C. S. Lewis’s Christian Apologetics Value Inquiry Book Series

Founding Editor

Robert Ginsberg

Executive Editor

Leonidas Donskis

VOLUME 286

Philosophy and Religion

Edited by

Kenneth A. Bryson

The titles published in this series are listed at brill.com/par C. S. Lewis’s Christian Apologetics

Pro and Con

Edited by

Gregory Bassham

LEIDEN | BOSTON Cover illustration: Morguefile.

Library of Congress Control Number: 2015943879

This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters covering Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see www.brill.com/brill-typeface. issn 0929-8436 isbn 978-90-04-30125-2 (paperback) isbn 978-90-04-30165-8 (e-book)

Copyright 2015 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

This book is printed on acid-free paper.

Philosophy and Religion (PAR)

Kenneth A. Bryson Editor

Other Titles in PAR

Alana M. Vincent. Jewish Thought, Utopia, and Revolution. 2014. VIBS 274

David C. Bellusci. Amor Dei in the Sixteenth and Seventeenth Centuries. 2013. VIBS 265

Jim Kanaris, Editor. Polyphonic Thinking and the Divine. 2013. VIBS 257

William Sweet and Hendrik Hart. Responses to the Enlightenment: An Exchange on Foundations, Faith, and Community. 2012. VIBS 241

Avi Sagi. Tradition vs Traditionalism: Contemporary Perspectives in Jewish Thought. Translated from Hebrew by Batya Stein. 2008. VIBS 197

Brendan Sweetman. The Vision of Gabriel Marcel: Epistemology, Human Person, the Transcendent. 2008. VIBS 193

Constantin V. Ponomareff and Kenneth A. Bryson. The Curve of the Sacred. 2006. VIBS 178

Deane-Peter Baker and Patrick Maxwell. Editors. Explorations in Contemporary Continental Philosophy of Religion. 2003. VIBS 143

Rem B. Edwards. What Caused the Big Bang? 2001. VIBS 115

Editorial Board of PAR

Rod Nicholls (webmaster) Harriet E. Barber Deane-Peter Baker Stephen Clark Alana M. Vincent Gwen Griffith-Dickson G. Elijah Dann Jim Kanaris Russ Dumke William Sweet Carl Kalwaitis Pawel Kawalec Ruby Ramji Esther McIntosh Gregory MacLeod Ludwig Nagl

To Alvin Plantinga, wisest and best of my official teachers

CONTENTS

EDITORIAL FOREWORD xii KENNETH A. BRYSON

PREFACE xv

INTRODUCTION: OXFORD'S BONNY APOLOGIST 1 GREGORY BASSHAM

Part One: THE ARGUMENT FROM DESIRE

ONE Pro: A Defense of C. S. Lewis’s Argument from Desire 27 PETER S. WILLIAMS

TWO Con: Quenching the Argument from Desire 45 GREGORY BASSHAM

THREE Reply to Gregory Bassham 57 PETER S. WILLIAMS

FOUR Reply to Peter Williams 69 GREGORY BASSHAM

Part Two: THE ARGUMENT FROM REASON

FIVE Pro: The Argument from Reason Defended 75 VICTOR REPPERT

SIX Con: Naturalism Undefeated 91 DAVID KYLE JOHNSON

SEVEN Reply to David Kyle Johnson 105 VICTOR REPPERT

EIGHT Reply to Victor Reppert 113 DAVID KYLE JOHNSON x Contents

Part Three: THE MORAL ARGUMENT

NINE Pro: The Moral Argument is Convincing 121 DAVID BAGGETT

TEN Con: A Critique of the Moral Argument 141 ERIK J. WIELENBERG

ELEVEN: Reply to Erik Wielenberg 153 DAVID BAGGETT

TWELVE Reply to David Baggett 163 ERIK J. WIELENBERG

Part Four: THE TRILEMMA ARGUMENT

THIRTEEN Pro: A Defense of C. S. Lewis’s “Trilemma” 171 DONALD S. WILLIAMS

FOURTEEN Con: Lewis’s Trilemma: Case Not Proven 191 ADAM BARKMAN

FIFTEEN Reply to Adam Barkman 201 DONALD S. WILLIAMS

SIXTEEN Reply to Donald Williams 205 ADAM BARKMAN

Part Five: THE PROBLEM OF EVIL

SEVENTEEN Pro: The Problem of Pain Defended 211 PHILIP TALLON

EIGHTEEN Con: C. S. Lewis on Evil: At Best a Likely Story 227 DAVID L. O’HARA

NINETEEN Reply to David L. O’Hara 237 PHILIP TALLON Contents xi

TWENTY Reply to Philip Tallon 243 DAVID L. O’HARA

WORKS CITED 249

THE CONTRIBUTORS 267

INDEX 269

EDITORIAL FOREWORD

The work of C. S. Lewis is challenging because it contains several layers of complexity. The problem of the suffering of innocent victims is a case in point. Lewis’s approach to the problem of evil faces one of the central problems of Theodicy: how can we reconcile the existence of an unlimitedly powerful and perfectly good God with the suffering of innocent victims? At first brush, Lewis’s explanation in The Problem of Pain sounds overly simplistic. Human suffering is the result of the fall from God’s grace; no need for a forced choice between divine attributes. Once we decided to rely on self- will rather than on God’s will “we fell under the control of ordinary biochemical laws and suffered whatever the inter-workings of those laws might bring about in the way of pain, senility and death” (77). But Lewis raises the argument to a deeper level when he wonders why we presume to know how God thinks. We cannot be sure why God permits evil. We are familiar with the ways we cause suffering by disempowering the environment and other persons, but this behavior is at odds with God’s grace and forgiveness of sin. In my opinion, the deepest layer of Lewis’s apologetics surfaces in his humble recognition of the consequence of not knowing the ways of the divine mind. This is clearest in A Grief Observed as he attempts to make sense of the suffering and death of his wife. Why would a good and loving God permit this? We cannot presume to identify our experience of pain with the role God provides for pain in the lives of animals and humans. In final analysis, Lewis’s grief enables him to identify with Job’s mystical experience as Job presses God to explain the suffering of the innocent (himself). Lewis’s passionate reflection on the loss of his dear wife brings him to the same place; an existential place in subjectivity where he moves beyond intellectualization to recognize his own insignificance and necessary dependence on the Almighty. I am delighted to welcome Gregory Bassham’s volume on Lewis’s Christian Apologetics to the Philosophy and Religion family of scholarly books. This book is a great catch, not only for Rodopi-Brill, but for lovers of C. S. Lewis worldwide. Kenneth A. Bryson Editor PAR special series Value Inquiry Book Series

PREFACE

C. S. Lewis is unquestionably the most influential Christian apologist over the past century. It is therefore surprising that—with a few notable exceptions— his arguments for Christian belief have received so little sustained scholarly attention. This book seeks to fill that gap. Here five leading defenders of Lewis’s apologetics go head-to-head with five acute critics in a pro/con debate format that permits in-depth discussion of the relevant issues. Our aim has been to produce a book that examines Lewis’s main arguments for Christianity with depth and rigor but is also accessible to general readers. The result, we hope, is a book that gives readers a clearer understanding of Lewis’s arguments and permits them to draw their own informed conclusions about whether those arguments are properly convincing. In editing this book, I have accumulated many debts. I am grateful to Dave Baggett, Erik Wielenberg, and Adam Barkman for advice on how to structure the book; to the contributors for their patience, professionalism, and grace under fire (friendly and otherwise); to my colleague Bernard Prusak for judicious feedback on various parts of the manuscript; to Robert Arp and Ken Bryson for much helpful advice on formatting issues; and to my wife, Mia, for her love, support, and generous spirit over a long period as this volume took shape. The book is dedicated to Alvin Plantinga, with whom I had the great good fortune to study in graduate school. Al was more than just an extraordinary scholar and a gifted teacher; he was a delightful companion, a fountain of whimsical good humor, and a model of how the Christian intellectual life should be lived. No one has contributed more to the remarkable flowering of philosophy of religion and Christian philosophy over the past half-century than Al has. My debt to him, both personal and professional, is great. Gregory Bassham King’s College

