have gone wrong. The problem as Devitt sees it is that the support for the third premises in both arguments is a priori and intuitive. Rationalist philosophy leads to a dead end; therefore the solution is to give up rationalism and more generally turn away from all nonnaturalistic ways of knowing. Whereas I have interpreted Quine as offering naturalized epistemology as a research program, Devitt is arguing that ‘‘there is only one way of knowing, the empirical way that is the basis of science’’ (1999, p. 96). Once we accept that, we can safely ignore both the Cartesian and Kantian arguments as relying on a discredited methodology for their credibility. I am sympathetic to Devitt’s support of realism as he has deﬁned it. However, in defending realism the way that he has, he goes too far out on a limb. Devitt’s concerns are about our knowledge of the empirical world—if Cartesian skepticism is right, then we have no knowledge of tokens of extramental commonsense and scientiﬁc types. We can save such knowledge, he claims, only at the price of a bizarre Kantian metaphysics replete with ineffable things-in-themselves and innate mental categories. Notice that Devitt is wholly unconcerned with philosophical propositions. That is, it’s not necessary truths that trouble him; it’s not whether we can have knowledge of the nature of the mental, the divine, the good life, and so on that is at stake. Asking how empirical knowledge is possible given skeptical possibilities may lead to the rejection of realism, but Cartesian skepticism is not similarly efﬁcacious against philosophical knowledge. So even if Devitt is right that the rationalist approach is bad news for realism and scientiﬁc knowledge, there is no analogous reason to think that rationalism is the wrong methodology to address properly philosophical matters. Maybe we shouldn’t try to give an a priori justiﬁcation for the scientiﬁc method. But we shouldn’t try to give an a priori justiﬁcation, or intuition-based reasons, for accepting some proposed moral principle, or some theory of properties, or some treatment of art? That won’t follow from Devitt’s arguments. Devitt wants to know how we can save science and scientiﬁc knowledge from the clutches of rationalist philosophers and their wacky ideas. His answer is to dismiss the rationalist belief-acquiring method of rationalist intuition. In doing so, he follows his maxim of settling the realism issue before addressing epistemic or semantic issues. That is, Devitt

The Naturalists’ Challenge

155

presumes that the correct approach is to put metaphysics ﬁrst, and once we have in hand an empirical metaphysics, we can proceed to derive or infer an empirical epistemology. He writes, ‘‘naturalized epistemology takes science, and hence its posits pretty much for granted’’ (Devitt 1997, p. 76). In so assuming, Devitt comes awfully close to Lewontin’s position, or perhaps to G. E. Moore’s. Lewontin’s solution to the problem of the criterion was to be a methodist and assume that the way to gain knowledge is through the scientiﬁc method. Devitt seems to be similarly begging the question, except by assuming a Moorean particularism: the posits of science are largely correct, and once we assume that they are items of knowledge, we can go on to infer what the proper epistemic methods are for acquiring more of the same. However, if Devitt’s argument ultimately collapses into Lewontin’s, then it is hard to see why partisans of nonnaturalistic belief-acquiring methods and perspectives ought to ﬁnd his reasoning compelling. If naturalists and nonnaturalists are merely begging the question against each other in addressing the problem of the criterion, then neither side is making much headway. Insofar as Devitt’s arguments are effective, they are so with respect to empirical knowledge. As noted above, rationalism does not lead to Cartesian skepticism or Kantian metaphysics when its purview is restricted to philosophical necessities. Of course, naturalists aren’t going to be satisﬁed with a separate-but-equal doctrine, with the scientiﬁc method the sole route to knowledge of contingent truths and rational intuition the way to attain knowledge of philosophical propositions. The naturalists’ move at this point is to avail themselves of a Quinean resource: deny that there are logical or metaphysically necessary truths. At best nomological necessities—the laws of nature and their entailments— are in the ofﬁng. Since all propositions can in principle be overthrown by recalcitrant experience, no logically necessary truths are forthcoming. If the truth-values of all propositions are contingent, and it is only through the methods of science that we can acquire knowledge of contingencies, then philosophers have only one alternative: we must refashion our understanding of philosophical propositions so that they too are contingent matters of fact. Before pursuing this challenge, let’s make sure that the naturalist claim that ‘‘all propositions are logically contingent’’ is a sensible one. If not, then the Quinean strategy of denying that there are necessary truths can be safely set aside. ‘‘All propositions are logically contingent’’ means that for all propositions p, possibly p and possibly not-p. This proposition itself is of course logically contingent by its own lights. If we

156

Chapter 4

substitute ‘‘all propositions are logically contingent’’ for p, we derive ‘‘possibly all propositions are contingent, and possibly it is not the case that all propositions are contingent.’’ The second conjunct is equivalent to ‘‘possibly, there is a noncontingent proposition.’’ If we adopt the familiar S5 system of modal logic, then whatever is possibly necessary is necessary. It follows that there is at least one noncontingent proposition, and so there is a necessary truth after all. Therefore, it is false that all propositions are logically contingent. There are two possible naturalist replies to this self-refutation argument. The ﬁrst is to reject S5 in favor of a weaker modal logic like S4, thus blocking the inference. The second is to reject logical necessity entirely. Now, the sole difference between S5 and weaker modal logics is that in S5, accessibility is an equivalence relation. So a naturalist wishing to give up S5 would need to provide some empirical evidence that the accessibility relation is either not reﬂexive, not transitive, or not symmetrical. Perhaps the evidence simply is that there are good empirical reasons to accept that all propositions are logically contingent, and therefore the correct system of modal logic must contain an accessibility relation that is not an equivalence relation. There seems to be little interest in this approach, however. As Kneale and Kneale make clear in their epic The Development of Logic, debate over which modal logic is to be preferred has been based on logical intuitions and what seems ‘‘natural,’’ not any sort of empirical evidence or scientiﬁc reasons.3 The more common naturalist response, and the one Quine himself makes, is to reject alethic modality entirely. There is no reason to debate which system of modal logic is the correct one, when all are pointless. The reason to give up necessity and possibility is explanatory parsimony: a rejection of necessity has a simpler explanatory ﬁt in our web of belief than an acceptance of modality and its ﬂorid semantics. However, it turns out to be quite difﬁcult to accommodate such a normative, theoryguiding principle in a purely naturalistic fashion. Nonnaturalistic normativity has a way of sneaking back in when one tries to do so. To his credit, Quine acknowledges this fact, writing, ‘‘a normative domain within epistemology survives the conversion to naturalism, contrary to widespread belief. . . . The most general of its norms are perhaps conservatism, or the maxim of minimum mutilation, and simplicity, familiar in ontological contexts as Ockham’s Razor’’ (1995, p. 49). Yet Quine does not see, I believe, how damaging it is to the overall naturalist project that the scientiﬁc method presumes aesthetic principles that function as normative contraints on the development of scientiﬁc theories, and that these

The Naturalists’ Challenge

157

principles cannot themselves be given a wholly empirical, naturalistic justiﬁcation. In the ﬁrst place, if modal locutions are rejected on the grounds of parsimony, and parsimony is a nonscientiﬁc normative principle, then necessity and possibility are abhorred on normative grounds. And unless we can convert the normative to the scientiﬁc, we seem back in the familiar terrain of selecting modal logics on the basis of rational intuition, not as a result of empirical inquiry. Let us look more carefully into the problem normativity poses for naturalism—not the problem of how science can accommodate morality or the normative element of epistemic justiﬁcation, but the seemingly ineliminable normative component of science itself. Naturalists, philosophers of science, and scientists themselves are in widespread agreement that the pursuit and advance of science relies on nonscientiﬁc value judgments about beauty, symmetry, falsiﬁability, and parsimony. Consider the matter of parsimony. A great many scientiﬁc-minded philosophers have endorsed Ockham’s Razor. David Armstrong writes, ‘‘in general, the theory that explains the phenomena by means of the least number of entities and principles (in particular, by the least number of sorts of entities and principles) is to be preferred’’; Russell calls Ockham’s Razor ‘‘the maxim that inspires all scientiﬁc philosophizing’’; Geoffrey Hellman, Frank Thompson, and William Lycan equate it with no less than ‘‘sound scientiﬁc procedure’’ itself.4 Quine’s own tastes for ‘‘desert landscapes’’ is well known; he writes, ‘‘of itself multiplication of entities should be seen as undesirable, conformably with Occam’s Razor’’ (Quine 1976b, p. 264). Should be seen as undesirable. Precisely how do such unlikely items as shoulds and desires ﬁt into the construction of a scientiﬁc picture of the world? One would think that an account of normative language and human desires would turn up at the end of some scientiﬁc story, not as a priori constraints on the very way in which scientiﬁc theories get constructed. At least, one would think so on a full-blooded naturalism. Ockham’s Razor is not the only aesthetic principle promoted by Quine. In The Web of Belief he lists no fewer than ‘‘ﬁve virtues that a hypothesis may enjoy, in varying degrees’’ (Quine and Ullian 1970, p. 66). One of these virtues is parsimony, and the others are conservatism (the less a new hypothesis conﬂicts with prior beliefs, the better it is), modesty (the less outrageous a new hypothesis, the better it is), generality (the more widely applicable and less ad hoc a new hypothesis, the better it is), and refutability (new hypotheses should be falsiﬁable) (Quine and Ullian 1970, ch. 6). Yet it is difﬁcult to see how such normative principles,

158

Chapter 4

ones that are means of evaluating empirical hypotheses, can be presented as empirical hypotheses. A Quinean naturalist might hold out a candle for some kind of bootstrapping—assume various aesthetic criteria in the construction of scientiﬁc theories and then hope that those theories are subsequently able to subsume and explain our aesthetic choices. Normativity and aesthetics are the proverbial Wittgensteinian ladder, to be thrown away once we have scaled the scientiﬁc heights. Apart from the risk of circularity, such an assumption is a hope, a wish, not the result of a rigorous scientiﬁc proof or even a strong inductive inference. We saw an example of this phenomenon in chapter 1, with Kornblith’s hand-waving in the direction of causal theories of reference in an attempt to accommodate intuition. Quine’s own approach is to insist that his epistemic virtues are virtuous because of their ‘‘predictive efﬁcacy’’ (Quine and Ullian 1970, p. 135), although he provides no real empirical evidence for this claim. Can some story about ‘‘efﬁcacy’’ save the Quinean virtues? Let’s consider one of his principles in more detail, namely the principle of parsimony. I will argue that Ockham’s Razor is not only terribly hard to take seriously as a guiding principle, but is neither especially truth-conducive nor pragmatically helpful. In other words, the virtue of parsimony is purely an aesthetic one; scientiﬁc procedure itself seems to presuppose the normative. David Lewis gives perhaps the clearest accounting of parsimony: Distinguish two kinds of parsimony, however: qualitative and quantitative. A doctrine is qualitatively parsimonious if it keeps down the number of fundamentally different kinds of entity: if it posits sets alone rather than sets and unreduced numbers, or particles alone rather than particles and ﬁelds, or spirits alone rather than both bodies and spirits. A doctrine is quantitatively parsimonious if it keeps down the number of instances of the kinds it posits; if it posits 10 29 electrons rather than 10 37 , or spirits only for people rather than spirits for all animals. I subscribe to the general view that qualitative parsimony is good in a philosophical or empirical hypothesis, but I recognize no presumption whatever in favor of quantitative parsimony. (Lewis 1973, p. 87)

In short, Ockham’s Razor enjoins us to try to minimize the number of entities or types of entities in our ontology. Put another way, it tells us that when considering two theories, we are to choose the one, ceteris paribus, with fewer entities, or types of entities. The ﬁrst problem for Ockham’s Razor arises when we try to use it to choose among competing global ontologies. By ‘‘global’’ here, I mean a complete accounting of what there is. The problem is that given plausible and popular ontologi-

The Naturalists’ Challenge

159

cal commitments, it is impossible to use Ockham’s Razor to select one ontology over another. Let us ﬁrst consider how this arises for Lewis’s quantitative version of parsimony. The argument begins by recognizing there is an inﬁnity of things, for example, events. On Kim’s treatment of events, an event is the exempliﬁcation of a property at a time. For example, there is the event of your sitting as you read this sentence. Let us call this event Y1 . There is also the event of your sitting as you read this sentence. Let us call this event Y2 . If we assume that time is a continuum (that there are @1 moments in time), then there is an inﬁnity of events between Y1 and Y2 . Or consider the problem of not-being. Suppose I look in my kitchen larder and discover a dearth of bread. Shall we say then that there is one dearth of bread in the larder? My larder contains three shelves. I look on the top shelf, ﬁnd no bread there, and conclude that there is a dearth of bread on the top shelf. I examine the second shelf and ﬁnd the same thing. So too with the bottom shelf. Now it seems that there are really three dearths of bread in my larder—one on each shelf. Of course, I might reexamine the contents of the larder more precisely. I look ﬁrst at the right half of the top shelf, and then at its left half. Both contain a dearth of bread. By this method we come to the conclusion that the larder contains six dearths. Quickly we can see that, given the continuity of space, the larder contains an inﬁnite number of dearths of bread, each occupying an inﬁnitesimally small region. One might argue that Kim’s theory is mistaken, or that dearths are not real, but this is to miss the larger picture. Any decent global ontology will be committed to an inﬁnity of things. It is tempting to conclude straightaway that all global ontologies thus have the same cardinality, and hence that Ockham’s Razor is of no use selecting among them.5 However, Cantor taught us that there are different orders of inﬁnity. @0 is the cardinal number of the smallest inﬁnite set; this is how many natural numbers there are. The second smallest cardinal number is @1 , the cardinal number of the power set of a set containing @0 elements.6 This is the cardinal number of the real numbers. The cardinal number of the power set of a set containing @1 elements is @2 . This procedure can be iterated—it is through power sets that one ascends the ladder of inﬁnity. The key feature of this for our purposes is that @1 > @0 , @2 > @1 , and so on. Thus Ockham’s Razor is able to recommend, for example, that we prefer a global ontology with @0 entities instead of @1 entities. While it is difﬁcult to see the advantage of a theory committed to @0 entities instead of @1 entities, there is also another problem. We all

160

Chapter 4

know that on pain of Russell’s Paradox, there is no cardinal number of all the numbers; there is no set that contains them. There are, in Quine’s terminology, transcendently many numbers.7 If we are comparing two global ontologies, both of which accept realism about numbers, it is clearly it is impossible to use Ockham’s Razor to adjudicate between them. Both are committed to transcendently many things. We must be mathematical nominalists for Ockham’s Razor to have hope for a purchase. This is obviously a particular problem for Quine, who is both a realist about mathematical objects and a strong promoter of Ockham’s Razor. Now, those who accept the principle of mereological fusion or conjunctivism (as Quine does8) hold that any combination of things is a legitimate thing. This means, in essence, that if you have a set S of things, all of the members of the power set of S will be legitimate things. If you add the result above that there are at least @0 basic entities, fusers will conclude that there are at least @1 things. If we reapply the fusion principle to the set containing @1 elements, we get a set of things with @2 elements, and so on. In short, mereological fusion þ @0 basic entities ¼ a transcendent universe. If Horatio was a fuser, then, Hamlet was surely wrong to advise him that there are more things in heaven and earth than are dreamt of in his philosophy. There may be different things, but it is impossible for there to be more. To avoid transcendency and thereby keep Ockham’s Razor capable of preferring one inﬁnite set to another larger one, we (and Hamlet) must eschew both number realism and mereological fusion. The transcendence problem arises for Lewis’s qualitative parsimony as well. Lewis cheerfully quantiﬁes over kinds, and I will follow his lead. Cubes and spheres are both kinds of objects, and they are different kinds. There is a continuum of polyhedra between cubes and spheres; some more cubelike, some more spherelike. Each is a kind of object, even if only some have names like ‘‘dodecagon.’’ It would be pretty strange to argue that there isn’t an inﬁnity of kinds solely on the grounds that there isn’t an inﬁnity of words in natural languages. We can name all the polygons anyway, if we avail ourselves of the language of mathematics and assume that language is combinatorial. Now we have established that there is an inﬁnity of kinds of objects. Eliminating some kinds of objects—tetragons, say—still leaves us with an inﬁnity of kinds.9 If we add the fuser view that any combination of kinds of things is a kind, we arrive shortly at a transcendent universe. Regrettably for Lewis’s qualitative parsimony, he does exactly that. He maintains both that there is an

The Naturalists’ Challenge

161

inﬁnity of properties, and that any conjunction of properties is a property.10 Thus any ontology Lewis would ﬁnd palatable is committed to transcendently many kinds of things, and so it is impossible for him to use parsimony as a desideratum for choosing one ontology over another. One might protest that all the objects from cubes to spheres are all polyhedra, and are not fundamentally different kinds of objects. Thus an inﬁnity of them is not a problem for Lewis’s qualitative parsimony, which aims to ‘‘keep down the number of fundamentally different kinds of entity.’’ Yet this objection will not work. ‘‘Fundamental’’ is a relative term. Presumably, if two types of entities are at the same taxonomic level in our ontology, then they are equally fundamental. If a type appears at a higher (genus) level, then it is more fundamental. Thus cubes and spheres are equally fundamental, but we might consider that concrete entities are more fundamental than events. Likewise, horses and zebras are not fundamentally different kinds, since they are both species of equus. Yet one might think that sets and unreduced numbers are at the same ontological level as well, as they are both abstract mathematical objects. If so, then they are not fundamentally different kinds of entities. But Lewis explicitly prefers getting rid of numbers in favor of sets, and regards this as an appealing ontological diet. Yet locating entities in a taxonomy does not reduce the number of entities in our overall ontology. The fact that zebras and horses both fall under the genus equus does not ‘‘reduce’’ zebras to horses, nor does it decrease the number of quadruped land animals, or the number of kinds of them. While adopting a thoroughgoing nominalism about abstracta and rejecting the fusion principle will allow a defender of Ockham’s Razor to escape the transcendence problem, there is a related problem close at hand. This we might call the plain old inﬁnity problem. Imagine that we have an ontology O committed to an inﬁnity of basic objects such as events. Our mystical friend Shirley recommends that we adopt ontology O-NO, which is just like our ontology except that it posits a guardian angel for every event. We might appeal hopefully to Lewis’s quantitative parsimony as a way of rejecting O-NO in favor of O. Regrettably, it is no help at all. @0 events þ @0 angels ¼ @0 objects. Shirley’s O-NO has no more things in it than our ontology O. At this point Lewis’s qualitative parsimony looks promising— adding the angels seems to increase the number of kinds of things, and so on qualitative grounds O is preferable to O-NO. Unfortunately, this approach faces a dilemma. Either we are nominalists about kinds/types of things, holding that kinds are ultimately reducible to sets of individuals

162

Chapter 4

(or some such), or not. If we adopt nominalism, then O-NO adds a type of entity in a loose sense only. According to nominalism there are not really types, only individuals, and hence in a strict sense O-NO is really only positing more individuals. Thus we are back trying (vainly) to use the quantitative version of Ockham’s Razor to eliminate the angels. If we reject nominalism about types, then the inﬁnity of kinds of polygons issue makes a reappearance. Yet as soon as we have an inﬁnity of kinds, then qualitative parsimony is dead in the water. @0 kinds of polygons þ some kind of angels ¼ @0 kinds of things. Ockham’s Razor seems incapable of preferring ontology O to O-NO. How then are we to close our borders to the entia non grata? Perhaps we have erred in focusing solely on the numerical aspect of Ockham’s Razor. Ockham’s Razor also instructs us not to increase entities beyond necessity. If you need events or limits or numbers or virtual particles to understand the world, then by all means, help yourself. But don’t add superﬂuous items. As Quine puts it, ‘‘pad the universe with classes or other supplements if that will get you a simpler, smoother overall theory; otherwise don’t’’ (Quine 1976b, p. 264). Quine’s Razor is like a sign above a buffet: take all you want, but eat all you take. If this is all Quinean parsimony comes to, then it is trivial. It no longer urges us to be ontological misers (since this turned out to be impossible), but rather not to engage in conspicuous consumption. That is, Ockham’s Razor says to include in a theory only things that are essential to make it work; for example, don’t add guardian angels to event theory. This is trivial because it is a necessary truth about explanation that if A, B, and C explain Z, and A and B explain Z just as well, then C plays no role in the explanation of Z. Thus to say, as Ockham’s Razor does, ‘‘get rid of C on the grounds that it is not necessary to explain Z’’ is trivial, because it was never part of the explanation to begin with. It never helped explain anything, therefore it could not have been part of an explanation of anything; it was never part of the theory about Z. It would be a complete mistake to imagine that we once had a theory about Z that involved A, B, and C, and now we have a new, superior theory about Z that is C-free. This may be what Wittgenstein is getting at when he wrote in the Tractatus, ‘‘If a sign is useless, it is meaningless. That is the point of Occam’s maxim’’ (Wittgenstein 1961, §3.328; cf. §5.47321). Think of it this way. Imagine that we have long been eating lettuce because of its health beneﬁts. Now someone discovers that lettuce is fatfree, and touts the new and improved fat-free lettuce. Should we prefer the new lettuce to the old? It is fat-free! Obviously there is no basis for

The Naturalists’ Challenge

163

preference; it is the same old lettuce, which has always been fat-free. Even if we had once mistakenly believed that lettuce contained some fat, and so the new discovery is of epistemic interest to us, there is no physical difference between the new lettuce and the old. So too with theories. It may be of epistemic interest to us that the Z-theory doesn’t require C, but this only shows that C never played an explanatory role in the Ztheory. We do not have a new theory about Z, and Ockham’s Razor has not pared anything away. Even if Ockham’s Razor cuts no ice in the previous scenario, one might argue that Ockham’s Razor has a role to play in the following sort of case. Suppose that A and B explain Z pretty well, but adding C helps the explanation of Z to some degree. We are then faced with a decision: do we stick with just A and B as our postulated entities, or do we add C? There is a cost associated with adding further explanans, namely a decrease in explanatory parsimony. So the function of Ockham’s Razor is to point out the price to be paid, and to urge us not to adopt a new kind of entity unless it is worth it. Here I am tempted to wonder, along with John Haugeland, what makes ontology so expensive, or what limits our budget, or indeed, what exactly we are spending.11 It begs the question to assume that adding C to the Z-theory is in any real sense a cost at all. Such a claim assumes that Ockham’s Razor is a proper guide to theory-making or ontologizing. Yet this is the very issue at hand. So it seems that Ockham’s Razor cannot be retained as border guard against unwanted ontologies. Fortunately there are other ways of resisting the addition of undesirable entities. The appeal to parsimony is often a red herring, disguising the real approach to ontology-formation. For example, I suspect that Quine’s true strategy, and probably Russell’s as well, is get rid of the suspicious stuff. If intensions seem like ‘‘creatures of darkness,’’ then, well, try to design an ontology that excludes them. Ockham’s Razor may be appealed to as a way to do it, but the preceding arguments show that numbers are not really the issue. The number of posited entities does not provide an independent reason for preferring an ontology. One suspects that some purely aesthetic, normative criteria are sneaking back in. Even if global ontologies have entities in such numbers that Ockham’s Razor cannot choose among them, there still may be a use for the razor on a smaller scale. For example, if we are choosing between Ptolemaic and Copernican astronomy, Ockham’s Razor might be just the tool to eliminate those unwieldy epicycles. Why should one prefer the less cumbersome Copernican system? One reason is that doing so—using

164

Chapter 4

Ockham’s Razor—will help us to gain truth and avoid error. Parsimonious theories are likelier to be true than unparsimonious ones.12 This means that between two competing theories with a ﬁnite number of entities or types of entities, the one with the fewer number is likelier to be true. Here is an argument against this alethic conception of Ockham’s Razor. Either Ockham’s Razor is to be applied ceteris paribus, or it is not. If not, then it will select the ancient Greek list of elements, with its four types (earth, air, water, and ﬁre), over the modern periodic table, with its 110 elements. Modern chemistry is much less parsimonious than ancient chemistry. Yet it is the modern theory that is universally accepted, and, we may suppose, that one that is much likelier to be true. So as an unqualiﬁed principle, Ockham’s Razor is false. Suppose then that it is a ceteris paribus rule, not an inviolable axiom for theory preference. That is, if you have two competing theories, equal in all other empirical respects such as data retrodiction and prediction of novel phenomena, then you should prefer the one with fewer items or types of items. It is the one more probably true. While this version of Ockham’s Razor is not obviously false, it is difﬁcult to see what would count as evidence of its truth. Even if we are able to judge theory A as superior to theory B on parsimonious grounds, why should this lead us to conclude that A is more likely to be true? Remember, we are supposing that the two theories are equal in all respects save the number of entities they countenance. They both predict and retrodict the same data. Without simply assuming that parsimony grants the mantle of truth, it is hard to see why we should think B is likelier to be false than A. Alternatively, one might argue inductively as follows. Historically, we have seen competing explanatory theories that, at some time in the past, were thought to be equal in data retrodiction, prediction of novel phenomena, and so on. At some later time, one of the theories proved to be superior at (say) predicting novel phenomena, and so was adopted over its competitor. When this has happened, the winning theory has always (or usually) been the more parsimonious theory. Thus parsimony is an indicator of the future success of a theory. This argument could be correct. However, there is a fair amount of historical research that must be done to establish its premises. There is no a priori reason to accept any one of them, especially the claim that parsimony has attended success. Indeed, a cursory glance at the history of ideas suggests a progression from relatively sparse ontologies to

The Naturalists’ Challenge

165

more plentiful ones—the discovery of more types of particles, more biological species, and more galaxies seems the rule, not the exception. As Santayana puts it, ‘‘is nature sparing of atoms or seeds, or depths of organization and interrelation beyond the reach of human thought?’’ (Santayana 1942, p. 510). It could also be the case that Ockham’s Razor is a good indicator of future success in some domains and not others. For example, it could turn out to be contingently true that parsimony has been a good predictor in geology but not evolutionary theory. In any case, it would mean that Ockham’s Razor is not a universal virtue of hypotheses, as Quine claims. Even if Ockham’s Razor does not help us in the pursuit of truth, it may still have some pragmatic or prudential value. Consider our earlier example of two theories each with a ﬁnite number of postulated entities or types of entities. Suppose that theory A includes fewer entities than theory B. We have seen that Ockham’s Razor gives us no reason think that A is likelier to be true, but perhaps it gives us a reason to prefer A to B on the grounds that A is simpler, less cumbersome, and more practical to use. We should expect a theory that postulates two or three basic entities to be easier to understand and employ than a theory with a dozen. Thus Ockham’s Razor can prefer Copernican astronomy to Ptolemaic just on the grounds that it is just easier to work with. All the phenomena are saved with a more digestible, and hence preferable, theory. Again we must consider whether Ockham’s Razor, in its new pragmatic guise, is to be understood ceteris paribus or not. If not, then it will still select ancient Greek chemistry (with its four elements) over the modern periodic table (with its 110). This may be ﬁne in everyday life where an awareness of ﬁre is more important than a sensitivity to Hahnium, with its half-life of less than two seconds, but it is clearly a serious mistake if Ockham’s Razor is taken as a real guide to which theory we should accept as true. Rigid adherence to Ockham’s Razor would lead to a science that is wildly askew. Nature is not always easy, and pragmatically insisting that it be so is sure to stiﬂe our understanding.13 Let us suppose, then, that our pragmatic Ockham’s Razor is to be understood ceteris paribus. Sometimes a theory with fewer basic types of entities will be easier to use than a theory that does all the same work with more. However, this is not always the case, and as a universal generalization, even the pragmatic ceteris paribus reading of Ockham’s Razor is false. Recall Lewis’s parsimonious preference for doctrines that posit sets alone, a` la

166

Chapter 4

Russell and Whitehead, rather than both sets and unreduced numbers. How could this preference possibly be rooted in pragmatic considerations? That would be to suppose that there is a practical advantage to employing a theory the length and complexity of Principia Mathematica over the old, bad, doctrine of sets and unreduced numbers. Principia is a magniﬁcent achievement of considerable epistemic interest and even beauty, but surely it did not make mathematics easier or simpler in any prudential sense. The appeal of Principia is not a pragmatic one. In the end it is quite difﬁcult to see how a naturalist might defend parsimony as an epistemic virtue and do so in a nonnormative manner. Quine’s hope that it has a ‘‘predictive efﬁcacy’’ has not panned out. It is interesting that scientists, even Nobel laureates like Feynman and Weinberg, accept the important role of aesthetics and normativity in the development of science, without feeling particularly compelled to make apologies or try to give an empirical justiﬁcation for their aesthetic choices. Weinberg, for example, argues that it is the holistic unity of a theory along with its seeming inevitability and symmetry in its laws that help make one theory superior to another. He writes that when we ‘‘have a complete theory of the sort we want . . . it will not matter very much how many kinds of particle or force it describes, as long as it does so beautifully, as an inevitable consequence of simple principles’’ (Weinberg 1992, p. 149). Feynman concurs, writing ‘‘you can recognize truth by its beauty and simplicity’’ (Feynman 1965, p. 171). Indeed, beauty is of such importance that Weinberg concludes that ‘‘in any case, we [physicists] would not accept any theory as ﬁnal unless it were beautiful’’ (Weinberg 1992, p. 165). For Weinberg and Feynman it seems that ﬁrst philosophy has not been overthrown or replaced at all, but instead has a valuable and perhaps even an ineliminable role to play in the development of naturalistic theories of the world. Both physicists disparage metaphysics and epistemology as useless for science, but aesthetics and axiology, oddly, have much relevance. Thus although naturalism aims toward a scientiﬁc explanation of everything, it does not look like it can get along without a good bit of traditional philosophy. Local Naturalism Even if global naturalism is difﬁcult to sustain, this does not mean that no naturalistic arguments could be effective against the relativist treatment of philosophical propositions presented here. For example, a natu-

The Naturalists’ Challenge

167

ralist might present empirical arguments that purport to show that nonnaturalistic methods of acquiring philosophical beliefs are untrustworthy, or naturalists might try to explain philosophical claims and methods as secretly empirical. (The attempt to show that all philosophical claims are noncognitive met a bitter end when the logical positivists’ veriﬁability criterion of meaning committed suicide.) The remainder of this chapter will be an assessment of naturalist or naturalist-inspired arguments designed to establish one of these two things. I will argue that naturalists present arguments that are frequently a welcome corrective to the empirical naivete´ of armchair a priorists, but that their arguments do not demonstrate the stronger conclusions above. Of course, criticizing speciﬁc arguments will not amount to a general defeat of naturalism; at best it will send naturalists back to the drawing board to develop new and improved arguments. But that’s ﬁne. The advance of philosophy consists in the ongoing dialectic. Continualism One naturalistic tack is to co-opt rationalism and ﬁnd a way to convert rational intuition into a naturalistic method: rationalism is good as far as it goes, but that’s only because it really relies on scientiﬁcally examinable cognitive processes. This is the view of Goldman (1999; Goldman and Pust 1998), who thinks that cognitive science will show that a priori intuitions about philosophical concepts will reduce to empirical facts about cognition and concept-formation. Rational intuition is scientiﬁcally deconstructed and eliminated as a source of evidence; what seemed like a priori philosophical propositions are empirical and a posteriori after all. A ‘‘continualist’’ form of naturalism, a` la Goldman, which attempts to vindicate the deliverances of intuition empirically (conceptual analysis meets cognitive science) might produce an argument like this: (1) it is absolutely true that science is our best method for forming beliefs about the empirical world; (2) philosophy has developed in such a manner as to avoid conﬂicts with science, but revelation and mystic insight have not done so; therefore (3) we ought to have greater conﬁdence in philosophy’s deliverances than in those of revelation or mystic insight. Presumably the truth of (2) is due to the general reliability of the cluster of cognitive processes that undergird what we the scientiﬁcally naive folk call ‘‘rational intuition.’’ As Goldman and other naturalists claim, it is the task of science to uncover and evaluate these psychological mechanisms. Thus rationalism is both continuous with science and superior to

168

Chapter 4

revelation and hallucinogen use as a means of gaining beliefs about philosophical propositions. Unfortunately, both conjuncts of the second premise are extremely dubious. With respect to the ﬁrst conjunct, there has been a lot of wild and woolly philosophy based on pure reason—‘‘The World Spirit,’’ windowless monads, solipsistic idealism, the indivisible absolute mind, modal realism. If these things avoid conﬂict with science, it is only because scientists are too busy laughing at philosophers to care. It was really only with the British empiricists that any philosophers stopped claiming epistemological priority and started putting scientiﬁc discovery ﬁrst. In other words, it has been a rather recent development. Regarding the second conjunct, both Christianity and hallucinogenic traditions have been inﬂuenced by science. With respect to Christian revelation, mainstream Christianity (particularly Catholicism and the major Protestant denominations) have recognized the need to accommodate heliocentrism, natural selection, big bang cosmology, and nondemonic mental illness as real features of the world. Their theologies have developed in a manner so as to avoid (overt) conﬂicts with science. The usual approach, dressed in philosophical clothing, is outlined by Alston (1991, pp. 239–244). In Christendom, only fundamentalists and evangelicals are serious antiscience holdouts. Partisans of hallucinogens as basic belief-forming methods have also tried to square their methods with science—not so much by clearing doctrinal content past the scientiﬁc board, but primarily by connecting those traditions with evolutionary theory. Terence McKenna (1992) has argued that psilocybin mushrooms played an important role in human evolution, particularly through increasing sexual arousal and stimulating the linguistic areas of the brain, thus conferring an adaptive advantage on its consumers. He maintains that the development of human consciousness was coeval with primate consumption of hallucinogenic plants, and that symbolic language is intimately and organically connected with the metabolites of diet. In this way McKenna argues that the Ur-tree of knowledge is most likely a fungus. It is not obvious that McKenna’s admittedly speculative story about human cognition is less plausible than Alston’s religious acclimatization of science (to say nothing of Swinburne’s just-so stories about revelation) or the causal theory of reference, a supposedly empirical theory routinely defended on a priori grounds. Which approach has fewer intractable conﬂicts with science: rational intuition, revelation, or hallucinogens? I don’t know. But I do