INTRODUCTION: OXFORD’S BONNY APOLOGIST

Gregory Bassham

C. S. Lewis (1898-1963) is the world’s best-known and most-admired Christian apologist. More than fifty years after his death, his books continue to sell in the millions, and countless readers around the globe have been influenced and inspired by his writings. In a 2004 Christianity Today survey, Lewis’s Mere Christianity (1952) was named the best religious book of the twentieth century. Yet, curiously, Lewis is not generally considered a major figure by academic theologians (MacSwain & Ward, 2010, pp. 1-2). There are very few books that discuss his religious writings with the scholarly depth and rigor they deserve (but cf. Purtill, 2004; Beversluis, 2007; Williams, 2013; Baggett, Habermas, and Walls, 2008). This book attempts to do just that. Lewis himself was not a professional theologian or philosopher, but a highly regarded professor of English at Oxford and Cambridge, specializing in medieval and Renaissance literature. After being an atheist for many years, Lewis converted to Christianity in his early thirties. Over a period of approximately two decades, Lewis published a series of widely-read works of Christian apologetics, including The Pilgrim’s Regress (1933), The Problem of Pain (1940), The Abolition of Man (1943), Miracles (1947), and Mere Christianity (1952). As a testament to Lewis’s growing popularity as a Christian writer, Time magazine featured him on the cover of its September 8, 1947 issue. Since his death in 1963, Lewis’s reputation as a Christian author and apologist has continued to grow. What accounts for Lewis’s success as an apologist? First and foremost, I believe, clarity and logic. Lewis was a master of clear, forceful, chiseled prose. Moreover, like Franklin Roosevelt in his famous fireside chats, Lewis was able to communicate difficult ideas in simple, direct language that people of all backgrounds could understand. As a communicator, Lewis was arguably one of the giants of the twentieth century. Lewis was also a master of logical reasoning and intellectual combat. For more than a decade, he served as president of the Oxford Socratic Club, a student organization dedicated to open discussion of the arguments for and against Christianity. At nearly every meeting, Lewis was there, “sniffing the imminent battle and saying ‘Aha!’ at the sound of the trumpet” (Farrer, 1992, p. 142). Despite regularly locking horns at these meetings with such secular heavyweights as A. J. Ayer, Antony Flew, Bernard Williams, and Gilbert Ryle, Lewis would “always maintain the cause,” and “no one could put Lewis down” (Farrer, 1992, p. 142). 2 GREGORY BASSHAM

Despite Lewis’s immense popularity as an apologist, one must still ask: Are his apologetical arguments sound? Can they stand up to rigorous intellectual scrutiny? That is the central focus of this book. Lewis defended Christian belief on many fronts, and it would be impossible to consider all his apologetical arguments here. To keep things manageable, we have selected what we consider to be Lewis’s five most important apologetical arguments: the argument from desire, the argument from reason, the moral argument, the trilemma argument, and Lewis’s response to the problem of evil. The first four arguments are examples of positive apologetics: the attempt to provide positive rational justification for religious belief. The last argument—Lewis’s answer to the problem of evil— is a piece of negative apologetics: the attempt to defend religious belief by responding to objections. So what are these five arguments, and are they rationally convincing? Let us begin with the argument from desire.

1. The Argument from Desire

Lewis’s argument from desire (or argument from joy, as some call it) is a modified version of a stock medieval and Renaissance argument for life after death. Versions of the argument are offered by Aquinas (1945, p. 692), Ficino (1948, pp. 211-212), Hooker (1845, p. 201), the authors of The Spectator (1907, p. 126), and many others. There are two basic forms of the argument, one philosophical and the other broadly theological. The first appeals to the Aristotelian dictum that “nature does nothing in vain.” Aquinas offers a typical statement of this philosophical version of the argument:

[I]t is impossible for natural desire to be unfulfilled, since “nature does nothing in vain.” Now, natural desire would be in vain if it could never be fulfilled. Therefore, man’s natural desire is capable of fulfillment, but not in this life . . . . So, it must be fulfilled after this life. Therefore, man’s ultimate felicity comes after this life (Aquinas, 1956a, p. 166).

The second version of the traditional argument from desire appeals not to the supposedly naturally knowable notion that all natural desires must be satisfiable, but to the confidently posited goodness and faithfulness of God as revealed in both Scripture and human experience. The Cambridge Platonist John Smith (1618-1652) offers an eloquent version of this form of the argument in his Selected Discourses (1660):

[The soul] knows that God will never forsake his own life which he hath quickened in it; he will never deny those ardent desires of a blissful fruition of himself, which the lively sense of his own Goodness hath Introduction 3

excited within it; those breathings and gaspings after an eternal participation of him are but the Energy of his own breath within us; if he had any mind to destroy it, he would never have shown it such things as he hath done; he would not raise it up to such Mounts of Vision, to shew it all the glory of that heavenly Canaan flowing with eternal and unbounded pleasures; and then tumble it down again into that deep and darkest Abyss of Death and Non-entity. Divine goodness cannot, it will not, be so cruel to holy souls that are ambitious suitors for his love (quoted in Willey, 1970, p. 179).

Lewis’s argument from desire is different from both of these traditional arguments. Instead of focusing on an enduring natural desire for eternal happiness, Lewis appeals to a specific type of ardent and fleeting spiritual longing, which he calls “Sweet Desire,” Sehnsucht, or “Joy.” Lewis argues that Joy is evidence of a transcendent reality. So what exactly is Lewis’s argument from desire? It is not easy to be precise, because Lewis nowhere states the argument in detail, and he sometimes formulates it in different ways. Slightly different versions of the argument appear in Mere Christianity (1952), the Preface to the third edition of The Pilgrim’s Regress (1933; 3rd ed., 1943), Surprised by Joy (1955), and “The Weight of Glory” (1940). There are also a few passages in the Chronicles of Narnia (1950-56), Till We Have Faces (1956), his early poems (notably “Sweet Desire” and “Joy”), and other literary works which speak, directly or indirectly, to various aspects of the argument. Finally, there is an early, speculative, and significantly different version of the argument (or perhaps a cognate argument) in a 1931 letter to his brother Warnie (Lewis, 2004, pp. 7-8). Let us begin with what Lewis means by “Joy” (hereafter Joy, with a capital “J,” but without quotation marks or italics). Lewis uses Joy in a specialized sense to refer to a particular form of desire, longing, or emotional response that he assumes will be familiar to at least most of his readers. Joy is a form of desire, Lewis claims, but of a unique sort. Experiences of Joy are brief, intense, thrilling “pangs” or “stabs” of longing that are at once both intensely desirable and achingly painful (Lewis, 1956, pp. 16-17, 72, 166). Physically, Joy is experienced as “a kind of kick or flutter in the diaphragm” which, for Lewis, felt exactly the same as the sensation that “accompanies great and sudden anguish” (Lewis, 1949, p. 20). Though Joy is a form of desire, it differs from all other desires in two respects. First, whereas other desires “are felt as pleasures only if satisfaction is expected in the near future,” with Joy “the mere wanting is felt to be somehow a delight” (Lewis, 1977, p. 7). Joy thus “cuts across our ordinary distinctions between wanting and having. To have it is, by definition, a want: to want it, we find, is to have it” (Lewis, 1977, p. 8). Second, Joy differs from all other desires in the mysteriousness or elusiveness of its object(s). With Joy, it is not clear exactly what is desired, 4 GREGORY BASSHAM and false leads are common. Many suppose, mistakenly, that Joy is a desire for some particular worldly satisfaction (sex, aesthetic experience, and so forth). But all such satisfactions, Lewis argues, turn out to be “false Florimels,” delusive images of wax that melt before our eyes and invariably fail to provide the satisfaction they seem to promise (Lewis, 1977, p. 8). It is this second unique feature of Joy—the fact that it is a strangely indefinite desire that apparently cannot be satisfied by any natural happiness attainable in this world—that provides the linchpin for Lewis’s argument from desire. So what, more precisely, is Lewis’s argument from desire? As John Beversluis notes, Lewis seems to offer both deductive and inductive versions of the argument (Beversluis, 2007, p. 40). The deductive version can be formulated as follows:

(1) Every natural desire has an object that can satisfy it. (2) Joy is a natural desire. (3) So, Joy has an object that can satisfy it. (4) But nothing in the natural world can satisfy Joy. (5) So, Joy has a supernatural object that can satisfy it. (6) The only supernatural object that can satisfy Joy is God. (7) Therefore, God exists.

Some writers have denied that Lewis gives a deductive version of the argument from desire (Holyer, 1988, pp. 68-69; McGrath, 2014, p. 118.) In fact, some Lewis scholars deny that he meant to offer a formal “argument” from desire at all (Smilde, 2014). But Lewis certainly appears to be offering an argument; he seems to make claims defended with reasons. Moreover, Lewis often uses modest, tentative-sounding language that strongly suggests that he meant the argument from desire to be understood as inductive (that is, probabilistic). Here are two representative passages, with the pertinent wording italicized:

Creatures are not born with desires unless satisfaction for these desires exists. A baby feels hunger: well, there is such a thing as food. A duckling wants to swim: well, there is such a thing as water. Men feel sexual desire: well, there is such a thing as sex. If I find in myself a desire which no experience in this world can satisfy, the most probable explanation is that I was made for another world (Lewis, 2001a, pp. 136- 137).

[W]e remain conscious of a desire which no natural happiness will satisfy. But is there any reason to suppose that reality offers any satisfaction of it? . . . A man’s physical hunger does not prove that that man will get any bread; he may die on a raft in the Atlantic. But surely a man’s hunger does prove that he comes of a race which repairs its body Introduction 5

by eating and inhabits a world where eatable substances exist. In the same way, though I do not believe . . . that my desire for Paradise proves that I shall enjoy it, I think it a pretty good indication that such a thing exists and that some men will (Lewis, 1949, p. 6).

If Lewis does indeed mean to offer an inductive version of the argument from desire, the argument might be briefly formulated as follows:

(1) Humans have a natural desire for the transcendent. (2) Most natural desires have objects that are capable of satisfying them. (3) So, there probably is something transcendent.

Is this version of the argument from desire, or any version, properly convincing? Peter S. Williams, author of C. S. Lewis vs. the New Atheists (2013), believes so. In fact, Williams, in his contribution to this volume, argues that there are no less than five compelling versions of the argument from desire. These include not only the deductive and inductive versions discussed above, but also what he calls the a priori argument, the abductive argument, and the reductio argument. In my responses to Williams in this volume, I argue that each of these five versions of the argument is faulty. Among the issues I raise are: How clear is it that Joy is a form of desire? If Joy is a desire, is it, in the relevant sense, a natural desire? Is it true that all (or even most) natural desires have objects that can satisfy them? Even if it is true that most natural desires have objects that can satisfy them, are there special reasons for doubting whether Joy can be satisfied?

2. The Argument from Reason

The argument from reason is an argument against metaphysical naturalism— the view, roughly, that nature or physical reality is all that exists (Goetz and Taliaferro, 2008, p. 67). Lewis did not invent the argument from reason; in fact, he refers to it as a “venerable philosophical chestnut” (Lewis, 2004, p. 715). However, Lewis is by far the best-known and most influential defender of the argument. Lewis lays out the argument most fully in the first and second editions of Miracles (1947, 1960), but there are also notable discussions in “De Futilitate” (1967), “Is Theology Poetry?” (1965), “Bulverism” (1970), “On Living in an Atomic Age” (1987), and in several of Lewis’s letters (see, for example, Lewis, 2004, pp. 714-15; Lewis, 2007, pp. 1338-1339, 1351). Let us first get clear on what the argument is, and then examine what can be said for and against it.

6 GREGORY BASSHAM

A. Setting the Stage

The aim of Lewis’s argument from reason is to show that naturalism is both false and (in a certain sense) self-refuting. It is false because it implies something that is obviously false—namely, that reasoning does not exist. It is self-refuting because naturalism itself can only be rationally believed and defended if reasoning is possible. Many thinkers have offered similar arguments against naturalism (also called “materialism” or “physicalism,” though some would distinguish these views). Aquinas, for example, argues that naturalism must be false because it cannot explain mental phenomena such as the mind’s ability to grasp general concepts, to apprehend infinite things, to know things other than bodies, and to reflect upon itself (Aquinas, 1956b, pp. 146-49). Kant argues that we must “postulate,” though we cannot know, that there is a nonphysical supersensible reality to make sense of our notions of moral duty and moral worth (Kant, 1993, pp. 49-61). Versions of the claim that naturalism is self-refuting have been defended by Arthur Balfour, G. K. Chesterton, J. B. S. Haldane, C. E. M. Joad, Victor Reppert, William Hasker, and Alvin Plantinga (Balfour, 1879, chap. 13; Chesterton, 1959, p. 33; Haldane, 1929, p. 209; Joad, 1933, pp. 58-59; Hasker, 1999, pp. 67-71; Reppert, 2003; Plantinga, 2011, chap. 10). There is nothing especially original about Lewis’s argument; much of it is borrowed Balfour. But Lewis’s version is justly famous for its clarity and force. Lewis defines his target, naturalism, as “the doctrine that only Nature— the whole interlocked system—exists” (Lewis, 2001b, p. 18). Later, it becomes clear that Lewis sees naturalism as involving three central claims: (1) that nothing supernatural exists (no God, no souls, no heaven, for example); (2) that everything that exists is either some form of physical “stuff” or is fully explainable in terms of physical stuff; and (3) that the whole of reality is “interlocked” in the sense that everything that occurs in nature is part of a closed, deterministic causal network. This last claim—that naturalism implies determinism—is debatable. As Lewis himself notes, many scientists now believe that quantum physics has shown that certain sub-atomic events are not causally determined. Lewis is dubious of this view, suggesting that scientists don’t really believe this, or if they do that science will someday show that they were wrong (Lewis, 2001b, pp. 18-19). Lewis can be challenged here. Though Einstein was also skeptical of quantum indeterminism—quipping that “God does not play dice with the universe”—it is strongly supported by experimental evidence, and most physicists now accept it. As we shall see, however, nothing in Lewis’s argument requires a linkage between naturalism and determinism. Moreover, by defining naturalism so narrowly, Lewis significantly restricts the scope of his argument, leaving unscathed widely-held forms of naturalism that do not Introduction 7 presuppose determinism. For these reasons, many defenders of Lewis’s argument drop his assumption that naturalism implies determinism.

B. Lewis’s Argument in the First Edition of Miracles

Lewis offered the first sustained presentation of the argument from reason in his 1947 book, Miracles: A Preliminary Study. (Lewis, it should be noted, never uses the term “argument from reason”; it was coined by John Beversluis in his 1985 book, C. S. Lewis and the Search for Rational Religion.) The following year, Lewis’s argument was powerfully criticized by Elizabeth Anscombe in a famous exchange with Lewis in a meeting of the Oxford Socratic Club. Lewis accepted many of Anscombe’s criticisms, and a longer and significantly improved version of the argument was published in the second edition of Miracles (1960). Lewis’s argument in the first edition of Miracles can be summarized as follows (cf. Reppert, 2009, pp. 353-54):

(1) No inferred belief is valid if it can be fully explained as the result of irrational causes. (2) If naturalism is true, then all inferred beliefs can be fully explained as the result of irrational causes. (3) So, if naturalism is true, then no inferred belief is valid (from 1 and 2). (4) Naturalism can be rationally believed only if it is a valid inferred belief. (5) So, naturalism cannot be rationally believed (from 3 and 4). (6) There are many valid inferred beliefs. (7) Any view that implies that no inferred beliefs are valid is false. (8) So, naturalism is false (from 6 and 7). (9) Therefore, naturalism is false and cannot be rationally believed (from 5 and 8).

In her response to Lewis, Anscombe offered a number of astute objections. The three most important concern (1) problems with Lewis’s use of the term “valid,” (2) Lewis’s failure to distinguish irrational from nonrational causes of belief, and (3) Lewis’s failure to distinguish two important senses of the word “because.” Anscombe was understandably puzzled by Lewis’s talk of “valid” and “invalid” thoughts. When Lewis studied logic in the 1920’s, it was common to speak of valid beliefs and inferences, as well as valid arguments (Bradley, 1883, p. 521; Balfour, 1915, p. 139; Russell, 1927, p. 185; Russell, 1948, p. 182). The restriction of “valid” to arguments became firmly established in logic only with the publication of Irving Copi’s enormously popular textbook, 8 GREGORY BASSHAM

Introduction to Logic, in 1956. It was also common in Lewis’s day for logicians to distinguish various senses in which inferences or arguments could be valid. Bradley, for instance, distinguishes between inferences that are “formally” valid (where the premises logically imply the conclusion) and those that are “really” valid (where the inference reliably tracks or copies reality) (Bradley, 1883, p. 521). Unless inferences are valid in this substantive, realist sense, Bradley argued, human reason threatens to self- destruct. He writes:

If in inference the conclusion is made what it is by an arbitrary act, how can any such process be true of reality? Our knowledge of the cause will itself be dragged down in the common ruin of all our reasoning, and in the end we must doubt whether there is such a thing as valid inference (Bradley, 1883, p. 497).

Anscombe, who studied under Wittgenstein and was trained in modern logic, faulted Lewis for not spelling out clearly what he meant by “valid.” If he meant “an argument in which the conclusion follows necessarily from the premises,” then some arguments are indisputably valid and some are not. If, on the other hand, he meant something normative, like an epistemically “rational” or “justified” inference, then naturalists, like anyone else, can recognize that such inferences exist (Anscombe, 1981, pp. 226-27). To confuse the genesis of an argument—its causal origins—with its rational strength would be to commit the genetic fallacy. As Anscombe saw it, whether a piece of reasoning is “rational” or “justified” usually has little or nothing to do with how the reasoning was caused. “If a man has reasons, and they are good reasons . . . for thinking something—then his thought is rational, whatever causal statements we make about him” (Anscombe, 1981, p. 229). Moreover, Lewis seemed to be suggesting that if naturalism is true, then no inferred belief is valid. But what sense does it make to contrast valid and invalid reasoning if no reasoning can be valid? If a bit of reasoning can go wrong, must it not be possible for it also to go right? In short, Anscombe charged, Lewis fails to identify any clear sense of “valid” that would pose a special problem for naturalists. In a brief response to Anscombe published in the Socratic Digest, Lewis conceded that “valid” was “a bad word” for the positive epistemic quality he had in mind—something like “veridical” would have been better (Anscombe, 1981, p. 231). Later, as we shall see, Lewis made clear that what he was talking about was neither deductive validity nor (pace Reppert 2003, 2012) epistemic rationality or justification, but instead what Bradley had called “real validity”: the value of reasoning “as a means of finding truth” (Lewis, 2001b, p. 23). Lewis’s aim was to show that naturalism implies an epistemic disconnect with reality that makes genuine reasoning (and hence most forms of knowledge) impossible. Introduction 9

Anscombe’s second major criticism of Lewis was that he failed to note an important ambiguity in the term “irrational.” Lewis, following Balfour (1915, p. 140), claimed that for naturalists “[a]ll thoughts whatever are . . . the results of irrational causes” (Lewis, 1947, p. 28). Anscombe pointed out that “irrational” can mean either “contrary to reason” (for example, inconsistent or wishful thinking) or nonrational (for example, an electrical impulse in the brain). Naturalists, she noted, may be committed to believing that all thoughts have nonrational causes, but they need not (and should not) admit that all thoughts have irrational causes (Anscombe, 1981, 224-25). Lewis accepted this criticism, noting that Anscombe had shown that his argument at best “proved the truism that irrational thought is invalid, not that all thought (being based on the non-rational) is invalid” (Lewis, 2007, p. 1338). He was careful in later presentations of the argument to distinguish these two senses of irrational, and to note that naturalists were committed only to the claim that all inferred beliefs could be fully explained in terms of nonrational causes. Anscombe’s final major criticism was that Lewis failed to note an important distinction between two senses of “because.” Consider the two statements:

(1) Alicia saw spots because she bumped her head. (2) Eating meat is wrong, because it causes unnecessary suffering.