The Naturalists’ Challenge

169

propose this: the answer is certainly neither evident nor a slam-dunk victory for analytic rationalism or for naturalism. Against Intuition Of the three nonnaturalistic belief-acquiring methods we have considered, namely rational intuition, Christian revelation, and the ritual use of hallucinogens, it is a safe bet that naturalists are least sympathetic to the last two. In chapter 2, I argued that, as far as the acquisition of philosophical knowledge is concerned, the methods of revelation and hallucinogens are no worse off than intuition. If naturalists are able to successfully argue that intuition has no epistemic value either, then they can chuck the whole messy lot overboard and focus on empirical approaches to philosophical questions. This is the approach that Jonathan M. Weinberg, Shaun Nichols, and Stephen Stich take with respect to the use of rational intuition in epistemology (Weinberg, Nichols, and Stich 2001). According to traditional epistemology, we discover normative principles about how we ought to form and revise our beliefs through a process of consulting our preanalytic intuitions about various epistemic cases and formulating (or refuting) principles on the basis of these intuitions. We commonly think that looking at a stopped clock never justiﬁes a belief that it is 2:00, no matter when we look at it. To make the normative element clear, we ought not form beliefs about the time on the basis of stopped clocks; we should ﬁnd a better method. Any account of epistemic justiﬁcation that entailed that one would have a justiﬁed belief about the time in the stopped clock case is thereby refuted. Our judgments about such cases and the resulting principles will then be adjusted to try to achieve reﬂective equilibrium. Weinberg, Nichols, and Stich argue that preanalytic intuitions are neither universally held nor somehow ‘‘pure’’ and uncontaminated by folk theory. The value of their work is that it is done empirically, utilizing the techniques of modern social science. At ﬁrst blush, this empirical approach might not seem like anything new; we saw many examples in chapter 1 of beliefs based on rational intuition being refuted by empirical discoveries. It is yesterday’s news that philosophers and their intuitions have been caught poaching on scientiﬁc territory. The innovation of Weinberg, Nichols, and Stich is that they are pushing for something quite a bit stronger. They want to show that science can debunk rational

170

Chapter 4

intuition. So not only is rational intuition fallible and open to possible refutation by empirical experience, but even more it is positively useless. Intuition has all the value for epistemology that psychic surgery has for medicine or that voodoo has for warfare. Like most naturalists, they consider rational intuition a magical, occult method that, once defenestrated, leaves naturalism as the only legitimate approach in epistemology. This is a bold program. Can it be made to work? I will argue no—even if science can defeat the intuitive justiﬁcation of speciﬁc propositions, it is powerless to debunk rational intuition as a general method. The strategy of Weinberg, Nichols, and Stich is to show that epistemic intuitions about familiar cases—Gettier’s examples, Lehrer’s Mr. Truetemp, Dretske’s zebras vs. cleverly disguised mules in a zoo— systematically vary across cultural background and socioeconomic status. Weinberg, Nichols, and Stich conducted surveys of Rutgers University undergraduates as to whether the subjects in Gettier, Truetemp, and Zebra cases ‘‘really know’’ or ‘‘only believe’’ the relevant proposition. The well-nigh universal view among professional philosophers is that one does not really know in a Gettier case, even though justiﬁed true belief is present. Weinberg, Nichols, and Stich show that a statistically signiﬁcant majority of students of East Asian descent do not share this intuition; they think that one really knows in a Gettier case. Over 70 percent of students of Western descent think that one only believes in a Gettier case. Weinberg, Nichols, and Stich found similar differences between the epistemic intuitions of those with an East Asian cultural background and those of a Western background in the Truetemp and Zebra cases as well. Weinberg, Nichols, and Stich expanded their study off-campus to random adults on the street, this time controlling for socioeconomic status (SES). Subjects were counted as low SES if they had never attended college, high SES otherwise. Again the researchers found that the epistemic intuitions of these groups were signiﬁcantly different from each other. To cite one instance, three times as many low SES subjects as high SES subjects polled thought that in the Dretske zebra case, one had knowledge that a perceived animal at the zoo was a zebra, even when the possibility of its being a cleverly disguised mule was raised. Weinberg, Nichols, and Stich consider couple of different explanations: perhaps low SES subjects just have lower standards for knowledge than high SES subjects, or perhaps the epistemic intuitions of high SES subjects are more sensitive to possible but nonactual states of affairs. Weinberg, Nichols, and Stich don’t endorse any particular theory.

The Naturalists’ Challenge

171

What Weinberg, Nichols, and Stich do conclude from the empirical studies is that the rationalist project of deriving normative epistemic principles from intuitions about hypothetical cases—the very bread and butter of the epistemological literature—is bankrupt. There are no neutral, preanalytic intuitions that everyone can consult to arrive at a common epistemology. Continuing to pursue the rationalist approach would lead to an absurd Balkanization of norms: ‘‘epistemic norms appropriate for the rich [would be] quite different from the epistemic norms appropriate for the poor, and . . . the epistemic norms for white people [would be] different from the norms appropriate for people of color’’ (p. 449). Whatever the merits of other forms of relativism, they argue, one that makes knowledge radically relative to ethnic background, class, and whatever else the social scientists discover is ridiculous. The historical hope of convergence and agreement on philosophical propositions through the relentless application of rational intuition is shown to be a fantasy. At this point the traditional epistemologist is well advised to give up the understanding of philosophical intuition as a Socratic drawing out of commonsense ideas universally possessed by the average citizen. Weinberg, Nichols, and Stich write: It may well be that upper middle class Westerners who have had a few years of graduate training in analytic philosophy do indeed all have strong, modalitylinked intuitions about Gettier cases. But since most of the world’s population apparently does not share these intuitions, it is hard to see why we should think that these intuitions tell us anything at all about the modal structure of reality, or about epistemic norms or indeed about anything else of philosophical interest. (2001, p. 452)

Fortunately, though, a rejection of Socratic commonsensism does not lead to this extreme conclusion. Rather, we should acknowledge that not all intuitions are created equal. Intuitions are and should be sensitive to education and training in the relevant domain. For example, the physical intuitions of professional scientists are much more trustworthy than those of undergraduates or random persons in a bus station. Scientists have and rely on physical intuitions, intuitions that are trained, educated, and informed and yet are good indicators of truth for those very reasons. In the same way, the modal intuitions of professional philosophers are much more reliable than either those of inexperienced students or the ‘‘folk.’’ In other words, philosophical intuitions are more like the physical intuitions of professional scientists than they are like Chomskian

172

Chapter 4

grammatical intuitions. It frankly doesn’t matter what the surveys of nonphilosophers show about whether one ‘‘really knows’’ or ‘‘only believes’’ in a Gettier case. These surveys shouldn’t lead us to dismiss the intuitions of philosophers any more than Richard Feynman would dismiss his intuitions about quantum electrodynamics on the basis of polling his Introduction to Physics students. Thus Weinberg, Nichols, and Stich aren’t really showing much of interest. At best they are showing that there is a naive conception of rational intuition that epistemologists should avoid. But this is far from proving their debunking hypothesis. Weinberg, Nichols, and Stich might reject the analogy I have drawn between the rational intuitions of professional philosophers and the empirical intuitions of professional scientists. They could argue that philosophical intuition is more like revelation than perception. Theologians’ experiences of the divine are not testable empirically and are seemingly irresolvable by argument, whereas the physical intuitions of scientists can in principle be proven wrong by experiment. There are plenty of examples of intuitively elegant theories that were junked in light of experience, but there is nothing to refute the intuitions of philosophers besides more appeals to intuition. It’s all just intuitions dancing on the head of a pin.14 There are several replies that can be made to this approach. First, the account of rational intuition presented here is one in which intuition delivers epistemically basic yet fallible data upon which reason operates to produce philosophical beliefs in wide reﬂective equilibrium. It is hardly true that philosophers accept nothing as refuting their intuitions. In fact, as Weinberg, Nichols, and Stich explicitly acknowledge, the justiﬁed-true-belief account of knowledge was widely accepted among philosophers until refuted by Gettier. Furthermore, it is also false that empirical discoveries can never disprove what is held on the basis of rational intuition. Exactly how they can do so was discussed at some length in chapter 1. There it was argued that rationalist philosophy proceeds by showing that a philosophical contention jp is false because Uq, and Uq ! Usp; therefore it is not the case that jp. The best way to show that something q is possible is to show that it is actual, and that is where scientiﬁc ﬁndings come in. I have already considered several examples. Thus the intuitions of the rationalists are open to disproof, even by empirical evidence. Second, if one rejects the contention that (like the empirical intuitions of professional scientists) the rational intuitions of professional

The Naturalists’ Challenge

173

philosophers are superior to the intuitions of the untrained masses on the grounds that rational intuitions are not testable empirically, then this bodes ill for mathematics and pure logic. The claims in those ﬁelds are not empirically testable either, but I rather doubt that Weinberg, Nichols, and Stich would liken mathematicians to theologians. Besides which, as in math and pure logic, philosophy has made genuine advances; lots of arguments thought persuasive by earlier generations have been shown to be inconsistent, dependent on shaky presuppositions, or otherwise ﬂawed. Slow advance is not no advance. The intuitions of philosophers have in many ways converged in the same ways that the physical intuitions of scientists have become more uniform. Third, there is a burden-of-proof problem here. Weinberg, Nichols, and Stich want to argue that 2,500 years of rationalist epistemology has been a waste of time. To effectively make this case they need to show that the trained intuitions of professional philosophers are no better than the untrained intuitions of random persons at a bus stop. A quick analogy with theologians is hardly an example of the hard scientiﬁc proof they think will kill off rationalist epistemology. In fact, it smacks strongly of an appeal to intuition. But in any case it seems that the burden of proof is on Weinberg, Nichols, and Stich to show that no one’s epistemic intuitions are any better than anyone else’s. Now, Weinberg, Nichols, and Stich might deny that this burden of proof is theirs on the grounds that there is no neutral way to say that some epistemic intuitions are ‘‘better’’ than others anyway. Suppose we unpack ‘‘better’’ in consequentialist terms: A’s intuitions are better than B’s if and only if normative principles of reasoning extracted from A’s intuitions (e.g., by a process of reﬂective equilibrium) lead to more true beliefs than normative principles extracted from B’s. If this is the line taken, then the traditional intuition-driven epistemologist needs to show that his or her (typically his) white, Western intuitions honed by years of philosophical training do in fact lead to more truth. Though the data on this are inconclusive, there is some reason to think that sophisticated thinkers in other (e.g., Asian) traditions would have signiﬁcantly different intuitions. If it is indeed the case that philosophers and other sophisticated thinkers in other traditions have different intuitions, surely it requires some heavy-duty argument to show that normative principles extracted from the intuitions of Western philosophers do a better job at leading to truth or some other desirable end. If the traditional epistemologist does not adopt a consequentialist interpretation of ‘‘better,’’ then he is duty-bound to explain what he does mean when he suggests

174

Chapter 4

his intuitions are better than those of Asians or less well-educated Westerners.15 My response is that it doesn’t matter how we unpack ‘‘better.’’ Pick a theory of justiﬁcation, any theory. The Weinberg, Nichols, and Stich argument is 1. The naive intuitions of the person on the street are different from the trained intuitions of professional philosophers. 2. The naive intuitions of the person on the street (along any epistemic dimension) are no worse than the trained intuitions of professional philosophers. 3. Therefore, professional philosophers are wasting their time by appeals to intuition. They provide credible empirical data in support of premise (1), and the logic of the argument seems persuasive. But no argument for the second premise has been offered. I have supposed that they might retort that it requires heavy-duty argument on the part of traditional epistemologists to show that normative principles extracted from the intuitions of Western philosophers do a better job at leading to truth or some other desirable end. While I agree that showing the instrumental value of intuition-based normative epistemic principles needs defense, I ﬁgure that since Weinberg, Nichols, and Stich are the ones trying to show that most of Western epistemology has been a big mistake, again they are ones who have the burden of proof here. Otherwise their argument is just a fallacious appeal to ignorance—something like, you traditional epistemologists can’t show that your methods are epistemically superior, therefore they aren’t. I have been arguing so far that the arguments given by Weinberg, Nichols, and Stich don’t in fact establish their radical conclusion that rational intuition has been debunked as a way to acquire knowledge. I also want to argue that their arguments cannot be adequately repaired. Note that Weinberg, Nichols, and Stich have freely helped themselves to epistemic norms, namely the norms of respectable social-scientiﬁc data collection and assessment, without bothering to justify these norms in the slightest. They spurn the epistemologist’s use of rational intuition to generate epistemic norms, insisting that it reﬂects no more than the prejudices of overeducated white Western males, while remaining oddly conﬁdent that their own epistemic norms are socially transcendent and absolute. Why? Is it that they think the normative principles of socialscientiﬁc reasoning will lead to more true beliefs than whatever epistemic

The Naturalists’ Challenge

175

principles are employed by Asians or those with low socioeconomic status? Let’s suppose that social-scientiﬁc reasoning really is objectively better at yielding true beliefs. This (undefended) supposition alone won’t show that Weinberg, Nichols, and Stich ought to reason in a socialscientiﬁc way unless they add the assumption that truth has intrinsic value. Unfortunately this assumption is problematic—one of the coauthors of ‘‘Normativity and Epistemic Intuitions’’ (2001), Stephen Stich, has argued that truth does not have intrinsic value.16 So at least he is faced with the problem of explaining exactly what valuable product tends to be produced by social-scientiﬁc reasoning. Now, Stich might argue that good social-scientiﬁc reasoning is likelier than competitor methods to lead to beliefs with high pragmatic utility, and it is this property, not truth, that ought to be promoted. Let’s suppose for the sake of argument that this can be persuasively demonstrated. It still remains to be shown that one ought to acquire beliefs with a high pragmatic utility. Why should one care about having beliefs that are pragmatically useful? One answer is a purely instrumental account of rationality. We just have various goals of inquiry, and, given these goals, social-scientiﬁc reasoning is the best way to achieve them. The goals themselves are beyond defense, since to argue that we ought to have some cognitive goals instead of others is to reintroduce normativity. The most that Stich could do is claim that there are desired practical consequences maximized by social-scientiﬁc reasoning, but he cannot claim that they are desirable ones. One ought to care about acquiring beliefs with high pragmatic utility because of the pragmatic beneﬁts, but that’s rather like answering the question ‘‘why be moral?’’ with ‘‘because it is the right thing to do.’’ The answer has the veneer of correctness but is fundamentally tautological. Another way to put the point is this: the supposed product of socialscientiﬁc reasoning is certain propositions, belief in which has pragmatic beneﬁts; and one ought to reason in that way because of the beneﬁts. If the ‘‘ought’’ is a pragmatic one, the claim is tautologous. The only way to make the claim nontrivial is if the ‘‘ought’’ is nonpragmatically normative. Thus, the normative element still remains. So it doesn’t look like Weinberg et al. are able to dispense entirely with normativity; their own method presupposes it. Perhaps instead of defending their own epistemic norms, Weinberg, Nichols, and Stich might head for the time-honored approach of their intellectual ancestors and deny that normative claims have a truth-value at all. Then they won’t need to show that it is true that social-scientiﬁc