As Anscombe notes, the “because” in (1) indicates a cause-effect relationship, whereas the “because” in (2) refers to a relationship of evidential support, or what she calls a “ground” (Anscombe, 1981, p. 225). Lewis, she claims, confuses these two senses of “because,” and consequently fails to see that the causes that produced a particular bit of reasoning are irrelevant to its logical strength or cognitive value (Anscombe, 1981, pp. 227-28). Lewis’s reaction to this last criticism was mixed. He agreed that it was important to distinguish cause-and-effect connections from what he came to call “ground-consequent” relationships. In fact, this distinction plays a critical role in his revised argument from reason in the second edition of Miracles. However, Lewis did not agree with Anscombe that questions of causal origin are irrelevant to the evidential or cognitive value of acts of reasoning. On the contrary, as we shall see, Lewis argued that naturalists are committed to an account of thinking that undermines the “truth-tracking” value of reasoning.

C. Lewis’s Revised Argument

In the second edition of Miracles, Lewis offered a significantly expanded and improved version of the argument from reason. The argument can be stated as follows: 10 GREGORY BASSHAM

(1) Naturalism can be rationally believed or defended only if it can be rationally inferred from good evidence. (2) Reasoning requires insight into real, objective relations of logical support. (3) Naturalism implies that all acts of inference can be fully explained in terms of nonrational causes. (4) If an act of inference can be fully explained in terms of nonrational causes, then it does not involve any insight into real, objective relations of logical support. (5) So, if naturalism is true, then reasoning does not exist (from 2-4). (6) If reasoning does not exist, then naturalism cannot be rationally inferred from good evidence. (7) So, if naturalism is true, then it cannot be rationally inferred from good evidence (from 5-6). (8) So, if naturalism is true, then it cannot be rationally believed or defended (from 1 and 7). (9) Any theory that implies that reasoning does not exist is false. (10) Therefore, naturalism is false, and even if were true, it could not be rationally believed or defended (from 5 and 9, and 8)

A quick aside here: This is a reconstruction of what I take to be the central thread of Lewis’s argument. The argument is complex, and there are elements I have left out. These include an argument that evolution could never explain how human reason emerged (Lewis, 2001b, pp. 33-34; cf. Nagel, 2012, pp. 45-47); a claim that naturalism may not be able to explain the “intentionality” of human thought—the way that thoughts are “about” things and have a capacity to represent things as being a certain way (Lewis, 2001b, pp. 25-26); and a lengthy argument, in Chapter 4 of Miracles, that human reason must have a “taproot” in an eternal, self-existent, rational Being (not necessarily God, but some sort of Cosmic or super-Cosmic Mind). These are all complicated (and contentious) issues that I cannot address here. Back, then, to the central argument. Clearly, the crucial premises in the argument as I have formulated it are 2-4. Before we look at how Lewis defends them, let me say a word about the overall structure of Lewis’s argument. Lewis’s argument from reason is a version of what philosophers call a “transcendental argument.” A transcendental argument is a type of argument that seeks to show that one thing, X, is a necessary condition for another thing, Y; and that since Y is the case, X logically must be the case as well (Robb and Heil, 2013). Well-known examples of transcendental arguments in philosophy include Kant’s attempt to show that there must be built-in categories of the human understanding to explain how we know synthetic a priori truths (Kant, 1961, pp. 129-175), and Descartes’s attempt to show that a Introduction 11 perfect God must exist to explain our idea of such a Being (Descartes, 1958, pp. 195-211). Metaphysical worldviews are natural targets for transcendental arguments, because it is easy to formulate big, overarching theories of reality that do not match up with the world as we concretely experience it. This was the crux of Kiekegaard’s critique of Hegel’s absolute idealism (Kierkegaard, 1968, pp. 98-113). An even better example is provided by Parmenides’s suggestion that the whole of reality consists of a timeless, changeless, undifferentiated sphere (Guthrie, 1965, pp. 43-49). What Philosophy 101 student could not have a field day thinking up transcendental arguments against that? Lewis’s argument is a transcendental argument aimed at metaphysical naturalism. But it is also what some philosophers call a “recoil argument” (Blackburn, 2005, p. 27). These are arguments that attempt to show that some claim is self-refuting in some respect. There are a variety of ways in which claims can be self-refuting. One familiar way is when a claim is either directly self-contradictory (“It can be proved that nothing can be proved”) or logically implies its contradictory (“All propositions are false”). In the first edition of Miracles, Lewis apparently believed that naturalism is self-refuting in this sense, since he titled the chapter in which he presented the argument from reason “The Self-Contradiction of the Naturalist” (Lewis, 1947, p. 23). Later, Lewis seems to have realized that his argument does not show that naturalism is self-contradictory or that it implies a contradiction. Rather, what it shows is that naturalism is epistemically self-undercutting in the sense that, if it is true, it cannot be rationally believed or defended. Recoil arguments of this sort are fairly common in philosophy. One much-discussed recent example is Alvin Plantinga’s attempt to show that evolutionary naturalism is epistemically self-undermining, because if it is true, then the reliability of our cognitive faculties is probably low, and thus its defenders have no good reason to accept evolutionary naturalism or pretty much anything else (Plantinga, 2011, pp. 307-349). Another notable example is Hume’s critique of Descartes’s hypothesis of an “evil genius.” Descartes famously hypothesized that he might be a victim of a powerful deceptive being who not only flooded his mind with as many false beliefs as possible, but warped even his very powers of thinking and reasoning, so that Descartes could not be sure even of seemingly self-evident things like the proposition that 3+2=5 (Descartes, 1958, pp. 179-181). Hume correctly pointed out that a skepticism this deep is a kind of black hole from which no escape is possible. Descartes’s subsequent attempt to refute the evil genius hypothesis by means of complex “proofs” of the existence of an all-powerful, truth-loving God presupposes the reliability of the very faculties (such as reason and memory) that the evil genius hypothesis calls into question (Hume, 1977, p. 103). So, Hume argued, Descartes shoots himself in the foot. The conjunction of theism with initial radical skepticism is epistemically self-undermining, because if 12 GREGORY BASSHAM

God does exists, His existence cannot be rationally believed given Descartes’s method of doubt. In sum, then: Lewis’s argument from reason is both a transcendental argument that aims to show that naturalism is false, and a recoil argument that seeks to show that it is irrational to accept naturalism, even if should turn out that naturalism appears to be true. Now let us look at how Lewis defends the key premises in his argument, beginning with (2), the claim that reasoning requires insight into real, objective relations of logical support. Reasoning is the mental process of drawing conclusions or inferences from beliefs, assumptions, hypotheses, or other data (cf. Paul, 1993, p. 485). Reasoning is often equated with inference, but they may not completely overlap. For one thing, reasoning seems to be a form of explicit inferring, whereas inference can rest on wholly implicit assumptions and be subconscious (James, 1950, p. 326; Paul, 1993, p. 485). Moreover, reasoning is nearly always conceived of as a mental process, whereas inference can refer either to a process of reasoning (“the inferential process”) or to the completed set of propositions that were involved in the mental sequence (“the inferential content”) (Audi, 2011, p. 177). Lewis, at any rate, seems to use the terms “reasoning” and “inference” pretty much interchangeably, and to think of these as mental processes. Most of his examples suggest that he is thinking mainly of deductive reasoning, but his (exaggerated) claim that “all possible knowledge” (Lewis, 2001b, p. 21) depends on the reliability of reasoning means that he must include nondeductive forms of inference as well. So why does he think that all forms of reasoning, whether demonstrative or probabilistic, require an insight into real, objective principles of inference? Lewis notes that all forms of reasoning involve a presumed “logical relation between beliefs or assertions” (Lewis, 2001b, p. 23). In reasoning, we recognize—implicitly or explicitly—that some things we believe or know are reasons for drawing certain conclusions. Reasoning thus presupposes what Lewis calls a “ground-consequent relation” (Lewis, 2001b, p. 23). Some things we believe (the ground) are seen as entailing, or providing evidence for, a conclusion (the consequent). What properly connects the ground with the consequent are principles of inference. There are a wide variety of such principles. The most basic are the so-called “laws of thought,” such as the principle of noncontradiction (no statement can be both true and false) and the principle of identity (if A then A). These principles are tautologies (in other words, logical entailment relationships that are necessarily true in virtue of their truth-functional form) and lie at the foundation of all logical or coherent thought. Other principles of inference include familiar rules of truth- functional reasoning (modus ponens, modus tollens, disjunctive syllogism, and the like). The corresponding conditionals of these argument forms are also tautologies. But there are many principles of inference that are not true merely in virtue of their logical form. These include hard-to-formalize principles of Introduction 13 informal logic, such as how to recognize good analogical reasoning, make good inferences to the best explanation, and avoid fallacious appeals to authority. They also may include normative principles of the sort discussed in, say, Descartes’s Discourse on Method, Locke’s Essay Concerning Human Understanding, and in contemporary discussions of critical-thinking standards and intellectual virtues. Epistemic norms such as “Avoid wishful thinking,” “Don’t stereotype,” “Be open-minded,” “Make sure all your premises are relevant,” “Check your conclusions carefully,” and “Don’t accept extraordinary claims without extraordinary evidence” can also figure in, and properly guide, acts of reasoning. They are not principles of inference in the same sense as, say, the rule of double negation (the rule that A implies not- not-A, and vice-versa). They are normative principles, not “laws of reasoning” that license or entail particular inferences from premises to conclusions. Moreover, unlike tautological entailment relationships, it is unclear whether they are knowable a priori. Nevertheless, they do clearly play an important role in intelligent and successful reasoning. They are part of network of formal and informal action-guiding epistemic principles that some philosophers call “norms” (Blackburn, 2005, p. 29). All reasoning, therefore, presupposes principles of inference. But what are principles of inference? What is their ontological status? Are they simply feelings, or instincts, or psychological dispositions, or linguistic stipulations, or social conventions, or causally determined neurochemical events in the brain? If so, Lewis argues, then we have no reason to think that reasoning reliably leads to truth. As Lewis puts it:

If the feeling of certainty which we express by words like must be and therefore and since is a real perception of how things outside our own minds really “must” be, well and good. But if this certainty is merely a feeling in our own minds and not a genuine insight into realities beyond them—if it merely represents the way our minds happen to work—then we can have no knowledge. . . . It follows that no account of the universe can be true unless that account leaves it possible for our thinking to be a real insight (Lewis, 2001b, p. 21).

Put otherwise: Reasoning presupposes the reliability of reason itself and the objective, truth-tracking validity of logic. If logic does not “connect” us to reality via a genuine insight, then “it has no value as a means of finding truth” (Lewis, 2001b, p. 23). To drive home Lewis’s point, imagine a world in which something like Descartes’s evil genius really does exist. Suppose you are kidnapped by aliens, taken to their home world, and plugged into a virtual reality machine called the Scrambler. This machine not only creates totally realistic virtual- reality hallucinations, but also scrambles your powers of logical reasoning. Every time you think that some conclusion follows from, say, modus tollens, 14 GREGORY BASSHAM it causes you to reason fallaciously and redirects your thoughts to some false conclusion. (If you think that you would quickly discover some “glitch” in this virtual world and wise up, imagine that the machine modifies your memories as needed to maintain the illusion.) Pretty much everything you believe in this world is false, including your belief that logic “works.” You might, on some accounts, be epistemically “justified” in your beliefs. You might be trying your hardest and not be at fault in any way. Yet clearly you have (practically) no knowledge. Your “reasoning” is disconnected from reality and provides no means to truth. Reasoning thus presupposes the objective validity, normativity, and “veridicalness” of ground-consequent relations. Logic cannot be simply a feeling, a subjective preference, or a social or linguistic convention. Logic must be a reliable pointer to reality. It must connect us with truth. And this means that when we reason we must “see” (intuit, apprehend) logical connections that exist, as Lewis says, “outside our own minds” (Lewis, 2001b, p. 23). (Contra Plato and Descartes, Lewis assumes that inferential principles are not innate. Moreover, treating them as innate would seem to fit poorly with naturalism.) It is not enough if we merely think we see objectively valid principles of inference, as the victim of the Scrambler did. We must really see them. There must be what Lewis calls “insight.” Lewis next asks: What must be true for such acts of insight to occur? Here is where Anscombe’s distinction between “causes” and “reasons” comes into play. Acts of inference are, of course, caused (all events are caused, Lewis believes). But causes, he argues, can be of two sorts. They can be ordinary causes in nature’s “interlocked” system of cause and effect (Lewis, 1970, p. 275). Or they can be rational causes in somebody’s mind. Reasoning presupposes that logical inferences have rational causes—we draw a particular conclusion because we see that the conclusion follows from the premises. This does not mean that acts of inference cannot also have physical causes, or at least “conditions” (Lewis, 1970, p. 275). My inference that, because I see smoke, there must be fire, has many physical causes (the fact that my optic nerve is working properly, for example). But Lewis believes that for trustworthy reasoning to occur, there must be a sense in which rational causes are not merely contributory or even predominant, but fully determinative. As Lewis writes:

A belief which can be accounted for entirely in terms of causes is worthless. The principle must not be abandoned when we consider the beliefs which are the basis of others. Our knowledge depends on our certainty about axioms and inferences. If these are the result of causes, there is no possibility of knowledge. Either we can know nothing or thought has reasons only, and no causes. . . . [T]hought has no father but thought (Lewis, 1970, p. 275). Introduction 15

Elsewhere, he says this about the respective roles of reasons and causes in acts of cognition:

An act of knowing must be determined, in a sense, solely by what is known; we must know it to be thus solely because it is thus. . . . The act of knowing has no doubt various conditions, without which it could not occur: attention, and the states of will and health which this presupposes. But its positive character must be determined by the truth it knows. If it were totally explicable from other sources, it would cease to be knowledge. . . . Any thing which professes to explain our reasoning fully without introducing an act of knowing thus solely determined by what is known, is really a theory that there is no reasoning (Lewis, 2001b, pp. 26- 27).

In a nutshell: Reasoning requires an act of cognitive insight into real, objective, logical relations that exist outside the mind, and such insight is possible only if acts of inferential knowing have a determinative rational cause. So much for Lewis’s defense of premise (2). What of premise (3), the claim that naturalism implies that all acts of inference can be fully explained in terms of nonrational causes? Here Lewis’s argument is quite straightforward. Premise (3) follows from Lewis’s definition of naturalism, together with his understanding of what it is to offer a “full” naturalistic explanation of a causal event. As we have seen, Lewis’s target is a form of naturalism that asserts that everything that exists is either physical or fully explicable in terms of the physical. This excludes nonreductionistic explanations in terms of “rational” causes, such as the fact that Jones drew a particular conclusion because he “saw” that the principle of modus ponens was logically valid. Lewis assumes, that is, that naturalists must believe that everything in nature, whether physical or nonphysical (for example, thoughts, if thoughts are nonphysical), must have physical causes— a version of what philosophers of mind call the causal closure principle (Dickerson, 2011, p. xxiii). For naturalists of this stripe, any so-called “higher level” aspects of human consciousness (for example, thoughts, purposes, reasons) can all be fully explained in terms of “lower level” sciences such as physics, chemistry, and neurophysiology. Any notion of mutually consistent “layered explanations” is rejected. So, too, is any possibility of causal “overdetermination” of mental events by invoking both physical and non- physical mental causes. Lewis’s intuition—widely shared in contemporary philosophy of mind (Robb and Heil, 2013)—is that “full” physical explanations of mental phenomena leave no room for mental events to do any causal work. As Lewis notes, even if reasons or grounds do exist, “what exactly have they got to do with the actual occurrence of the belief as a 16 GREGORY BASSHAM psychological event? If it is an event it must be caused. . . . [S]ince causes work inevitably, the belief would have had to arise whether it had grounds or not. . . . How could such a trifle as lack of logical grounds prevent the belief’s occurrence or how could the existence of grounds promote it?” (Lewis, 2001b, pp. 24-25). Thus, premise (3) follows from Lewis’s definition of naturalism, plus a plausible and widely shared view of what it means to offer a “full” physicalistic explanation. By now it should be clear that Lewis’s argument is not directed at all forms of naturalism, as that term is widely used today. His focus is on (1) “reductionistic naturalism,” the view that everything that exists is physical or reducible to the physical, and (2) “epiphenomenalistic naturalism,” the idea that everything that exists is either physical or a causally inert product, or epiphenomenon, of what is physical (Kim, 2011, pp. 96-97; cf. Lewis, 1967, pp. 99-100). If Lewis were writing today, he would need to address broader forms of naturalism, including so-called “emergentist” or “nonreductive” varieties. Could it be that nonreducible and causally efficacious nonphysical things (thoughts, meanings, values, evidential grounds) really exist, but only insofar as they “supervene” upon, or are “realized” in, physical structures (Clayton and Davies, 2006; Bedau and Humphreys, 2008)? What would Lewis say to naturalists of this ilk? As a Christian, Lewis would, of course, have rejected all forms of naturalism, at least inasmuch as naturalists exclude supernatural realities, like God. As noted earlier, Lewis argues in Miracles that reason could never have emerged from nonreason by means of an evolutionary process (Lewis, 2001b, pp. 27-30); that naturalism cannot account for the way thoughts can be “about” things (Lewis, 2001b, pp. 25-26); and that some kind of rational supernatural being must exist to explain why humans can reason (Lewis, 2001b, pp. 37-51). As I said, for reasons of space I shall not explore those complex topics here. But from what Lewis says in response to stricter forms of naturalism, I suspect that Lewis would have argued that emergentist naturalism is also epistemically self-undermining. Emergentism allows for non-physical thoughts and causal relations, but could it explain how material beings with “computers made out of meat” could ever “see” necessary truths (including normative truths, such as epistemic norms and, as he elsewhere argues, objective moral values) that exist outside their own minds? How is this commerce between electrochemical events in the brain and abstract, necessary propositions to be explained? Could emergent naturalism explain how propositional content—for example, the belief that no logical contradiction can be true—not only can cause mental acts, such as acts of inference, but can be the sole or determinative cause of such acts? How could something abstract, something that has no physical properties, such as propositional content, cause physical events in a naturalistic world? I will have more to say about this issue below. Introduction 17

Finally, let us turn to the last key premise in Lewis’s argument, (4): the claim that if an act of inference can be fully explained in terms of nonrational causes, then it does not involve any insight into real, objective relations of logical support. Lewis’s thought here is that certain kinds of causal explanations can act as defeaters for justified claims of “seeing,” “insight,” or more generally causal interaction with real objects. Lewis gives the example of hearing. If I know that the ringing in my ears “can be fully explained from causes other than a noise in the outer world,” “such as, say, the tinnitus produced by a bad cold” (Lewis, 2001b, p. 27), this information acts as a defeater for my thinking that I am actually hearing sounds that others can also hear. The same point can be made about putative cases of religious or moral perception. Suppose I have an experience of (what I think might be) God speaking to me from the hydrangea bush near my front porch. If I later learn that this experience can be fully explained by my having eaten a marijuana- laced brownie my cousin from Colorado sent me, I can have no rational confidence that it really was God speaking to me. Likewise with cases of putative moral perception. To adapt Hume’s famous example (Hume, 1978, pp. 468-469): Mel and his friend Pete are quarreling over whether Pabst Blue Ribbon is a better beer than Miller Lite. Mel flies into a rage and stabs Pete to death. Seeing this, I react with horror, indignation, “disapprobation”—the full panoply of negative Humean emotions—and the thought springs into my mind: “That was morally wrong.” Reflecting on this later, I wonder if this judgment reflects an act of genuine moral insight. Was there a mental perception of objective moral wrongness? Was my mind in contact with moral reality? Later, I discover that Max, my mad-scientist neighbor, implanted a microchip in my brain that caused me to think “That was morally wrong” just after I witnessed the stabbing. Max convinces me that this was the full explanation of why I had the thought. I now have a defeater for my belief that there was any genuine moral perception or insight—any contact with objective moral reality. I can have no confidence that there was any act of moral “seeing.” Such examples make clear why Lewis believes that naturalism undermines the normativity of reason as a connector to truths that are independent of our beliefs. We naturally trust reasoning to hook us up with truth, including necessary truths. But if all our acts of reasoning are fully explainable in terms of causal necessities involving neurochemical events in the brain, what reason do we have for thinking that the necessary hooks exist? Precisely the same neurochemical events could occur while we are plugged into the Scrambler, and all our beliefs would be false. Moreover, how is it that these neurochemical events in the brain (“C-fibers firing”) are able to make the necessary contact with reality (including normative and abstract reality)? Reasoning transcendentally requires insight (“seeing”) into objective inferential relations, and confidence that principles of inference connect us to 18 GREGORY BASSHAM reality. Naturalism is a defeater for such confidence and for belief in such insights. Therefore naturalists cannot consistently believe in reasoning. Yet naturalism can only be rationally believed and defended if reasoning exists. Thus, naturalism undercuts itself and is not rationally believable. Moreover, any view that implies that reasoning does not exist is false and should be rejected. Therefore, Lewis argues, naturalism is both false and not rationally believable.

D. Evaluating the Argument from Reason: The Central Points in Dispute

Is the argument from reason sound? In this volume, Victor Reppert, author of a widely read monograph on the argument from reason (Reppert, 2003), argues that it is. More precisely, he argues that a rather complex, updated version of Lewis’s argument can be successfully defended. Drawing upon some distinctions offered by the contemporary metaphysician William Hasker, Reppert argues that naturalism is epistemically self-undermining because it involves ontological commitments that conflict with certain epistemological commitments that any successful defense of naturalism would necessarily presuppose. The relevant ontological commitments are (1) that all events, at the base level, can be explained mechanistically; (2) the causal closure of the physical (that is, the principle that only physical events can cause other physical events); and (3) a principle of supervenience that entails that anything non-physical that exists in space and time must strongly supervene upon the physical, and thus be a necessary consequence of the physical. The relevant epistemological commitments are (a) that rational insight into logical relations plays a crucial role in acts of inference; (b) that thoughts sometimes possess qualities like intentionality, purposefulness, normativity, and a particular qualitative “feel”; (c) that propositional content (in other words, what a given thought is about) is operative in mental causation; and (d) that it is possible to form beliefs on the basis of evidence. In response to Reppert, David Kyle Johnson argues that Reppert relies on a loaded definition of naturalism to generate the alleged conflicts he claims. Johnson argues that there are varieties of naturalism, such as non- reductive materialism and mind-brain identity theory, which can embrace all of the epistemological commitments Reppert claims. Moreover, Johnson argues, naturalists need not accept a principle of causal closure that denies that mental events can have causal effects in the physical world. All that the naturalist must deny is that the physical is closed to anything outside the material world, not that physical events cannot have mental causes. Finally, Johnson challenges Reppert’s (and Lewis’s) assumption that reliable reasoning processes require a non-physical rational insight into logical relations. According to Johnson, computers can reason and produce reliable Introduction 19 beliefs even though they lack consciousness and operate purely mechanistically. As the debate between Reppert and Johnson makes clear, whether the argument from reason is sound turns crucially on (a) how “naturalism” should be defined and (b) what is required for reliable (and truth-tracking) reasoning to occur. Both are issues that continue to be hotly debated today.

3. The Moral Argument

Many sorts of moral arguments have been proposed as evidence for the existence of God; most reason from some feature of moral experience and conclude that God, or some higher power, is the best, or only, explanation of that feature. Versions of the moral argument have been offered by Kant (1956, p. 130), John Henry Newman (1979, pp. 98-108), Hastings Rashdall (1907, pp. 206-220), W. R. Sorley (1921, pp. 346-353), A. E. Taylor (1930, pp. 206- 210), and Frederick Copleston (1964, pp. 82-91); more recent defenses include Linda Zagzebski (2004), Mark Linville (2009), C. Stephen Evans (2010, pp. 107-148), David Baggett and Jerry L. Walls (2011, pp. 7-29), and Angus Ritchie (2012). Lewis’s version of the moral argument is probably the most famous and influential ever offered. The complete argument is complex and is laid out most fully in the first five chapters of Mere Christianity. In a nutshell, the argument runs as follows: Mankind is aware of a moral law, a real Right and Wrong, which is not of our own making, which we know we ought to obey, and which we cannot sincerely deny, forget, or suppress. Attempts to explain, or to explain away, this sense of objective morality in terms of “herd instinct,” social convention, or other such naturalistic explanations, all fail. Thus, we must acknowledge that there are values in the world, not just facts (Lewis, 2001, p. 27). At the same time, all of us are aware that we have failed to keep the moral law, and we feel guilty when we do. Something must lie behind this real moral law; a bit of matter cannot give moral commands or instructions. Only something mind-like could do that. Thus, some mind-like Power must exist that directs the universe, gives us our sense of right and wrong, and makes us feel responsible and uncomfortable when we do wrong. In this argument, Lewis draws attention to three alleged facts of moral experience: our awareness of an objective moral law not of our own making, our sense that this law must be obeyed, and feelings of guilt when we inevitably fall short of obeying the moral law. Suppose, following Erik Wielenberg, we call these three features “Lewisian moral phenomena.” We can then formulate the kernel of Lewis’s moral argument, as Wielenberg does (Wielenberg, 2008, p. 63), as follows:

20 GREGORY BASSHAM

(1) Lewisian moral phenomena exist. (2) The best explanation of the existence of Lewisian moral phenomena is the existence of a Higher Power that created the universe. (3) So: There is a Higher Power that created the universe (from 1 and 2). (4) The Higher Power issues instructions and wants us to engage in morally right conduct. (5) If (4), then there is a good, mindlike Higher Power that created the universe. (6) Therefore, there is a good, mindlike Higher Power that created the universe (from 4 and 5).