176

Chapter 4

norms are likely to lead to useful beliefs, or that it is true that pragmatically useful beliefs have intrinsic value. If some sort of noncognitivism or quasirealism is right, then the proposition that it is a good thing to acquire beliefs with a high pragmatic utility is neither true nor false; it is simply an endorsement of utility, or something of the sort. One arrives at the sort of emotivist epistemology that Rorty has occasionally endorsed.17 Now, either Weinberg et al. will argue for normative noncognitivism, or they will presuppose it. If they presuppose the truth of noncognitivism in order to make their case that it is only through cross-cultural studies and surveys that we can come to an understanding of how people use normative language, then their argument begs the question. Such a procedure would assume the truth of a controversial theory of normativity (noncognitivism) to underwrite their social-scientiﬁc methodology, which would in turn yield a theory of normative language! For naturalists, a theory of normativity ought to be the result of empirical research, not a controversial philosophical assumption one begins with. Suppose, then, that Weinberg et al. provide an argument that noncognitivism is true. Their argument would presumably involve surveys, polls, data analysis, and the other trappings of social science (a purely rationalist argument of the sort one generally sees for noncognitivism would, for Weinberg et al., be self-defeating). In other words, they would employ the accepted norms of social-scientiﬁc research to argue for noncognitivism, just as they have used these norms in an attempt to debunk rational intuition. Again, we would have to ask what justiﬁes the use of those norms. Clearly in this instance the reply ‘‘we don’t need to justify them; norms are noncognitive’’ is not satisfying. Noncognitivism is the very thing under discussion. Even setting this methodological issue aside, naturalists should be wary of a simple ‘‘boo–hurray’’ approach to normativity. Boo-attitudes are famously idiosyncratic and relative to the personal psychology of the person who has them. If science relies on epistemic norms, and anything normative is a boo–hurray sort of affair and radically relative to personal judgments, then science itself is infected with a deep, come-as-you-are relativism. This is sure to make hard-nosed realists/naturalists shudder. Naturalism faces a host of problems involving normativity, as we saw earlier with respect to Quinean parsimony. The one to which I am now drawing attention is a branching dilemma. The ﬁrst dilemma is that the epistemic norms of social science either must be adopted without justiﬁcation or must themselves be justiﬁed. On the former horn, it seems that rationalists can simply accuse the naturalists of begging the question

The Naturalists’ Challenge

177

in favor of the scientiﬁc method and go on about their business. The latter horn—that the epistemic norms of social science must themselves be justiﬁed—branches into another dilemma. Either the justiﬁcation of social scientiﬁc norms will proceed via the scientiﬁc method or via some other method. If the former, then the validity of the scientiﬁc method is being presupposed in order to validate the scientiﬁc method. While this is bound to be successful, it’s about as persuasive as defending the doctrine of papal infallibility by asking the Pope. If one tries to justify social scientiﬁc norms by some nonnaturalistic method, this is for naturalists to throw in the towel and concede that there are methods other than the scientiﬁc method that produce knowledge. Even if, in the end, naturalists and rationalists are reduced to begging the question against each other, there is no reason to suppose that naturalists have somehow debunked the use of rational intuition. Weinberg, Nichols, and Stich might wonder why I am trying to so hard to rebut their position. What they do is use ethnographic data in an attempt to undermine the sanctity of rational intuition. Isn’t that what I am doing as well? After all, chapter 2 included an extended discussion of anthropological ﬁeld reports about the use of hallucinogens among traditional peoples to gain philosophical beliefs, and how there aren’t any good arguments to privilege rational intuition as a better method. So perhaps I should welcome Weinberg, Nichols, and Stich as fellow-travelers, and view their ﬁndings as complementary to my own. Unfortunately, the similarity between the use of ethnography in Weinberg, Nichols, and Stich and here is superﬁcial. They are attempting to debunk rational intuition as a legitimate method of gaining philosophical knowledge. They think intuition has no value and that this can be established empirically—systematically varying conﬂicts among basic epistemic intuitions show that intuition has no role to play in determining epistemic norms. The view I’ve defended here is that intuition does have value; it is just neither uniquely nor supremely valuable. The real issue isn’t conﬂicts among competing intuitions, but conﬂicts among competing perspectives with their concomitant basic belief-acquiring methods. It is at the interperspective level that relativism enters in, not within a perspective. Against Ethnographic Data The way in which familiar arguments for relativism proceed is through the consideration of what people in other societies and other cultures

178

Chapter 4

believe. Once differences between the beliefs of indigenous tribe X and those of twenty-ﬁrst-century Anglophone ﬁrst-world countries are noted, it is then typically argued that the differences in beliefs are due to incommensurable differences in concepts. One such example is the famous Sapir-Whorf hypothesis, which holds that dissimilar language speakers carve the conceptual world along different joints; in Whorf’s words, ‘‘we dissect nature along lines laid down by our native languages’’ (Whorf 1956, p. 213). Thus there is no language-neutral way to understand reality, and variances in language become different but equally valid worldviews. In support of this hypothesis, linguistic/ ethnographic data are cited, such as the Wintu use of tense, the varying nominalization of verbs, the Hopi classiﬁcation of events by duration type, and so on. The most notorious bit of evidence cited by proponents of the Sapir-Whorf view is the many words for snow purportedly used by the Eskimos. A legend embellished in the retelling, the Yupik and InuitInupiaq families of languages are supposed to have up to four hundred different words for snow, compared with the relative paucity in English. Whorf’s casual remarks in a paper published for a quasi-popular audience helped to canonize the view that this ‘‘fact’’ establishes that snow is of such great cultural importance for the Eskimos that they split up the conceptual world into distinct classes that English speakers are unable to recognize with their crude and limited tools.18 What kinds of snow are there? Well, the answer is relative to the diverse cultural and linguistic concepts of different speakers; there is no absolute, languageneutral fact of the matter. Both the claim about Eskimos’ words for snow and the argument into which it ﬁgures have been roundly and deﬁnitively criticized.19 In the ﬁrst place, the Eskimos do not have hundreds of words for snow. They have between two and around a dozen, depending on how the counting is done—in other words, about the same number as English. The inﬂated ﬁgures are due to careless scholarship that captured the public imagination, a sureﬁre recipe for starting an urban legend. Furthermore, even if the Eskimos really did have hundreds of words for snow, this wouldn’t show anything compelling about conceptual schemes or relativism, any more than do the mundane facts that English-speaking painters have many words for shades of beige (taupe, eggshell, tan, ecru, bone, cream, etc.) or printers have various names for different fonts (Geneva, Times, Helvetica, Caslon, Courier, etc.). Nonpainters and nonprinters may not recognize such ﬁne distinctions, but this is due to lack

The Naturalists’ Challenge

179

of interest, not lack of conceptual ability. They are not prisoners of language, but of education. I offer this thumbnail sketch of Sapir-Whorf conceptual relativism and one sort of criticism of the linguistic ethnography on which it is sometimes based for purposes of contrast. It is a fascination with the exotic Other that has made the claims about the Eskimo words for snow seem like ﬁne evidence for the Sapir-Whorf view. As Pullum puts it, ‘‘only the link to those legendary, promiscuous, blubber-gnawing hunters of the ice packs could permit something this trite to be presented to us for contemplation’’ (Pullum 1991, p. 166). While I have used ethnographic information to buttress my arguments for relativism about philosophical propositions, this is certainly not due to the charm of the curious. Revelation, for example, is a method of belief-acquisition quite familiar to educated Westerners. It is true that, despite its ancient pedigree, the ritual use of hallucinogens to gain beliefs about the mind, the divine, or the good life may appear bizarre, particularly to analytic philosophers strongly wedded to the virtues of sober rationalism. However, it is not the quality of strangeness that matters: I am not arguing, as does Whorf, that the Jı´varo or Huichol shamans possess alien concepts that parse reality in a way orthogonal to our own. The version of relativism that I defend has nothing to do with concepts—it is the method of gaining noninferential beliefs that matters. My argument is epistemic, not linguistic. The irrelevance of the Sapir-Whorf debate does not mean that every argument critical of relying on ethnography to support relativism is similarly irrelevant to the treatment of relativism proposed here. For example, Michele Moody-Adams has recently defended moral realism in the teeth of anthropology-based arguments in support of cultural relativism. She maintains that the most compelling arguments for philosophical moral relativism depend on certain empirical claims being true, claims that she labels descriptive cultural relativism (Moody-Adams 1997, p. 15). Descriptive cultural relativism is the thesis that differences in the diverse moral practices of diverse social groups generate fundamental, irresolvable moral disputes, ones that are neither reducible to nonmoral disagreement nor capable of rational resolution. MoodyAdams takes this to be an empirical contention, one based on the anthropological study of various cultures. ‘‘Culture’’ is of course difﬁcult to deﬁne, and she glosses it simply as any social group’s more or less distinctive way of life (ibid.). The importance of descriptive relativism, according to Moody-Adams, is that it underwrites all arguments for

180

Chapter 4

metaethical relativism. That is, unless it is true that the empirical study of different cultures reveals deep, irresolvable moral differences, then it cannot be the case that some moral judgments are true only relative to the practices or conventions of some society or cultural group. She argues both that descriptive cultural relativism is false, and that even if it were true, it would not establish relativism for metaethical claims. I will examine her arguments to see what force, if any, they have against the sort of relativism that I am defending for philosophical propositions in general (Moody-Adams is interested only in ethics). I count her among naturalist critics because her main arguments are centered around the role that anthropology and ethnographic ﬁeld studies play in the development and support of philosophical relativism. As a prolegomena to her critique of the evidence for descriptive cultural relativism, Moody-Adams defends Karl Duncker’s idea of situational meanings. In a 1939 Mind article, Duncker argued that any social practice has associated with it a situational meaning, which incorporates a complex set of nonmoral beliefs about not only the objective, causal properties of the practice, but its subjective affective associations as well (Moody-Adams 1997, p. 35). Situational meanings, according to Duncker and Moody-Adams, are essential to the individuation of actions. For example, consider the action of killing one’s aged parents. According to descriptive cultural relativism, there are cultures, that of the Eskimos, say, in which intentionally killing one’s elderly parents is morally permissible, whereas in other cultures this is regarded as horriﬁcally wrong. Duncker argues that the action description ‘‘intentionally killing one’s elderly parents’’ is incomplete, and does not pick out a speciﬁable action unless its situational meaning is ﬁlled in. As Moody-Adams writes, ‘‘on Duncker’s view killing one’s aged parents where that practice has the socially sanctioned meaning of sparing them a lingering death simply is not the same action as killing aged parents where this would be viewed as causing the early and unnecessary death of innocent persons’’ (1997, p. 35). Therefore, the very possibility of cross-cultural comparisons of the moral assessment of ‘‘the same act’’ requires a nuanced awareness of situational meanings. Duncker claims that close attention to situational meanings would show that no two societies have ever disagreed about the moral properties of a fully described act, and so descriptive cultural relativism is false. Moody-Adams is tempted by this view (she writes that ‘‘I deny that there is ever anything ‘new’ in moral thinking,’’ 1997, p. 62) but does not completely endorse it. She does, however, use the idea of situational

The Naturalists’ Challenge

181

meanings to highlight the complexity of actions embedded in a society and to argue that this complexity is a serious, perhaps insuperable, methodological obstacle to establishing descriptive cultural relativism. In addition to the argument from situational meanings, MoodyAdams stresses that moral intuitions are not culture-bound, but that even within a culture they can vary along lines of education and class. In this respect she is in agreement with Weinberg, Nichols, and Stich. Moody-Adams considers the example of cruelty to animals. What are the moral convictions of American culture on this topic? She argues that it isn’t sensible to say—there are innumerable various viewpoints on this issue represented in contemporary American society, and no one could be singled out as ‘‘American moral practice.’’ For example, most Americans are meat-eaters, although there is a substantial minority of vegetarians and quasivegetarians. Many Americans are unconcerned with slaughterhouse or factory farm conditions and yet treat their pets as children; some animals as viewed as noble, beautiful, and wrong to kill for those reasons (e.g., raptors), whereas others are perceived as dirty ‘‘nuisance’’ animals whose killing is expected (e.g., rats); and so forth. Pinning down an easily formulated ‘‘American moral attitudes toward nonhuman animals,’’ she argues, is well-nigh impossible, making unworkable the cross-cultural comparisons needed to establish descriptive moral relativism. A related issue is the matter of who should count as a representative spokesperson for the practices and beliefs of a culture. This is a methodological issue quite relevant in anthropology; ethnographical ﬁeldwork often relies on a few tribal informants who relate practices and attitudes that anthropologists proceed to ascribe to the tribe as a whole. Moody-Adams holds that this procedure is absurdly simplistic. Consider again the treatment of animals in the United States: whom should a naive anthropologist interview to ascertain American moral attitudes? Someone who would write letters to the New York Times? A Montana woodsman? George W. Bush? A professional philosopher (and then who? Peter Singer?)? Furthermore, even when there may be a majority view on some moral topic, the presence of a distinct minority voice may have great social signiﬁcance in that culture, a fact that can easily be missed or ignored when one relies on the reporting of some selected authority. Moody-Adams gives several examples from classic anthropological studies, which I will not recount here. To summarize Moody-Adams’s arguments, she holds that it is impossible to ever provide adequate evidence that descriptive cultural

182

Chapter 4

relativism is true. To fully assess the moral attitudes of persons in other cultures, we would have to overcome the obstacles of determining who shall count as the cultural spokesperson, how to determine the extent and role of subcultural moral dissidents, and how to establish the situational meaning of the various acts in that society. In short, the irreducible complexity of the various, integrated, moral practices within a given culture ‘‘consistently thwarts relativist efforts to conﬁne moral judgments to a single culture’’ (Moody-Adams 1997, p. 61). How might this be a threat to the sort of relativism I have defended? The arguments that I have presented for relativism about philosophical propositions have depended on the existence of diverse epistemic perspectives that incorporate different basic belief-acquiring methods. These perspectives are such that they can yield sets of beliefs about philosophical propositions such that each set of beliefs is capable of being accepted in reﬂective equilibrium and yet the belief-sets are mutually inconsistent with each other. If Moody-Adams is right and we can say so little about the moral beliefs of other cultures, then we will not be able assess whether their beliefs are in fact inconsistent with those produced by rational intuition. Indeed, we may not even be able to recognize the notion of a doxastic practice or an epistemic perspective at all, given the hurdles of ascribing beliefs to groups of individuals instead of to individuals themselves. Thus Moody-Adams presumably would reject a label such as ‘‘traditional Christians’’ as picking out any well-deﬁned group that uses revelation as a way to gain noninferential beliefs about the divine. Moody-Adams is persuasive that philosophers and anthropologists have often been hasty in ascribing broad, oversimpliﬁed, views to large classes of people. However, she errs on the opposite end in setting the bar so high that no generalization, no matter how ﬁne-grained, will count as true. Generalizations tend to simplify by their very nature, just as maps do. No map includes every bend in the river down to the nanometer, but it is an unfair criticism to insist that therefore no map is accurate. Furthermore, Moody-Adams seems caught on the horns of the following dilemma. Grant that because of situational meanings, dissident subcultures, and the problem of selecting a spokesperson we can’t say what a whole culture believes. Surely a more sensitive ethnography could examine the subcultures in a society, interview their leaders, and come to an understanding of what actions mean in different contexts with variable background assumptions. This would lead to a more complete and

The Naturalists’ Challenge

183

accurate assessment of moral (and other philosophical) attitudes within a society. No doubt it would still be the case that Moody-Adams is correct that descriptive cultural relativism is not in the ofﬁng for state-level cultures, but descriptive cultural relativism could be defended at the level of subcultures. There are diverse subgroups within a single society whose diverse moral practices generate fundamental, irresolvable moral disputes, ones that are neither reducible to nonmoral disagreement nor capable of rational resolution. Thus she has not defeated descriptive cultural relativism. Moody-Adams might rejoin that the same problems she raises for determining the moral attitudes of large-scale cultures reappear at the level of subcultures as well, and so we will never be able to ascribe particular moral practices to subcultures either. Again descriptive moral relativism is thwarted. This approach confronts the other horn of the dilemma: if we can’t generalize at all about what people in a subculture believe, then we can’t even individuate cultures. All we could talk about is what individual persons believe, or perhaps what individuals believe at a certain time, given certain prompts, in particular contexts, and so on. Moody-Adams herself deﬁnes ‘‘culture’’ in terms of a distinctive way of life. However, to reject the possibility of a reasonable generalization about the normative attitudes of a group of people is to reject the possibility of recognizing that group as a possessing a distinctive way of life and so as constituting a cultural unit. The conclusion is that either we can still get descriptive cultural relativism (just at the level of subcultures) or Moody-Adams is forced to conclude that there are no cultures at all, just hordes of individuals with varying beliefs and practices. I have focused on Moody-Adams’s critique of descriptive cultural relativism. It is worth mentioning that she also presents a more theoretical argument against relativism. Suppose that descriptive cultural relativism is true, whether at a cultural or subcultural level. Moody-Adams still doesn’t think that it will establish philosophical relativism. She maintains that there is always the possibility of the rational resolution of differences—currently unsettled disagreements do not prove the existence of rationally irresolvable disagreements. However, on my account of relativism, rational resolution is precluded, precisely because relativism stems from different perspectives, diverse doxastic practices with distinct methods of acquiring noninferential beliefs about philosophical propositions. When one group accepts revelation and another rejects it, or one accepts hallucinogens and another rejects it, the beliefs that are generated by those methods are not

184

Chapter 4

available to those who reject the method. More accurately, distinct doxastic practices lead to different (albeit overlapping) sets of beliefs. In chapter 2 the conﬂict thesis was defended at length, namely, the contention that the method of rational intuition leads to a belief that p (for many propositions p) and the competitor methods of the ritual use of hallucinogens and revelation lead to a belief that not-p. The problem isn’t, as Moody-Adams hopes, that there are unsettled philosophical disagreements awaiting rational resolution which are irresolvable. Agreement is possible only if the disputing parties ﬁrst agree on a common method for acquiring basic beliefs. But this is precisely what is not agreed on. Even if everyone settled on a common inference procedure and concurred about the value of reﬂective equilibrium, without agreement about what counts as legitimate data, there will still be irreconcilable differences at the end. Conclusion Nietzsche once wrote that all philosophers come to love their conclusions. I wish that he were right. The chief conclusions of the present book— 1. there is no reason to think that rational intuition is a better means of discovering which philosophical propositions are true than Christian revelation or even the ritual use of hallucinogens; and 2. the truth values of philosophical propositions are relative to doxastic perspectives, and may be true in one perspective and false in another; —are deeply disturbing. I ﬁnd it psychologically impossible to take the sensus divinitatis seriously, as an authentic means of coming to have knowledge of what must be true. Nor can I entertain taking hallucinogens as a way that I personally might come to understand the world. These methods seem primitive and absurd to me, prescientiﬁc, barbarian methodologies that anachronistically survive in the modern world. Even when I reﬂect on the fact that my feelings about these alternative methods are due to my own doxastic perspective of rationalist analytic philosophy, they are not live possibilities for me; they remain the incommensurable Other. I can sympathize with Hume, who although convinced by his own skeptical arguments, after a few drinks and backgammon with his friends, found his speculations so ‘‘cold and strain’d’’ that he hadn’t the heart to enter into them again.20 And yet I am persuaded by the arguments I have given that the truth about what my

The Naturalists’ Challenge

185

moral obligations are, the existence and nature of God, the nature of art, the analysis of knowledge, and the rest are all relative truths. It is a bitter pill, but one that intellectual honesty demands we swallow. At least relativism has this to be said for it: unlike Humean skepticism, it has a built-in error theory. A philosophical ediﬁce built on rational intuition appears so comprehensive, systematic, and closed that it is hard to imagine how a comparable structure could be erected on different foundations. It is high time that philosophers come to recognize that their own familiar foundations have no privilege and their ways of knowing have no monopoly on the truth. Indeed, the reason for this is that there is no single truth to be had, but many perspective-dependent truths—relativism arises from the very foundations of philosophy.

Notes

Introduction 1.

Compare Rorty 1986 with Sosa 1987.

2. Nagel (1997) develops this strain of thought to oppose subjectivism in various domains. 3. Examples of this general approach that differ from each other include Whorf 1956, Unger 1984, and Lynch 1998. 4. John MacFarlane is working along these lines in his unpublished work. Also, Andy Egan, John Hawthorne, and Brian Weatherson have a forthcoming article ‘‘Epistemic Modals in Context’’ in which they till this soil.

Chapter 1 1.

Portions of this chapter were ﬁrst published in Hales 2000.

2.

Bonjour 1998, pp. 98, 106.

3. Exactly how theory, intuition, and experiment interrelate is a complex matter that I will not enter into here. Steven Weinberg has a good discussion of this topic in Weinberg 1992. In Hales 2001a,c, I argue that truly excellent and well-conﬁrmed scientiﬁc theories cannot be overthrown by evidence without some explanation of why the theory is wrong. 4. Lewis 1986, pp. 240–241. See also the discussion of reﬂective equilibrium at the end of this chapter. 5. This sort of view ﬁnds voice in Wittgenstein 1953 and more recently in Lynch 1998. 6.

This is C. L. Hardin’s argument in Hardin 1999, p. 308.

7.

A defense of this view can be found in Rachels 2000.

8.

Sartre 1956, p. 51.

9.

Bealer 1996, pp. 16–17.

188

10.

Notes

Bealer 1996, p. 32, n. 26.

11. For more detailed discussions of the nature of the a priori, see Bonjour 1998, ch. 1 and Chisholm 1989b, ch. 4. 12. Compare Robert Audi’s principle that ‘‘if p is a necessary truth which, simply on the basis of understanding it, S attentively believes, then S’s belief that p is prima facie justiﬁed’’ (Audi 1993, p. 310). 13.

Kornblith 1998.

14.

Cf. Bealer 1996, p. 8, which defends this premise in a different way.

15.

Foley 1998, p. 242, passim.

16.

Ibid., esp. p. 254.

17. The twentieth century’s u¨berfoundationalist, Roderick Chisholm, makes clear this connection in Chisholm 1989b, pp. 28–99. 18. For a review of the vast and well-known literature concerning these stronger theses, see Plantinga 1993. 19. Weinberg’s criticisms are serious ones. Potshots at philosophy can be found in the writings of a number of scientists, for example Feynman (1986). 20. This is somewhat of an oversimpliﬁcation, since even if W ¼ 1, space is not uniformly ﬂat as a result of local relativistic distortion of spacetime by mass. Euclid had no notion of relativistic spacetime. An accessible discussion can be found in Smoot and Davidson 1993, pp. 158–163. 21. For discussion of Stich’s claims and his replies, see the symposium on his book in Philosophy and Phenomenological Research, vol. 51, no. 1, 1991. 22. A good critical discussion of Rawls on reﬂective equilibrium is Dworkin 1975. 23. Someone who admirably recognizes and defends moral intuitions as delivering foundational yet fallible beliefs which are then adjusted to achieve reﬂective equilibrium is McMahan (2000).

Chapter 2 1.

Boghossian 2001, p. 4.

2. According to a 2001 Gallup poll, 50% of Americans believe in ESP, over 40% believe that houses can be haunted, nearly 30% believe in astrology, and the same number believe in communication with the dead (Gallup Organization 2001). Furthermore, these percentages have increased over the last decade. This in our modern, scientiﬁc age! 3.

See esp. Swinburne 1992, ch. 5.

4. An excellent critique of Swinburne on this and related points is Byrne 1993.

Notes

5.

189

See Alston 1991, ch. 5.

6. Plantinga admits, though, that he is not arguing either that God exists or that the deliverances of revelation do in fact have warrant. His 500-page treatise Warranted Christian Belief is devoted to defending the conditional proposition that if there is a God, then Christian beliefs have warrant. See Plantinga 2000, pp. 186–190, 347. One is tempted to reply Az di bobe volt gehat beytsim volt zi geven mayn zeyde (if my grandmother had balls, she’d be my grandfather). 7.

See Blanshard 1975, p. 296, passim.

8.

Barth 1936, pp. 185–186.

9. A less conciliatory separate spheres doctrine is in Tillich. He writes: ‘‘There is no conﬂict between faith in its true nature and reason in its true nature. . . . Science has no right and no power to interfere with faith and faith has no power to interfere with science. One dimension of meaning is not able to interfere with another dimension’’ (1957, pp. 80–82). See also Tillich 1951, p. 130. 10. See esp. John Paul II 1998. Modern Catholic writers are not entirely in agreement on this, though, as we saw in the case of Cardinal Newman. 11. See Humanae vitae, §§14–15. For more on masturbation in particular, see the Roman Curia’s doctrinal document (Roman Curia 1975). 12. One philosopher who is a self-avowed dualist is Robert Almeder. Interestingly, however, Almeder defends dualism on empirical grounds, not on the basis of rational intuition. See Almeder 1992. I have criticized some of Almeder’s arguments in Hales 2001a, c. 13.

Nietzsche 1886 (1966), §46.

14.

See Wasson, Hofmann, and Ruck 1978, pp. 89–93.

15. See Mylonas 1961, pp. 224–227. Aristotle alludes to the Aeschylus matter in Nichomachean Ethics, book 3, ch. 1, 1111a. 16. The seminal study on the role of the ergot mushroom and the Eleusinian Mysteries is Wasson, Hofmann, and Ruck 1978. Ergot poisoning was a common event in medieval Europe. See Sidky 1997, pp. 167–183, and Schultes and Hofmann 1979, pp. 102–105. 17. The locus classicus is Wasson 1968. For criticism of the Soma ¼ ﬂy agaric view, see Rudgley 1993, ch. 3. 18. An excellent overview can be found in Schultes 1972. Schultes was the Harvard botanist responsible for identifying the teonanacatl of the Aztecs as the psilocybe mushroom. A more detailed study is Schultes and Hofmann 1979. 19. See ‘‘Shamanism and Peyote Use among Apaches of the Mescalero Indian Reservation’’ in Harner 1973b, pp. 53–66. 20. In the Vatican museums is a painting of a marble statue of one of the Roman gods, smashed on the ﬂoor. Behind the ruined statue, raised up on an altar, is a gleaming cruciﬁx. La plus c¸a change, la plus c’est la meme choses.

190

Notes

21.

This is the argument of Ehrman 2003.

22.

See Harner 1973b.

23.

Cited in Wasson 1972, p. 210.

24.

See, e.g., LaBarre 1972, and esp. Wasson et al. 1986.

25. Compare Isocrates, who, in his Panegyricus §28, praises Demeter for ‘‘two gifts, the greatest in the world—the fruits of the earth, which have enabled us to rise above the life of the beasts, and the holy rite which inspires in those who partake of it sweeter hopes regarding both the end of life and all eternity’’ (Isocrates 380 b.c.e. [1928]). 26.

My source on the Huichol is Furst 1972.

27. Further arguments along these lines and other examples of LSD experiences can be found in Pahnke 1967. 28.

Bealer 1996, p. 30, n. 15.

29. Plantinga defends Alston’s peer pressure argument in chapter 4 of Warranted Christian Belief (2000). 30.

See Plantinga 2000, p. 132, and Alston 1991, p. 220.

31.

Tidman 1996, pp. 169–170. Cf. Foley 1998, p. 242.

32.

Alston 1991, p. 153. Cf. Plantinga 2000, p. 130.

33.

He argues for these things in Stich 1990, esp. ch. 3.

34. The locus classicus on evolutionary spandrels is Gould and Lewontin 1979. 35.

Langford 1966, p. 41.

36. For example, John Paul II (1998) writes (Fides et ratio §82), ‘‘[a] relativist philosophy would be ill-adapted to help in the deeper exploration of the riches found in the word of God. . . . theology needs therefore the contribution of a philosophy which does not disavow the possibility of a knowledge which is objectively true.’’

Chapter 3 1. Portions of this chapter were originally published in Mind as Hales 1997a and Hales 1997b. 2.

I offer a detailed assessment of Lynch’s book in Hales 2001b.

3.

Nietzsche 1886 (1966), §18.

4.

Also see Aristotle, Metaphysics, bk. 4, chs. 4–5.

5. Margolis 1991, pp. 9–13. (N.B.: the kind of relativism Margolis considers ‘‘logically incoherent’’ and I show is not is what he calls ‘‘relationalism’’; Harris 1992, pp. 70–71, 82–84, 193; Nagel 1997, p. 15; Rorty 1991, p. 23).

Notes

191

6. P can be derived only within an axiomatic system of the sort presented in the appendix. As demonstrated there, P is a theorem provided commensurability is an equivalence relation. 7. The most sophisticated attempt along these lines is probably Bennigson 1999. 8. Some authors apparently use ‘‘self-refuting’’ and ‘‘paradoxical’’ indifferently. See Meiland 1980; Beach 1984; Margolis 1991. 9.

On this point, see (Lockie 2003, pp. 323–324).

10.

As argued in Plantinga 1974, pp. 51–54.

11. Davidson 1984, p. 190: ‘‘ ‘Incommensurable’ is, of course, Kuhn and Feyerabend’s word for ‘not intertranslatable.’ ’’ 12. Ibid., p. 184: ‘‘Studying the criteria of translation is . . . a way of focusing on criteria of identity for conceptual schemes.’’ See also p. 190: ‘‘The failure of intertranslatability is a necessary condition for difference of conceptual schemes. . . .’’ 13.

See Alston 1991, pp. 162–163, and p. 165, n. 37.

14. For a few of these opinions, see the essays in Loux 1979. See also Chisholm 1989a and Lewis 1986. 15.

Percival 1994.

16.

For pillorying, see Laudan 1990. For parodying, see Sokal 1996.

17.

For some discussion, see Kirkham 1992, sec. 3.2.

18. Wright 1992, p. 39. For essentially the same interpretation, see Lynch 1998, p. 103. 19. A view that he defends in Russell 1929. Actually Russell goes further than this—he thinks premarital sex is a good idea. 20. The basic argument is Sosa’s; see Sosa 1991, p. 7. It is developed in more detail in Plantinga 1982. 21. The presentation of this system is based on L. T. F. Gamut’s superb work in Gamut 1991. 22. See Gamut 1991, p. 26, for an explanation of why this expresses transitivity.

Chapter 4 1. A quick Internet search for the quotation from Lewontin will yield many examples. 2. Well, in point of fact Quine rejects Carnapian epistemology and calls for a scientiﬁc replacement. This is a typical Quinean strategy: criticize one theory in domain X, and then draw a grand, sweeping conclusion. Quine does this in

192

Notes

‘‘Two Dogmas of Empiricism’’ too: he criticizes a deﬁnition of analyticity in terms of referential transparency and concludes that since that deﬁnition didn’t work, the entire analytic–synthetic distinction is bogus. It is amazing how many philosophers go along with this form of argument. 3.

Kneale and Kneale 1962. See esp. pp. 556–568.

4. Armstrong 1989, pp. 19–20; Russell 1914b, p. 107; Hellman and Thompson 1975, p. 561; Lycan 1975, p. 224. 5.

Oliver (1996, p. 7) ﬂirts with this argument.

6. Obviously I am assuming Cantor’s continuum hypothesis, viz. that 2@0 ¼ @1 . Strictly, the former number expresses the cardinal number of the power set of a set containing @0 elements, and the latter is the cardinal number of the set of the well-orderings of the ﬁnite numbers. While I am assuming the continuum hypothesis for simplicity of exposition, nothing of philosophical interest here turns on it. 7.

Quine 1976a, p. 202.

8.

Quine 1960, p. 171.

9. There are other examples as well. Some particle theorists have held that there are inﬁnitely many types of hadrons, and appear not to be dissuaded from such spendthrift theories on parsimonious grounds. Steven Weinberg puts it, ‘‘in some theories there is an unlimited number of species: the number of types of particles will increase faster and faster as we explore higher and higher masses’’ (Weinberg 1977, p. 138). Cf. Hawking 1993, p. 110. 10. Lewis 1984, pp. 346, 364. Armstrong also accepts conjunctive properties, which bodes ill for his support of Ockham’s Razor. See Armstrong 1978, pp. 30– 42. Somewhat different criticisms of Lewis’s qualitative parsimony can be found in Oliver 1996, p. 7. 11.

Haugeland 1983, p. 4.

12. Endorsement of this interpretation is explicit in Sober 1990, where he offers further reasons to reject the alethic treatment of Ockham’s Razor. 13. These reﬂections seem to underlie Santayana’s suggestion that ‘‘Occam’s Razor . . . as a criterion of truth, is the weapon of a monstrous self-mutilation with which British philosophy, if consistent, would soon have committed suicide’’ (Santayana 1942, p. 510). 14. Stephen Stich suggested this response to me in conversation. I don’t know if he would continue to endorse it. 15.

Stephen Stich also suggested this line of argument to me in conversation.

16.

Stich 1990, ch. 5.

17. In, e.g., Rorty 1991, p. 24: ‘‘For the pragmatist . . . ‘knowledge’ is, like ‘truth,’ simply a compliment paid to the beliefs which we think so well justiﬁed that, for the moment, further justiﬁcation is not needed.’’

Notes

18.

193

See Whorf 1956, p. 216, and also his illustration on p. 210.

19. The seminal study is Martin 1986. Compelling follow-up criticisms can be found in Pullum 1991 and Pinker 1994, ch. 3. I repeat Pinker’s arguments in the present paragraph. 20.

Hume 1739 (1978), bk. I, pt. 4, §7.

References

Almeder, Robert. 1992. Death and Personal Survival: The Evidence for Life After Death. Lanham, Maryland: Rowman and Littleﬁeld. Alston, William P. 1991. Perceiving God: The Epistemology of Religious Experience. Ithaca: Cornell University Press. Aristotle. 1941. The Basic Works of Aristotle. Ed. Richard McKeon. Trans. W. D. Ross. New York: Random House. Armstrong, David M. 1978. A Theory of Universals, vol. 2: Universals and Scientiﬁc Realism. Cambridge: Cambridge University Press. ———. 1989. Universals: An Opinionated Introduction. Boulder: Westview. Audi, Robert. 1993. The Structure of Justiﬁcation. Cambridge: Cambridge University Press. Augustine. 400 (1952). The Confessions. Trans. E. B. Pusey. Ed. R. M. Hutchins. Vol. 18, The Great Books of the Western World. Chicago: The Encyclopedia Brittanica. Ayer, Alfred J. 1936. Language, Truth, and Logic. London: Gollancz. Barth, Karl. 1936. Credo. New York: Charles Scribner’s Sons. Beach, Edward. 1984. The Paradox of Cognitive Relativism Revisited. Metaphilosophy 15 (1):1–15. Bealer, George. 1996. On the Possibility of Philosophical Knowledge. In Philosophical Perspectives volume 10: Metaphysics, ed. J. E. Tomberlin. Cambridge: Blackwell. Bennigson, Thomas. 1999. Is Relativism Really Self-Refuting? Philosophical Studies 94:211–236. Blanshard, Brand. 1975. Reason and Belief. New Haven: Yale University Press. Boghossian, Paul. 2000. Knowledge of Logic. In New Essays on the A Priori, ed. P. Boghossian and C. Peacocke. Oxford: Oxford University Press. ———. 2001. How Are Objective Epistemic Reasons Possible? Philosophical Studies 106:1–40.