There are two important things to note about this argument. One is that it is an abductive argument—that is, an inference to the best explanation. Its success, therefore, depends on whether it is in fact possible to rule out, or prevail over, all competing explanations. The other important thing to note is that the argument purports to prove only the existence of a “good, mindlike Higher Power,” not the God of classical theism. As Lewis acknowledged (Lewis, 2001, p. 30), further arguments are needed to show that this Higher Power is a person who possesses all the traditional attributes of the Judeo- Christian God. In this volume, Lewis’s moral argument is defended by David Baggett, executive director of the moralapologetics.com Web site, and a leading authority on moral arguments for the existence of God. Baggett focuses, in particular, on the fact that some moral truths appear to be necessary truths. In his view, the existence of necessary moral truths fits comfortably within a theistic worldview, but poorly within a naturalistic one. Responding to Baggett, Erik Wielenberg argues that there is an objectionable obscurity in attempts to identify God with the Good or with the moral law. How can God (a person) be identical with a property (goodness) or a set of normative propositions (the moral law)? Moreover, Wielenberg argues, naturalists can account for the existence of necessary moral truths. They can do this, he argues, by embracing a view known as robust normative realism. On this view, moral properties like moral rightness or goodness necessarily supervene on natural properties, but are not reducible to those properties. For instance, Wielenberg argues, the fact that an act is a case of causing pain just for fun makes the act morally wrong. Wrongness is a normative property, distinct from the natural property of causing pain just for fun. But there is a strong causal dependence between the two properties. The fact that an act is a case of causing pain just for fun causes the act to be wrong in a particularly robust (and in fact necessary) fashion. This explains why the proposition “Acts that cause pain just for fun are (prima facie) morally wrong” is a necessary truth, even though there is no God who eternally and infallibly believes that proposition. Introduction 21

Wielenberg offers one standard way of responding to Lewis’s moral argument: To deny that naturalism cannot account for necessary moral truths. Another common strategy is to deny that “objective morality” exists at all. Could it be, as J. L. Mackie, Michael Ruse, and other naturalistic philosophers argue, that belief in objective moral values is an illusion foisted on us by our genes or by the bewitchment of language or culture? (Ruse, 1986, p. 252; Mackie, 1977, pp. 35-49). Are there plausible non-objectivist ethical theories that Lewis does not consider? (Beversluis, 2007, pp. 83-87). These are important questions that both defenders and critics of Lewis’s moral argument must thrash out.

4. The Trilemma Argument

The so-called trilemma argument is an attempt to show that Christ is God (or “Lord”). The argument is actually very old. Early versions of the argument are found in some of the Church Fathers (Brazier, 2011, pp. 106-107). Detailed formulations of the argument that preceded Lewis’s version are found in H. P. Liddon’s 1866 Bampton Lectures (Liddon, 1867, Lecture 4), W. K. Chesterton’s The Everlasting Man (1993, pp. 196-198), and Ronald Knox’s The Belief of Catholics (1958, pp. 91-95), among others. Notable recent defenses of the argument include Stephen Davis (2002, pp. 221-245; 2004, pp. 480-492), David A. Horner (2008, pp. 68-84), Peter S. Williams (2013, pp. 184-188), Josh McDowell (1979, pp. 103-107), and Peter Kreeft and Ronald Tacelli (1994, pp. 155-174). Lewis fullest and most explicit statement of the argument is in Mere Christianity. He there writes:

I am trying here to prevent anyone saying the really foolish thing that people often say about Him: I’m ready to accept Jesus as a great moral teacher, but I don’t accept His claim to be God. That is the one thing we must not say. A man who was merely a man and said the sort of things Jesus said would not be a great moral teacher. He would either be a lunatic—on the level with the man who says he is a poached egg—or else he would be the Devil of Hell. You must make your choice. Either this man was, and is, the Son of God: or else a madman or something worse. You can shut Him up for a fool, you can spit at Him and kill Him as a demon; or you can fall at His feet and call him Lord and God, but let us not come with any patronising nonsense about His being a great human teacher. He has not left that open to us. He did not intend to. . . . Now it seems to me obvious that He was neither a lunatic nor a fiend: and consequently, however strange or terrifying or unlikely it may seem, I have to accept the view that He was and is God (Lewis, 2001a, pp. 52- 53).

22 GREGORY BASSHAM

In short, Lewis argues, those who wish to say that Jesus was a great moral teacher but not God face a trilemma, or difficult three-fold choice. Often, this choice is put alliteratively as “Liar, Lunatic, or Lord” (or alternately as “Mad, Bad, or God.”). In this form, the argument runs: Jesus claimed to be God. Either he was God, or he was a liar or a lunatic. But Jesus was not a liar or a lunatic. Therefore he was God. Many critics have argued that formulating the argument as a trilemma is too simple. For instance, Daniel Howard-Snyder has argued that there is a fourth alternative that should be seriously considered—namely, that Jesus believed that he was in some sense divine, but was sincerely mistaken (Howard-Snyder, 2004, pp. 473-477). Expanding the argument in this way, Howard-Snyder helpfully reformulates Lewis’s argument as a “quadrilemma,” as follows:

(1) Jesus claimed, explicitly or implicitly, to be divine. (2) Either Jesus was right or he was wrong. (3) If he was wrong, then either (a) he believed he was wrong and he was lying, or (b) he did not believe he was wrong but he was institutionalizable, or (c) he did not believe he was wrong and he was not institutionalizable; rather, he was merely mistaken. (4) He was not lying, i.e. (a) is false. (5) He was not institutionalizable, i.e. (b) is false. (6) He was not merely mistaken, i.e. (c) is false. (7) So, he was right; i.e. Jesus was, and presumably still is, divine (Howard-Snyder, 2004, p. 457).

Is this argument sound? Some critics have challenged, or at least questioned, the claim that Jesus claimed to be God (Beversluis, 2007, pp. 116- 118). This is the main ground on which Adam Barkman criticizes Lewis’s argument in this volume. According to Barkman, Jesus may well have claimed to be some sort of divine, Messianic “Son of Man,” but there is no substantial evidence that he claimed to be God Incarnate. Other critics, pointing to Biblical passages that suggest that “many” of Jesus’ hearers thought he was “raving mad” (John 10:20; NIV), and that even Jesus’ own family apparently believed for a time that he was “out of his mind” (Mark 3:21), argue that Jesus may have been deluded about his own identity (Beversluis, 2007, pp. 125-127). In addition, Daniel Howard-Snyder has argued that the trilemma argument fails because of a “dwindling probabilities” problem. The difficulty, in brief, is this: The trilemma argument is complex, and succeeds only if we have strong grounds for affirming each of four distinct claims: (1) Jesus claimed to be God, (2) Jesus was not lying, (3) Jesus was not insane or delusional, and (4) Jesus was not sincerely mistaken. Introduction 23

According to Howard-Snyder, the proper way to assess complex, multi- pronged arguments of this sort is to apply the probability calculus, which requires that we multiply the epistemic probabilities of the individual conjuncts. This creates problems at two levels. One is that modern Biblical scholarship raises serious doubts about whether Jesus did, in fact, claim to be God (Young, 1977, pp. 13-47). Many of the key passages that allegedly support such a claim come from the Gospel of John, which many scholars regard as the least historically reliable of the four Gospels. But even if we set aside this problem and assign probabilities to each of the four conjuncts a good deal higher than most reputable New Testament scholars would be prepared to do, we may still wind up, when we multiply the conjuncts, with an overall probability that is too low to support the conclusion (Howard-Snyder, 2004, pp. 458-462; for a critique see Davis, 2004, pp. 481-485). In this way, Howard-Snyder argues, the trilemma falls victim to its own complexity. In this volume, the trilemma argument is vigorously defended by Donald T. Williams, a well-known Lewis scholar and President of the International Society of Christian Apologetics. Williams argues that the trilemma argument is basically sound, but is often misunderstood. In his view, Lewis did not attempt to prove the divinity of Christ, but merely posed a choice: Either we must accept Jesus’ claim to be God, or we must conclude that he was immoral or insane. Moreover, Williams argues, we must remember the original context of Lewis’s argument. The scripts that later became Mere Christianity were originally delivered as 15-minute radio broadcasts over the BBC. Lewis was speaking to a lay audience, and so naturally did not delve into many issues (such as the historicity of the Gospels) that would have required a fuller and more scholarly treatment if he were addressing his fellow academics. When all due allowances are made, Williams claims, Lewis’s Trilemma is still a strong argument.