196

References

Bonjour, Laurence. 1998. In Defense of Pure Reason: A Rationalist Account of A Priori Justiﬁcation. Cambridge: Cambridge University Press. ———. 2002. Epistemology. Lanham, Maryland: Rowman and Littleﬁeld. Byrne, Peter. 1993. A Defence of Christian Revelation. Religious Studies 29:381– 394. Calvin, John. 1559 (1845). The Institutes of the Christian Religion. Trans. H. Beveridge. Edinburgh: The Calvin Translation Society. Carnap, Rudolf. 1959. The Elimination of Metaphysics through the Logical Analysis of Language. In Logical Positivism, ed. A. J. Ayer. New York: The Free Press. The Catechism of the Catholic Church. Available at http://www.vatican.va/ archive/ccc/index.htm/. Chalmers, David. 1996. The Conscious Mind: In Search of a Fundamental Theory. Oxford: Oxford University Press. Chisholm, Roderick M. 1977. Theory of Knowledge, 2nd ed. Englewood Cliffs, N.J.: Prentice Hall. ———. 1982. The Foundations of Knowing. Minneapolis: University of Minnesota Press. ———. 1989a. On Metaphysics. Minneapolis: University of Minnesota Press. ———. 1989b. Theory of Knowledge, 3rd ed. Englewood Cliffs, N.J.: Prentice Hall. Cicero, Marcus Tullius. 45 b.c.e. (1928). De legibus. Trans. C. W. Keyes. Ed. T. E. Page, E. Capps, W. H. D. Rouse, L. A. Post and E. H. Warmington. The Loeb Classical Library. Cambridge, Mass.: Harvard University Press. Davidson, Donald. 1984. On the Very Idea of a Conceptual Scheme. In his Inquiries Into Truth and Interpretation. Oxford: Oxford University Press. Dennett, Daniel. 1984. Elbow Room: The Varieties of Free Will Worth Wanting. Cambridge, Mass.: MIT Press. DePaul, Michael R., and William Ramsey, eds. 1998. Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry. Lanham, Maryland: Rowman and Littleﬁeld. Descartes, Rene´. 1628 (1985). Rules for the Direction of the Mind, Trans. R. S. J. Cottingham and D. Murdoch. In The Philosophical Writings of Descartes, vol. 1 Cambridge: Cambridge University Press. ———. 1644 (1985). Principles of Philosophy. Trans. R. S. J. Cottingham and D. Murdoch. In The Philosophical Writings of Descartes, vol. 1 Cambridge: Cambridge University Press. Devitt, Michael. 1997. Realism and Truth, 2nd ed. Princeton: Princeton University Press.

References

197

———. 1999. A Naturalistic Defense of Realism. In Metaphysics: Contemporary Readings, ed. S. D. Hales. Belmont, Calif.: Wadsworth. Doblin, Rick. 1991. Pahnke’s ‘‘Good Friday Experiment’’: A Long-Term FollowUp and Methodological Critique. Journal of Transpersonal Psychology 23 (1):1– 28. Duncker, Karl. 1939. ‘‘Ethical Relativity? (An Inquiry into the Psychology of Ethics).’’ Mind 48:39–57. Dworkin, Ronald. 1975. The Original Position. In Reading Rawls, ed. N. Daniels. New York: Basic Books. Ehrman, Bart D. 2003. Lost Christianities. Oxford: Oxford University Press. Elgin, Catherine Z. 1996. Considered Judgment. Princeton: Princeton University Press. Escohotado, Antonio. 1999. A Brief History of Drugs: From the Stone Age to the Stoned Age. Trans. K. A. Symington. Rochester, Vermont: Park Street Press. Fernandez, James W. 1972. Tabernanthe Iboga: Narcotic Ecstasis and the Work of the Ancestors. In The Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger Publishers. Feynman, Richard. 1965. The Character of Physical Law. Cambridge, Mass.: MIT Press. ———. 1986. Surely You’re Joking, Mr. Feynman! New York: Bantam Books. Foley, Richard. 1998. Rationality and Intellectual Self-Trust. In Rethinking Intuition, ed. M. R. DePaul and W. Ramsey. Lanham, Maryland: Rowman and Littleﬁeld. Frege, Gottlob. 1884. Die Grundlagen der Arithmetik. Breslau: W. Ko¨bnner. Furst, Peter T. 1972. To Find Our Life: Peyote among the Huichol Indians of Mexico. In Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger Publishers. Galilei, Galileo. 1610. Sidereus nuncius. Venice: Apud Thomam Baglionum. Gallup Organization. 2001. Americans’ Belief in Psychic Paranormal Phenomena Is Up Over Last Decade. Princeton, N.J.: Gallup. Gamut, L. T. F. 1991. Intensional Logic and Logical Grammar, vol. 2: Logic, Language, and Meaning. Chicago: University of Chicago Press. Gilson, Etienne. 1938. Reason and Revelation in the Middle Ages. New York: Charles Scribner’s Sons. Goldman, Alvin. 1986. Epistemology and Cognition. Cambridge, Mass.: Harvard University Press. ———. 1999. A Priori Warrant and Naturalistic Epistemology. In Philosophical Perspectives, vol. 13: Epistemology, ed. J. E. Tomberlin. Oxford: Blackwell.

198

References

Goldman, Alvin, and Joel Pust. 1998. Philosophical Theory and Intuitional Evidence. In Rethinking Intuition: The Psychology of Intuition and its Role in Philosophical Inquiry, ed. M. DePaul and W. Ramsey. Lanham, Maryland: Rowman and Littleﬁeld. Goodman, Nelson. 1954. Fact, Fiction, and Forecast. Cambridge, Mass.: Harvard University Press. Gould, Stephen Jay, and Richard C. Lewontin. 1979. The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. Proceedings of the Royal Society of London 205, series B (1161):581– 598. Hacker, P. M. S. 1996. On Davidson’s Idea of a Conceptual Scheme. Philosophical Quarterly 46 (184):289–307. Hales, Steven D. 1997a. A Consistent Relativism. Mind 106 (421):33–52. ———. 1997b. Reply to Shogenji on Relativism. Mind 106 (424):749–750. ———. 2000. The Problem of Intuition. American Philosophical Quarterly 37 (2):135–147. ———. 2001a. Evidence and the Afterlife. Philosophia 28 (1–4):335–346. ———. 2001b. Lynch’s Metaphysical Pluralism. Philosophy and Phenomenological Research 63 (3):699–709. ———. 2001c. Reincarnation Redux. Philosophia 28 (1–4):359–367. Hales, Steven D., and Rex Welshon. 2000. Nietzsche’s Perspectivism. Urbana, Ill.: University of Illinois Press. Hardin, C. L. 1999. Color and Illusion. In Metaphysics: Contemporary Readings, ed. S. D. Hales. Belmont, Calif.: Wadsworth. Harner, Michael J. 1973a. Common Themes in South American Indian Yage´ Experiences. In Hallucinogens and Shamanism, ed. M. J. Harner. Oxford: Oxford University Press. ———, ed. 1973b. Hallucinogens and Shamanism. Oxford: Oxford University Press. Harris, James F. 1992. Against Relativism. La Salle, Ill.: Open Court. Haugeland, John. 1983. Ontological Supervenience. Southern Journal of Philosophy 22 (supplemental):1–12. Hawking, Stephen W. 1993. The Quantum Mechanics of Black Holes. In his Black Holes and Baby Universes and Other Essays. New York: Bantam Books. Hellman, Geoffrey, and Frank Thompson. 1975. Physicalism: Ontology, Determination, and Reduction. Journal of Philosophy 72 (17):551–564. Henderson, David, and Terry Horgan. 2000. What Is A Priori and What Is It Good For? Southern Journal of Philosophy 38 (supplement):51–86. Hesse, Herman. 1949. Magister Ludi. New York: Henry Holt.

References

199

Hintikka, Jaakko. 1999. The Emperor’s New Intuitions. Journal of Philosophy 96 (3):127–147. Hume, David. 1739 (1978). A Treatise of Human Nature, 2nd ed. Ed. L. A. Selby-Bigge. Oxford: Oxford University Press/Clarendon Press. Isocrates. 380 b.c.e. (1928). Panegyricus. Trans. G. Norlin. Ed. T. E. Page, E. Capps, W. H. D. Rouse, L. A. Post, and E. H. Warmington, The Loeb Classical Library. Cambridge, Mass.: Harvard University Press. Jackson, Frank. 1998. From Metaphysics to Ethics: A Defence of Conceptual Analysis. Oxford: Oxford University Press. Jaspers, Karl. 1967. Philosophical Faith and Revelation. Trans. E. B. Ashton. New York: Harper and Row. John Paul II. 1995. Evangelium vitae. Rome: Vatican. ———. 1998. Fides et ratio. Rome: Vatican. Kant, Immanuel. 1787 (1965). Critique of Pure Reason. Trans. N. K. Smith. New York: St. Martin’s Press. Kass, Leon. 1997. The Wisdom of Repugnance. New Republic 216:17–26. Kim, Jaegwon. 1988. What Is ‘Naturalized Epistemology’? In Philosophical Perspectives 2: Epistemology, ed. J. E. Tomberlin. Atascadero, Calif.: Ridgeview. ———. 1996. Philosophy of Mind. Boulder: Westview. Kirkham, Richard L. 1992. Theories of Truth. Cambridge, Mass.: MIT Press. Kneale, William, and Martha Kneale. 1962. The Development of Logic. Oxford: Oxford University Press. Kornblith, Hilary. 1998. The Role of Intuition in Philosophical Inquiry: An Account with No Unnatural Ingredients. In Rethinking Intuition, ed. M. R. DePaul and W. Ramsey. Lanham, Maryland: Rowman and Littleﬁeld. Krausz, Michael, and Jack W. Meiland, eds. 1982. Relativism: Cognitive and Moral. Notre Dame: University of Notre Dame Press. Kripke, Saul. 1980. Naming and Necessity, 2nd ed. Cambridge, Mass.: Harvard University Press. Kuhn, Thomas S. 1970. The Structure of Scientiﬁc Revolutions, 2nd ed. Chicago: University of Chicago Press. Ko¨lbel, Max. 1999. Saving Relativism from Its Saviour. Crı´tica, Revista Hispanoamericana de Filosoﬁa 31 (91):91–103. LaBarre, Weston. 1972. Hallucinogens and the Shamanic Origins of Religion. In Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger Publishers. Langford, Jerome J. 1966. Galileo, Science, and the Church. Ann Arbor: University of Michigan Press.

200

References

Laudan, Larry. 1990. Science and Relativism. Chicago: University of Chicago Press. Leibniz, Gottfried Wilhelm. 1696 (1949). New Essays Concerning Human Understanding. La Salle, Ill.: Open Court. Lewis, David. 1973. Counterfactuals. Cambridge, Mass.: Harvard University Press. ———. 1984. New Work for a Theory of Universals. Australasian Journal of Philosophy 62:343–377. ———. 1986. On the Plurality of Worlds. Oxford: Blackwell. Lewontin, Richard C. 1997. Billions and Billions of Demons. New York Review of Books (January 9):28–32. Locke, John. 1690 (1974). An Essay Concerning Human Understanding. Ed. A. D. Woozley. New York: Penguin Books. Lockie, Robert. 2003. Relativism and Reﬂexivity. International Journal of Philosophical Studies 11 (3):319–339. Loux, Michael J., ed. 1979. The Possible and the Actual: Readings in the Metaphysics of Modality. Ithaca: Cornell University Press. Lycan, William G. 1975. Occam’s Razor. Metaphilosophy 6 (3–4):223–237. Lynch, Michael P. 1998. Truth in Context: An Essay on Pluralism and Objectivity. Cambridge, Mass.: MIT Press. MacIntyre, Alasdair. 1985. Relativism, Power, and Philosophy. Proceedings and Addresses of the American Philosophical Association 59 (1):5–22. Margolis, Joseph. 1991. The Truth about Relativism. Oxford: Blackwell. Martin, Laura. 1986. ‘‘Eskimo Words for Snow’’: A Case Study in the Genesis and Decay of an Anthropological Example. American Anthropologist 88:418– 423. Luther, Martin. 1517 (1957). ‘‘Disputatio contra scholasticam theologiam.’’ In Luther’s Works, vol. 31. Ed. and trans. Harold J. Grimm. Philadelphia: Fortress Press. ———. 1536 (1960). ‘‘Disputatio de homine.’’ In Luther’s Works, vol. 34. Ed. and trans. Lewis W. Spitz. Philadelphia: Fortress Press. McGinn, Colin. 1991. The Problem of Consciousness. Oxford: Blackwell. McKenna, Terence. 1992. Food of the Gods: The Search for the Original Tree of Knowledge. New York: Bantam Books. McMahan, Jeff. 2000. Moral Intuition. In The Blackwell Guide to Ethical Theory, ed. H. LaFollette. Oxford: Blackwell. Meiland, Jack W. 1980. On the Paradox of Cognitive Relativism. Metaphilosophy 11 (2):115–126.

References

201

Monsay, Evelyn H. 1997. Intuition in the Development of Scientiﬁc Theory and Practice. In Intuition: The Inside Story, ed. R. D. Floyd and P. S. Arvidson. New York: Routledge. Moody-Adams, Michele M. 1997. Fieldwork in Familiar Places: Morality, Culture, and Philosophy. Cambridge, Mass.: Harvard University Press. Mylonas, George E. 1961. Eleusis and the Eleusinian Mysteries. Princeton: Princeton University Press. Nagel, Thomas. 1997. The Last Word. Oxford: Oxford University Press. Newman, John Henry (Cardinal). 1892. Apologia pro vita sua. Rev. ed. London: Longman, Green. Newton-Smith, W. 1982. Relativism and the Possibility of Interpretation. In Rationality and Relativism, ed. M. Hollis and S. Lukes. Cambridge, Mass.: MIT Press. Niebuhr, H. Richard. 1962. The Meaning of Revelation. New York: Macmillan. Nielsen, Kai. 1993. Relativism and Wide Reﬂective Equilibrium. Monist 76 (3):316–332. Nietzsche, Friedrich. 1882 (1974). The Gay Science. Ed. and trans. Walter Kaufmann. New York: Random House. ———. 1886 (1966). Beyond Good and Evil. Trans. W. Kaufmann. Ed. W. Kaufmann. New York: Vintage Books. ———. 1895 (1968). The Antichrist. Trans. R. J. Hollingdale. Ed. R. J. Hollingdale. New York: Viking Penguin. ———. 1910 (1968). The Will to Power. Ed. and trans. Walter Kaufmann. New York: Random House. Nozick, Robert. 2001. Invariances: The Structure of the Objective World. Cambridge, Mass.: Harvard University Press. Oliver, Alex. 1996. The Metaphysics of Properties. Mind 105 (417):1–80. Pahnke, Walter N. 1963. ‘‘Drugs and Mysticism: An Analysis of the Relationship between Psychedelic Drugs and the Mystical Consciousness.’’ Ph.D. dissertation, Harvard University, Cambridge, Mass. ———. 1967. LSD and Religious Experience. In LSD, Man, and Society, ed. R. C. DeBold and R. C. Leaf. Middletown, Conn.: Wesleyan University Press. Paul IV. 1968. Humanae vitae. Rome: Vatican. Percival, Philip. 1994. Absolute Truth. Proceedings of the Aristotelian Society 94:189–213. Pinker, Steven. 1994. The Language Instinct. New York: William Morrow. Plantinga, Alvin. 1974. The Nature of Necessity. Oxford: Oxford University Press.

202

References

———. 1982. How to Be an Anti-Realist. Proceedings and Addresses of the American Philosophical Association 56 (1):47–70. ———. 1993. Warrant: The Current Debate. Oxford: Oxford University Press. ———. 2000. Warranted Christian Belief. Oxford: Oxford University Press. Plato. 1987a. Euthyphro. Trans. Lane Cooper. In The Collected Dialogues of Plato. Ed. Edith Hamilton and Huntington Cairns. Princeton: Princeton University Press. ———. 1987b. Laws. Trans. A. E. Taylor. In The Collected Dialogues of Plato. Ed. Edith Hamilton and Huntington Cairns. Princeton: Princeton University Press. ———. 1987c. Republic. Trans. Paul Shorey. In The Collected Dialogues of Plato. Ed. Edith Hamilton and Huntington Cairns. Princeton: Princeton University Press. ———. 1987d. Theaetetus. Trans. F. M. Cornford. In The Collected Dialogues of Plato. Ed. Edith Hamilton and Huntington Cairns. Princeton: Princeton University Press. ———. 1987e. Timaeus. Trans. Benjamin Jowett. In The Collected Dialogues of Plato. Ed. Edith Hamilton and Huntington Cairns. Princeton: Princeton University Press. Pullum, Geoffrey K. 1991. The Great Eskimo Vocabulary Hoax and Other Irreverent Essays on the Study of Language. Chicago: University of Chicago Press. Putnam, Hilary. 1981. Reason, Truth, and History. Cambridge: Cambridge University Press. ———. 1983. Why Reason Can’t Be Naturalized. In his Realism and Reason. Cambridge: Cambridge University Press. Quine, W. V. O. 1953. ‘‘Two Dogmas of Empiricism.’’ In his From a Logical Point of View. Cambridge, Mass.: Harvard University Press. ———. 1960. Word and Object. Cambridge, Mass.: MIT Press. ———. 1969. Epistemology Naturalized. In his Ontological Relativity and Other Essays. New York: Columbia University Press. ———. 1976a. A Logistical Approach to the Ontological Problem. In his The Ways of Paradox and Other Essays. Cambridge, Mass.: Harvard University Press. ———. 1976b. On Multiplying Entities. In his The Ways of Paradox and Other Essays. Cambridge, Mass.: Harvard University Press. ———. 1976c. Quantiﬁers and Propositional Attitudes. In his The Ways of Paradox and Other Essays. Cambridge, Mass.: Harvard University Press. ———. 1995. From Stimulus to Science. Cambridge, Mass.: Harvard University Press.

References

203

Quine, W. V. O., and J. S. Ullian. 1970. The Web of Belief. New York: Random House. Rachels, James. 2000. Naturalism. In The Blackwell Guide to Ethical Theory, ed. H. LaFollette. Oxford: Blackwell. Rawls, John. 1971. A Theory of Justice. Cambridge, Mass.: Harvard University Press. Reichel-Dolmatoff, Gerardo. 1972. The Cultural Context of an Aboriginal Hallucinogen: Banisteriopsis Caapi. In Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger. Reid, Thomas. 1764 (1975). An Enquiry into the Human Mind on the Principles of Common Sense. In Thomas Reid’s Inquiry and Essays, ed. R. Beanblossom and K. Lehrer. Indianapolis: Bobbs-Merrill. Roman Curia. 1975. Declaration on Certain Questions Concerning Sexual Ethics—Persona Humana. Rome: Vatican. Rorty, Richard. 1986. From Logic to Language to Play. Proceedings and Addresses of the American Philosophical Association 59 (5):747–753. ———. 1991. Solidarity or Objectivity. In his Objectivity, Relativism, and Truth. Cambridge: Cambridge University Press. Rudgley, Richard. 1993. Essential Substances: A Cultural History of Intoxicants in Society. New York: Kodansha International. Russell, Bertrand. 1914a. On the Nature of Acquaintance. In Logic and Knowledge, ed. R. C. Marsh. London: Unwin Hyman. ———. 1914b. Our Knowledge of the External World. Chicago: Open Court. ———. 1927. Why I Am Not a Christian. In The Basic Writings of Bertrand Russell, ed. R. E. Egner and L. E. Denonn. New York: Simon and Schuster. ———. 1929. Marriage and Morals. London: Horace Liveright. Santayana, George. 1942. Realms of Being. New York: Charles Scribner’s Sons. Sartre, Jean-Paul. 1956. Being and Nothingness. Trans. H. E. Barnes. New York: Philosophical Library. Schmitt, Frederick F. 1995. Truth: A Primer. Boulder: Westview. Schultes, Richard Evans. 1972. An Overview of Hallucinogens in the Western Hemisphere. In Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger. Schultes, Richard Evans, and Albert Hofmann. 1979. Plants of the Gods: Origins of Hallucinogenic Use. New York: McGraw-Hill Book Company. Shogenji, Tomoji. 1997. The Consistency of Global Relativism. Mind 106 (424):745–747. Sider, Theodore. 2001. Four-Dimensionalism: An Ontology of Persistence and Time. Oxford: Oxford University Press.

204

References

Sidky, H. 1997. Witchcraft, Lycanthropy, Drugs, and Disease: An Anthropological Study of European Witch-Hunts. New York: Peter Lang. Smith, Huston. 1964. Do Drugs Have Religious Import? Journal of Philosophy 61 (18):517–530. Smoot, George, and Keay Davidson. 1993. Wrinkles in Time. New York: Avon Books. Sober, Elliott. 1990. Let’s Razor Ockham’s Razor. In Explanation and Its Limits, ed. D. Knowles. Cambridge: Cambridge University Press. Sokal, Alan D. 1996. Transgressing the Boundaries: Toward a Transformative Hermeneutics of Quantum Gravity. Social Text 14 (1–2):217–252. Sosa, Ernest. 1987. Serious Philosophy and Freedom of Spirit. Journal of Philosophy 84:707–726. ———. 1991. Knowledge in Perspective. Cambridge: Cambridge University Press. ———. 1996. Rational Intuition: Bealer on Its Nature and Epistemic Status. Philosophical Studies 81:151–162. Stace, W. T. 1960. Mysticism and Philosophy. London: Macmillan. Stich, Stephen P. 1990. The Fragmentation of Reason. Cambridge, Mass.: MIT Press. Stich, Stephen P., and Jonathan M. Weinberg. 2001. Jackson’s Empirical Assumptions. Philosophy and Phenomenological Research 62 (3):637–643. Swinburne, Richard. 1992. Revelation: From Metaphor to Analogy. Oxford: Oxford University Press. Swoyer, Chris. 1982. True For. In Relativism: Cognitive and Moral, ed. M. Krausz and J. W. Meiland. Notre Dame: University of Notre Dame Press. Tertullian. c. 210 (1950). On the Soul. In Apologetical Works, trans. E. A. Quain. Washington: The Catholic University of America Press. Tidman, Paul. 1996. The Justiﬁcation of A Priori Intuitions. Philosophy and Phenomenological Research 56 (1):161–171. Tillich, Paul. 1951. Systematic Theology, vol. 1. Chicago: University of Chicago Press. ———. 1957. Dynamics of Faith. New York: Harper. Unger, Peter. 1984. Philosophical Relativity. Minneapolis: University of Minnesota Press. Van Cleve, James. 1983. Conceivability and the Cartesian Argument for Dualism. Paciﬁc Philosophical Quarterly 64 (1):35–45. Wasson, R. Gordon. 1968. Soma: Divine Mushroom of Immortality. New York: Harcourt, Brace, and World.

References

205

———. 1972. What Was the Soma of the Aryans? In Flesh of the Gods: The Ritual Use of Hallucinogens, ed. P. T. Furst. New York: Praeger. Wasson, R. Gordon, Albert Hofmann, and Carl A. P. Ruck. 1978. The Road to Eleusis: Unveiling the Secret of the Mysteries. New York: Harcourt Brace Jovanovich. Wasson, R. Gordon, Stella Kramrisch, Jonathan Ott, and Carl A. P. Ruck. 1986. Persephone’s Quest: Entheogens and the Origins of Religion. New Haven: Yale University Press. Weinberg, Jonathan M., Shaun Nichols, and Stephen Stich. 2001. Normativity and Epistemic Intuitions. Philosophical Topics 29 (1–2):429–460. Weinberg, Steven. 1977. The First Three Minutes. New York: Basic Books. ———. 1992. Dreams of a Final Theory. New York: Pantheon Books. Whorf, Benjamin Lee. 1956. Language, Thought, and Reality. Ed. J. B. Carroll. Cambridge, Mass.: MIT Press. Wittgenstein, Ludwig. 1953. Philosophical Investigations. Trans. G. E. M. Anscombe. Oxford: Basil Blackwell. ———. 1961. Tractatus Logico-Philosophicus. Trans. D. F. Pears and B. F. McGuinness. London: Routledge and Kegan Paul. ———. 1969. On Certainty. Trans. D. Paul and G. E. M. Anscombe. Ed. G. E. M. Anscombe and G. H. von Wright. Oxford: Basil Blackwell. Wright, Crispin. 1992. Truth and Objectivity. Cambridge, Mass.: Harvard University Press.

Index

Abortion, 63–64 Absolutism, 99–101, 105, 121, 126– 128, 131, 142 Abstract intensional objects, 113 Accessibility relation, 106 Adam and Eve, 64 Aeschylus, 70 Agnosticism, 67 Alcibiades, 70 Alethic modality, 44, 102, 114 Alston, William P., 53–54, 66, 69, 75– 76, 81–83, 113–114, 168 Amahuaca, 73 Amanita muscaria, 71, 83. See also Fly agaric mushroom Analytic philosophy, 15, 22–23, 32, 41, 69, 79 Analytic rationalism, 119, 169 Animists, 51 Anselm, 72 Apaches of the Mescalero Reservation, 71 A posteriori justiﬁcation. See Justiﬁcation, a posteriori A priori intuition. See Intuition, a priori A priori justiﬁcation. See Justiﬁcation, a priori A priori knowledge. See Knowledge, a priori A priori propositions. See Propositions, a priori Aquinas, Thomas, 58, 62, 72 Aristotle, 20, 58, 87

Armstrong, David, 157 Aryans of Northern India, 70 Atheism, 67 Augustine, 54–55, 64–65 Ayahuasca, 71, 73 Ayer, A. J., 24–26, 46 Aztecs, 71, 76 Bacchus, 69, 72 Banisteriopsis caapi, 71, 73, 83 Barth, Karl, 59–60 Bealer, George, 20, 27–28, 36, 80–81, 113 Bennigson, Thomas, 102 ‘‘Big Chill,’’ the, 37 ‘‘Big Crunch,’’ the, 38 Birth control, 63–64 Blanshard, Brand, 68 Boghossian, Paul, 11, 34–35, 98 Bonjour, Laurence, 12, 20, 25, 29, 36, 89 British empiricists, 168 Broackes, Justin, 22–23 Bwiti, 73 Calviceps purpurea, 70 Calvin, John, 51, 58–60, 72, 150 Carnap, Rudolf, 95 Cartesian argument, 153, 154. See also Descartes, Rene´ Cartesian ﬁrst principles, 53. See also Descartes, Rene´ Cartesian skepticism, 28, 46–47, 155. See also Descartes, Rene´; Skepticism

208

Catechism of the Catholic Church, the, 64–65 Catholic Church, the, 52, 60, 62, 78 Catholic doctrine, 65 Catholic perspective, 133, 134, 141 Catholic theology, 79 Catholicism, 56, 58, 62–63, 168 Causal theory of reference, 31 Causation, 29, 39 Cephalus, 9 Chalmers, David, 11, 66 Chisholm, Roderick, 20, 149–150 Chomsky, Noam, 172 Christianity, 45, 52, 56, 61–62, 69, 71–73, 123, 168, 182 Christian belief, 54, 68 Christian doctrine, 67 Christian doxastic practice, 114 Christian ministers, 72 Christian Mystical Practice, 53, 81, 83 Christian perspective, 112–113, 119, 127 Christian revelation, 50–51, 77–78, 81–82, 87–88, 114, 125, 127, 129, 138, 148, 168–169, 184 (see also Revelation) Christian theologians, 62, 79 Chrysippus, 57 Cicero, 74 Circularity, 34 Cognitivism, 23 Coherence theory of justiﬁcation, 43, 80 Coherentism, 33–35 Color, 22–23 metamers, 38 wave-length theory of, 38 Commensurability, 106–111, 115– 118 Concept formation, 30 Concepts, 15–19 Conceptual analysis, 15–19 Conceptual optimism, 17 Conceptual relativism, 126. See also Relativism

Index

Conceptual-role semantics, 34 Conceptual scheme, 109–110. See also Lynch, Michael P. Conﬂict thesis, 55–56, 59, 62, 77 Conibo-Shipibo, 73 Conjunctivism, 160 Consequentialism, 39, 173 Consequentialist intuitions. See Intuition, consequentialist Conservatism, 157 Contextualists, 153 Contingency, 104 Contingent truths, 20 Continualism, 167 Contraception. See Birth control Correspondence theory, 140–141 Creationists, 149 Criteria for knowledge, 149–150 Cultural relativism, 179 Darwin, Charles, 37, 79 Darwinism, 22 Davidson, Donald, 105, 109–110 Dawkins, Richard, 22 De anima, 56. See also Tertullian Deduction, 15 Deductive inferences. See Inferences, deductive De legibus, 74 Dennett, Daniel, 11–12 Deontological intuition. See Intuition, deontological DePaul, Michael R., 12 Desana, 73 Descartes, Rene´, 14–15, 19–20, 25, 36–37, 39, 41, 66, 153 Descriptive cultural relativism, 179– 183 Determinism, 21 Development of Logic, The, 156. See also Kneale, William and Martha Devitt, Michael, 29, 152–155 Dispositionalism, 22 ‘‘Disputatio contra scholasticam theologiam,’’ 58. See also Luther, Martin

Index

‘‘Disputatio de homine,’’ 57. See also Luther, Martin Disquotational theory, 140 Divine, the, 51 Doctrine of the Trinity, 58 Doctrine of Transubstantiation, 60 Doxastic methods, 89 Doxastic perspectives, 127, 131, 184 Doxastic practice, 43, 54, 80–82, 86, 88, 125, 182–184 Doxastic system, 126, 134 Dreams of a Final Theory, 37 Dretske’s zebras, 170 Duncker, Karl, 180 Eboka plant, 74 Einstein, Albert, 37 Eleusinian hierophant, 150 Eleusinian Mysteries, 70, 72, 74 Elgin, Catherine Z., 44–45 Emotivist epistemology, 176 Empirical hypotheses, 158 Empirical intuition. See Intuition, empirical Empirical knowledge, 121, 154 Empirical propositions. See Propositions, empirical Empirical psychology, 92, 151. See also Naturalized epistemology Empirical relativism. See Relativism, empirical Empirical science, 30, 122 Empiricism, 33, 124, 148 Entheogenic substances, 74 Epicurus, 57 Epistemic intuitions. See Intuition, epistemic Epistemic justiﬁcation. See Justiﬁcation, epistemic Epistemic norms, 174, 177 Epistemic perspective, 148, 182 Epistemic skepticism, 147. See also Skepticism Epistemic theory of truth, 131–132, 135, 137–138, 140–141 Epistemic value, 44, 166

209

Epistemological naturalism, 24. See also Naturalism Epistemological propositions. See Propositions, epistemic Epistemology, 24, 27, 44, 46, 47, 92, 131–132, 135, 137–138, 140–141, 147, 148, 151, 166, 170, 174, 177, 182 ideal epistemic conditions, 135, 136, 138–139 naturalized, 151, 155 religious, 58 traditional, 168–169, 171, 173–174 Western, 174 Ergot mushroom, 70 Eschatology, 65 Escohotado, Antonio, 69, 74 Eskimos, 178–180 Essay Concerning Human Understanding, An, 60. See also Locke, John Ethics, 24, 39, 53, 65, 78, 180 Ethnography, 177–178 Eucharist, 72 Euclidean geometry, 36–37 Euthanasia, 63–64 Euthyphro, 15–16, 19, 148. See also Plato Evangelium vitae, 63. See also Pope John Paul II Externalism, 88–89 Externalist justiﬁcation. See Justiﬁcation, externalist Extraperspectival truth, 103 Extrasensory perception, 51 Feynman, Richard, 13, 166, 172 Fides et ratio, 55, 67. See also Pope John Paul II Fly agaric mushroom, 71, 76, 82 Fodor, Jerry, 135 Foley, Richard, 32, 82 Formalism, 106 Foundationalism, 32, 33, 34, 35, 47 Free will, 16, 98 Frege, Gottlob, 36

210

Fundamental theology, 55 Fundamental vocabulary, 18–19 Galileo, 79, 87–88 Gambler’s fallacy, 41, 51 Garden of Eden, 64 Gauss-Lobachevskian geometry, 37 Gay Science, The, 18. See also Nietzsche, Friedrich Gestalt switch, 107–108 Gettier, Edmund, 10, 16–18, 21, 35, 46, 170–172 Gettier case, 170–172 Gilson, Etienne, 53–54 Global absolutism, 113 Global naturalism, 151, 166 Global ontologies, 158–159 Global relativism. See Relativism, global God, 51–66, 69, 72, 75–76, 112, 127 God-perceptions, 75 Goldman, Alvin, 89, 167 Goldman’s process reliabilism, 88 Goodman, Nelson, 40–41 Hacker, P. M. S., 106 Hallucinogen-based moral beliefs, 75, 78 Hallucinogenic traditions, 168 Hallucinogens, 69, 72, 76, 79, 81, 83, 85, 87, 125, 127, 130–131, 168, 183 ritual use of, 50, 69, 71, 78, 85, 88– 90, 130, 138, 148, 152, 169, 179, 184 Harner, Michael, 73, 77 Haugeland, John, 163 Heaven’s Gate (cult), 54 Hell, 66 Hellman, Geoffrey, 157 Henderson, David, 39 Heraclitus, 57 Himalayan shamans, 76 Hintikka, Jaakko, 12, 19 Homosexuality, 63 Huichol of Mexico, 74, 78, 81, 112, 179

Index

Humanae vitae, 63. See also Pope Paul IV Human cloning, 10 Human evolution, 168 Hume, David, 28, 66, 150, 184 Humean skepticism, 185. See also Skepticism Ideal reasoner, 135–136 Immortal soul, 66 Incommensurability, 107–108 Indeterminacy of translation, 28 Inductive inferences. See Inferences, inductive Inferences, deductive, 40 Inferences, inductive, 40 Institutes of the Christian Religion, 59. See also Calvin, John Instrumental account of rationality, 175 Internalism, 88–89 Internalist justiﬁcation. See Justiﬁcation, internalist Internal relativism. See Relativism, internal Intuition, 10–12, 15–16, 20–22, 83 a priori, 29, 42, 87 consequentialist, 133 deontological, 133 empirical, 13, 20, 172 epistemic, 170, 173, 177 intuition-based arguments, 34 intuition-based philosophy, 35, 64 intuition-driven epistemology, 151 (see also Epistemology) intuitive knowledge, 36 moral, 64, 181 philosophical, 12, 41, 61, 77, 86, 147, 171 physical, 12–13, 171 problem of, 26–27 rational, 12–21, 24, 26, 29–44, 46–47, 50, 55–56, 59–62, 66–67, 77–78, 81–90, 130–131, 133, 138–139, 147–148, 152, 154–155, 167, 168–170, 172, 174, 177, 184– 185

Index

scientiﬁc, 12 sensible, 13 Jackson, Frank, 16–19, 66 Jaspers, Karl, 53 Jesus of Nazareth, 61, 65, 72 Jı´varo of Ecuadorian Amazon, 71, 73, 78, 134, 179 Jones, James, 54 Judaism, 72 Justiﬁcation, 24, 26 a posteriori, 28, 30 a priori, 28–29 epistemic, 26, 134, 157 externalist, 30, 86 internalist, 86 justiﬁed belief, 25, 138 justiﬁed true belief account of knowledge, 35, 172 justiﬁed true belief account of philosophy, 19 naturalized, 157 simpliciter, 137 Kant, Immanuel, 36–37, 153 Kantian metaphysics, 154–155 Kass, Leon, 10–11 Kim, Jaegwon, 34, 66, 151, 159 Kirkham, Richard, 137–138, 140 Kneale, William and Martha, 156 Knower paradox, 93–95, 119–120 Knowledge, a priori, 25, 29, 39, 47 Ko¨lbel, Max, 115, 117 Kornblith, Hilary, 30–31, 158 Krausz, Michael, 109. See also Meiland, Jack W. Kripke, Saul, 11, 16, 31 Kuhn, Thomas, 106–108, 111 Kykeo´n, 70 Laws, 69–70. See also Plato Lehrer, Keith, 170 Leibniz, Gottfried, 60–61 Levin, Michael, 64 Lewis, David, 13, 158–161, 165 Lewis’s quantitative parsimony. See Parsimony, Lewis’s quantitative

211

Lewontin, Richard, 149–151, 155 Libertarian free will, 21 Libri, Guilio, 87 Linguistic ethnography, 179 Local naturalism, 166 Locke, John, 36, 60–62, 150 Logical axioms, 129–130 Logical inference, 20, 25 Logically necessary propositions. See propositions, logically necessary Logically necessary truths, 155 Logical possibility, 22–23, 106 Logic of relativism, 107, 110 Lophophora williamsii, 71, 83. See also Peyote Luther, Martin, 57–58, 60 Lycan, William, 157 Lynch, Michael P., 98, 105, 110–111, 138–140 Lynch’s dilemma, 139–140 Masturbation, 63–64 Mathematical nominalists, 160 Mathematical propositions. See Propositions, mathematical Mathematics, 36, 41 McGinn, Colin, 140 McKenna, Terence, 168 Medical ethics, 63 Meiland, Jack W., 109. See also Krausz, Michael Metaethical relativism. See Relativism, metaethical Metaphysical necessity, 23, 38, 42, 147 Metaphysics, 53 Methodism, 150 Mind, 180. See also Plato Mint of Nature argument, 82–88 Miracle of Marsh Chapel of 1962, 76 Modality, 45, 100, 105 Modal logic, 106 Modal matters, 44 Modal propositions. See Propositions, modal Modal realism, 168 Modal relativism, Elgin’s, 45

212

Modal reliability, 27 Modal system, 99 Modern social science, 169 Monsay, Evelyn, 13 Moody-Adams, Michele, 179–184 Moore, G. E., 150, 155 Moorean particularism, 155 Moral beliefs, 81 Moral intuition. See Intuition, moral Moral judgments, 180 Moral propositions. See Propositions, moral Moral realism, 179 Moral theory, 39, 42 Moschion, 57 Mysterianism, 140 Mysterium tremendum, 74 Mysticism and Philosophy, 76. See also Stace, W. T. Naming and Necessity, 11–12. See also Kripke, Saul Native shamans, 79 Naturalism, 24–25, 29–33, 38, 46, 78, 83–85, 94, 123, 142, 148–152, 155–157, 166–167, 169, 176–177 Naturalistic justiﬁcation. See Justiﬁcation, naturalized Naturalistic methodism, 150 Natural law, 24 Natural reason, 56 Natural selection, 85 Natural theology, 59 Necessary propositions. See Propositions, necessary Necessary truth, 14, 20, 22, 44, 102, 104 Necessity, 20, 99, 106, 109, 156, 162 New Essays Concerning Human Understanding, 60. See also Leibniz, Gottfried Newman, Cardinal John Henry, 52, 59, 60 Newton-Smith, W., 105–106 Nichols, Shaun, 169–175, 177, 181 Niebuhr, Richard, 53, 55

Index

Nielsen, Kai, 43, 45, 79, 125–126 Nietzsche, Friedrich, 18, 45–46, 65, 68, 184 Nihilism, 50, 90, 94–95 Nominalism, 161–162 Nomological necessities, 155 Noncognitivism, 176. See also Normativism Noncontingent propositions. See Propositions, noncontingent Noninferential belief, 24, 35, 142 Nonnaturalistic belief-acquiring methods, 131, 138, 151–152, 155, 167, 169, 177 Nonnaturalistic normativity, 156 Nonnaturalistic perspectives, 151 Nonnaturalists, 155 Nonnormative psychology, 151 Nonperspectival truth, 103 Normative epistemic principles, 171. See also Epistemology; Justiﬁcation, epistemic Normative language, 157, 176 Normative principles, 173–174 Normative propositions. See Propositions, normative Normative psychology, 151 Normativism, 157–158, 175–176 ‘‘Normativity and Epistemic Intuitions,’’ 175. See also Stich, Stephen Nozick, Robert, 97 Ockham’s Razor, 156–165 Original sin, 64 Pagans, 51 Pahnke, Walter, 76–77 Paradigm shift, 108 Parsimonious theories, 164 Parsimony, 157–158, 161, 165–166 Lewis’s quantitative, 159–161 Particularism, 150 Peirce, C. S., 132 Peircean epistemic theory of truth, 132, 134, 136, 139. See also Putnamian epistemic theory of truth

Index

Perceptions, 40 Perceptual beliefs, 40 Percival, Philip, 118 Perspectives as abstracta, 114 Perspective scheme, 109. See also Conceptual scheme Peru, 73, 76 Peyote, 71, 74, 76, 78, 81, 89, 112 Philosophers of science, 157 Philosophical beliefs, 172. See also Intuition, philosophical Philosophical claims, 147, 167 Philosophical intuition. See Intuition, philosophical Philosophical knowledge, 31–32, 131, 141, 152, 177 Philosophical problems, scientiﬁc solutions to, 151 Philosophical propositions. See Propositions, philosophical Philosophical relativism. See Relativism, philosophical Philosophical truth, 128, 141 Philosophy of mind, 52, 77, 135 Physical intuition. See Intuition, physical Physical necessity, 24 Physical possibility, 23, 106–107 Physics, 41 Piety, 15–16, 19. See also Plato Plantinga, Alvin, 53–54, 66–69, 76, 81–83, 135–137, 140 Plato, 57, 69–70, 98 Platonic dialogues, 9. See also Plato Plurality problem, 80 Pope John Paul II, 55, 63, 67, 69, 88 Pope Paul IV, 63 Pope Paul VI, 133–134, 141 Positivism, 29, 37 Possibility, 99–100, 102, 106, 108– 109, 156 Possible truth, 104 Possible worlds, 104, 113 Post-Kuhnian relativism, 37 Pragmatic realism, 132. See also Internal relativism Pragmatic utility, 176

213

Premarital sex, 63, 133–134, 141 Principia Mathematica, 166. See also Russell, Bertrand Principle of mereological fusion, 160 Priority of reason, 85–88 Problem of evil, 117 Problem of induction, 28–33 Problem of intuition. See Intuition, problem of Problem of self-refutation, 44 Proof of the pudding argument, 78– 80, 86, 88 Proper names, 31 Propositions a priori, 29, 31, 33 empirical, 21, 28, 122, 124, 138 epistemic, 92, 137 logically necessary, 30 mathematical, 21, 129 modal, 31 moral, 39, 130 necessary, 14, 20, 21 noncontingent, 29 normative, 23, 124 philosophical, 20–24, 30, 36, 38, 42, 46, 50, 53, 62, 72, 77–85, 88, 89– 95, 97, 111, 119–121, 123–124, 127–131, 133–135, 138–142, 147– 148, 152, 154–155, 167, 179–180, 182, 184 self-justifying, 32 Protagoras, 57, 98. See also Plato Protestantism, 59, 62, 168 Psilocybe mexicana, 71, 83 Psilocybin, 76–77 Psychoactive drugs. See Hallucinogens Pullum, Geoffrey K., 179 Pure rationalist philosophy, 127 Pure reason, 168 Putnam, Hilary, 16, 98, 132–133 Putnamian epistemic theory of truth, 133–140. See also Peircean epistemic theory of truth Quasirealism, 176 Quine, W. V. O., 28, 92, 114, 151– 152, 154–158, 160, 162–163, 166

214

Quinean ‘‘naturalized epistemology,’’ 24 Quinean parsimony, 162, 176 Quine’s naturalism, 152, 158 Quine’s Razor, 162 Radical empiricism, 28–33 Ramsey, William, 12 Rational intuition. See Intuition, rational Rationalism, 29–30, 33–34, 38, 41, 43, 46, 78–79, 82, 124, 127, 167, 179 Rationalist philosophers, 123–124, 133–134, 141, 147–148, 150, 154, 172, 176 Rational reﬂection, 85, 87 Rawls, John, 40–43, 126 Realism, 152–154, 176 Reﬂective equilibrium, 41–42, 45, 50, 79, 113, 130, 131, 134, 147, 182, 184 Refutability, 157 Reid, Thomas, 82–85, 150 Relative possibility, 108 Relativism, 50, 90, 96 of content, 106 empirical, 119, 124 global, 99, 115, 118 internal, 132 logic of, 104, 109, 115, 129 metaethical, 180 moral, 179 philosophical, 124, 166, 180, 183 scientiﬁc, 108 speciﬁc local, 118 of truth, 106, 126 Reliabilism, 30 Republic, 9. See also Plato Rethinking Intuition, 12. See also DePaul, Michael R.; Ramsey, William Revelation, 52–69, 72, 79, 83, 85, 89– 90, 131, 133, 152, 168, 179, 183– 184 Christian (see Christianity, Christian revelation)

Index

Riemannian geometry, 38 Rig-Veda, the, 70 Ritual use of hallucinogens. See Hallucinogens, ritual use of Roman Catholic Church. See Catholic Church, the Roman Catholicism. See Catholicism Roman Curia’s document on sexual ethics, 63 Rorty, Richard, 176 Rules for the Direction of the Mind, 14, 36. See also Descartes, Rene´ Russell, Bertrand, 17, 37, 61, 133– 134, 141, 157, 163, 166 Russell’s Paradox, 40, 160 Rye ergot, 76 Salvation, 52 Santayana, George, 165 Sapir-Whorf conceptual realism, 179 Sapir-Whorf hypothesis, 178 Sartre, Jean Paul, 25 Schmitt, Frederick, 132, 137, 140 Schultes, Richard Evans, 74 Science, 9, 23–24, 28, 37–39, 149, 157, 166, 167–168, 171–172 Scientiﬁc discovery, 24, 39, 79, 95 Scientiﬁc knowledge, 154 Scientiﬁc method, 24, 138, 148, 154– 156, 177 Scientiﬁc perspective, 107 Scientiﬁc relativism. See Relativism, scientiﬁc Scientiﬁc story, 157 Secondary qualities, 22 Self-justifying propositions. See Propositions, self-justifying Self-refutation, 102, 105 Self-refutation argument, 99 Self-refutation objection, 98, 130 Self-refutation problem, 103, 128–129 Sense perception, 82–83, 87 Sensus divinitatis, 51, 59 Separate spheres doctrine, 56 Sexual ethics, 6 S5 relativist logic, 115–116 S5 system of modal logic, 156

Index

S4 relativist logic, 115 S4 system of modal logic, 156 Shamanistic societies, 123 Shamans, 71–72, 76–77, 81–82, 124, 179 Shogenji, Tomoji, 115 Sider, Ted, 38 Sidereus nuncius, 87. See also Galileo Situational meanings, 180–182. See also Duncker, Karl Skeptical possibilities, 154 Skeptical problem, 153 Skepticism, 44–46, 50, 88, 90–94, 119–121, 127–129, 150 Snow, 178–179 Social practice, 18 Social-scientiﬁc methodology, 176 Social-scientiﬁc norms, 176–177 Social-scientiﬁc reasoning, 174–175 Social scientists, 171 Socrates, 9, 15–16, 98, 148. See also Plato Socratic commonsensism, 17, 171 Socratic method, 17–18, 171 Socratic objection, 117 Socratic tradition, 147 Solipsistic idealism, 168 Soma, 70–71 Sosa, Ernest, 20, 42, 135–137, 140 Sosa–Plantinga argument, 135–137, 140 Speciﬁc local relativism. See Relativism, speciﬁc local Spinoza, Baruch, 41, 104 Spinozist, 104–105 Spirit ﬂights, 71 Stability of beliefs, 134 Stace, W. T., 76 States of affairs, 141. See also Philosophical states of affairs Stich, Stephen, 18, 40–41, 83–84, 97, 169–175, 177, 181 Students of East Asian descent, 170 Students of Western descent, 170 Substance dualism, 66, 77 Summum bonum, 19 Superstitions, 51

215

Swinburne, Richard, 51–54, 61, 66, 69, 168 Swoyer, Chris, 114 Telepathy, 51 Teonanacatl, 71 Tertullian, 56–57, 60 Theaetetus, 98. See also Plato Thompson, Frank, 157 Tidman, Paul, 82–83 Traditional rationalist philosophy, 122, 139 Transcendence problem, 161 Transcendency, 160 Tree of Knowledge of Good and Evil, 64 Truth, 98. See also Protagoras Truth, nature of, 131–140 Truth-functional truths, 25 Tukano, 73 Tukano shaman, 111, 134. See also Yaje´ Ucayali, 73 Ullian, Joseph, 157 Universality, 18 Urban legend, 178 Utilitarianism, 19, 42 act, 35, 65 hedonistic, 19 Van Cleve, James, 35 Vaupe´s, 73 Virtual absolutism, 105 Virtue theorists, 65 Vitalism, 56 Wasson, R. Gordon, 74 Web of Belief, The, 157. See also Quine, W. V. O. Weinberg, Jonathan M., 18, 166, 169–177, 181 Weinberg, Steven, 37 Western philosophers, 173–174 Wide reﬂective equilibrium, 39, 42– 43, 125–126, 131, 172. See also Reﬂective equilibrium

216

Windowless monads, 168 Wine, 69–70 Witchcraft, 71 Wittgenstein, Ludwig, 36, 162 Wittgensteinian ladder, 158 ‘‘World Spirit,’’ the, 168 Worldviews, 110, 111, 178. See also Lynch, Michael P. Wright, Crispin, 133 Xenocrates, 57 Yahweh, 51 Yaje´, 73, 79–80, 85, 111, 120

Index