5. The Problem of Evil

What academics call “the problem of evil” (or, less commonly, “the problem of pain”) is the problem of trying to square belief in God with the reality of evil. Is it unreasonable, as many atheists claim, to believe in God given all the evil and suffering in the world? If God does exist, why does he permit so much evil, and so much apparently pointless evil? For many people, the problem of evil is the single biggest challenge to religious belief. Lewis devoted an entire book (The Problem of Pain) to the problem of evil, and discussed it in many other places, including (most movingly and personally) in A Grief Observed. Lewis does not claim to offer a complete “solution” to the problem of suffering. In fact, he argues that we should not expect to fully understand why God permits evil. Humans cannot possibly see “the big picture” as God does. Nor, as finite, temporal creatures, can we truly grasp how “the sufferings of this present time are not to be compared with the 24 GREGORY BASSHAM glory which shall be revealed in us" (Rom. 8:18; quoted in Lewis, 2001c, p. 148). But the problem runs even deeper than this, Lewis argues. “[I]f God is wiser than we His judgments must differ from ours on many things, and not least on good and evil. What seems to us good may therefore not be good in His eyes, and what seems to us evil may not be evil” (Lewis, 2001c, p. 28). This does not mean that what we consider good could be radically different from what God considers good. That would make it empty to speak of God as “good,” and would take away all moral reasons for loving and obeying him (Lewis, 2001c, pp. 28-29). Still, it would be foolish to imagine that we, by a happy twist of fate, have nearly infallible moral radar, when so many others plainly do not. This point—that our modern ideas of good and evil may be mistaken in important respects—plays a crucial part in Lewis’s response to the problem of evil. Lewis’s reflections on the problem of evil are complex and multi-layered, but his major claims include the following:

(1) The inexorable nature thesis: Any society of embodied free-willed beings presupposes a common environment or Nature that is relatively fixed and unalterable. In such a shared world, the possibility of both evil thoughts (sin) and evil outward acts (for example, assaulting another person with a stick) cannot be excluded. No meaningful free-will would be possible if God were to continually perform miracles to prevent such evil thoughts and acts (Lewis, 2001c, pp. 19-22). (2) The tough-love thesis: God is supremely good, but not “kind” in the way many people today tend to assume. God is not a “senile benevolence” in the sky, interested mainly in “a good time being had by all.” Rather, God is infinite Love, and “love, in its own nature, demands the perfecting of the beloved” (Lewis, 2001c, p. 38). We are a Divine work of art, something God is making, and he will not be satisfied until we are molded in his likeness and share in his perfect love and goodness. Such a process requires total self- surrender and is inherently arduous and painful (Lewis, 2001c, pp. 34-47). (3) The horror thesis: The Bible teaches that we are sinners and a “horror to God” (Lewis, 2001c, p. 62). Yet many people today no longer feel a strong sense of sin. No real understanding of why God permits evil is possible unless we recover a sense of our true condition. (4) The shattering thesis: In our present condition, pain is not a sheer evil. Rather, it is “God’s megaphone to rouse a deaf world” (Lewis, 2001c, p. 91). In particular, pain benefits us by (a) sending a clear signal to the wicked that they must change (Lewis, 2001c, pp. 91- 94), (b) shattering our illusion that we can get along without God Introduction 25

(Lewis, 2001c, pp. 94-96), (c) letting us know when we are truly acting for God’s sake, instead of for some less worthy motive (Lewis, 2001c, pp. 97-104), and (d) providing opportunities for testing ourselves and for moral and spiritual growth through challenge and tribulation (Lewis, 2001c, pp. 104-109)—the ancient Christian themes of being made “perfect through suffering” (Heb. 2:10) and of this world as a “vale of soul making” (Lewis, 2001c, p. 109).

In addition to these four major theses, Lewis offers insightful reflections in The Problem of Pain on the Fall and original sin, hell, animal pain, and heaven. In this collection, Lewis’s response to the problem of evil is ably defended by Philip Tallon, Assistant Professor of Theology at Houston Baptist University and author of The Poetics of Evil: Toward an Aesthetic Theodicy (Oxford University Press, 2011). Tallon notes that The Problem of Pain is a short book aimed at lay readers, and so naturally does not go into great theological depth or develop points with the rigor that would be expected in a more scholarly treatise. Nevertheless, Tallon argues, The Problem of Pain is a great book: thoughtful, scriptural, steeped in the Western tradition, and remarkably wide-ranging. Those who fault it usually view the book as something it is not, or approach it from very different starting points. Tallon’s discussion partner in this volume, David O’Hara, is an example of a critic who disagrees with some of Lewis’s fundamental assumptions. O’Hara, who teaches philosophy and classics at Augustana College, argues that Lewis’s reflections in The Problem of Pain are “inadequate and of very limited value,” because they treat suffering as primarily an intellectual problem rather than as a call to awe, love, and compassion. As O’Hara sees it, over-intellectualizing the problem of evil can lead to complacency and distract us from what is far more important: a life that models Christ’s faithfulness and love in lived Christian service. Moreover, O’Hara contends, Lewis’s arguments can lull us into a false sense of having “figured God out.” As the Bible says, God’s thoughts are not our thoughts, and His ways are not our ways (Isaiah 55:8). Ultimately, O’Hara believes, our response to evil must be that of Job’s: to acknowledge our finitude and the limits of our understanding, to keep faith even in the dark night of the soul, to respond to God and to life with awe and gratitude, and to repent “in dust and ashes” (Job 42:6).

6. Conclusion

As a religious writer, Lewis tends to provoke polarized reactions: he is both widely loved and widely disdained. In this book, we try to approach Lewis’s Christian apologetics in a different spirit. Our goal is to take Lewis’s measure 26 GREGORY BASSHAM as an apologist, not from the standpoint of partisans in today’s science-and- religion wars, but as scholars seeking an honest and focused appraisal of the true strengths and weaknesses of his arguments. Readers will note that few of our “pro” contributors think that Lewis’s arguments are completely sound and unassailable, and few on the “con” side believe that the arguments are wholly without merit. Our goal, throughout, has been to explain Lewis’s arguments as carefully as possible, to unpack them when necessary, and to evaluate them with the same standards of scholarly rigor and care that would be used with any other intellectually challenging author. What emerges, we hope, is a deeper and more balanced understanding of Lewis’s apologetics and of Lewis the Christian intellectual: neither infallible sage nor bombastic sophist, but “Oxford’s bonny apologist,” whose writings, five decades after his death, continue to challenge and inspire.

One

PRO: A DEFENSE OF C. S. LEWIS’S ARGUMENT FROM DESIRE

Peter S. Williams

As a deer longs for a stream of cool water, so I long for you, O God. —Psalm 42:1

The “argument from desire” (AFD) is a family of arguments for theism from human desire. C. S. Lewis famously described and argued from a desire that he called “Joy.” I am sympathetic to the view that the AFD is, as Alister McGrath writes, an argument “primarily for the existence of Heaven or ‘another world,’ and only secondarily for the existence of God” (McGrath, 2014, p. 105). Indeed, “a close reading of Lewis’s works suggests that Lewis’s ‘argument from desire’ is primarily about the intuition of ‘another world’ – Heaven – rather than God” (McGrath, 2014, p. 115). Framing the AFD in terms of “a natural desire for transcendent fulfilment” (a move that fits the all-embracing nature of Joy as described by Lewis), McGrath notes: “This is not really an argument for the existence of God, in the strict sense of the term. For a start, we would need to expand Lewis’s point to include the Christian declaration that God either is, or is an essential condition for, the satisfaction of the natural human desire for transcendent fulfillment” (McGrath, 2012, p. 110). John Haldane likewise frames the AFD in terms of “a natural desire for transcendent fulfillment, which cannot be attained or experienced in the present world” (Haldane, 2007, p. 47). He concludes: “Neither Lewis’s discursive reasoning nor this argument refers immediately to God as the end of transcendent desire, but the original discussion makes it clear that what is at issue is a beatific afterlife” (Haldane, 2007, p. 47). Hence Haldane extends his AFD via the premise that “God is, or is an essential condition of the satisfaction of the natural desire for transcendent fulfillment” (Haldane, 2007, p. 47). In this McGrath and Haldane follow Robert Holyer, who argues that since “Joy is a desire . . . for ultimate happiness” and “since ultimate happiness is not possible without God, Joy is in a sense also a desire for God” (Holyer, 1988, p. 66). That said, some formulations of the AFD (especially when framed abductively, or as a reductio) invoke God more directly.

28 PETER S. WILLIAMS

1. Lewis on the Phenomenology of Joy

In Lewis’s autobiography, Surprised by Joy, he recounts how the Castlereagh Hills of Northern Ireland “taught me longing—Sehnsucht; made me for good or ill, and before I was six years old, a votary of the Blue Flower” (Lewis, 1998, p. 6). Alister McGrath explains that “leading German Romantic writers . . . used the image of a ‘Blue Flower’ as a symbol of the wanderings and yearnings of the human soul, especially as this sense of longing is evoked— though not satisfied—by the natural world” (McGrath, 2013, p. 16). The German term Sehnsucht describes this combined sense of longing for, and displacement or alienation from, the object of one’s desire. Sehnsucht is a “nostalgic longing” that arises when experience of something within the world awakens in us a desire for something beyond what the world can offer as a corresponding object of desire: “The thing I am speaking of is not an experience,” as Lewis writes in The Problem of Pain. “You have experienced only the want of it” (Lewis, 1977b, p. 118). Sehnsucht directs our attention towards the transcendent. Corbin Scott Carnell states that “Sehnsucht may be said to represent just as much a basic theme in literature as love” (Carnell, 1974, p. 23). Throughout Surprised by Joy, Lewis traces his experience of this “unsatisfied desire which is itself more desirable than any other satisfaction” (Lewis, 1998, p. 12), which he calls “Joy” (Lewis, 1998, p. 12). Lewis found that Joy was occasioned by his appreciation of various aspects of reality, but pointed beyond such experiences: “Only when your whole attention and desire are fixed on something else—whether a distant mountain, or the past, or the gods of Asgard—does the “thrill” arise. It is a by- product. Its very existence presupposes that you desire not it but something other and outer” (Lewis, 1998, p. 130). Joy is thus a state of consciousness that is occasioned by various objects of experience, but is not a desire for any of the objects of experience that occasion it. Nor is it merely a mediated positive experience of a transcendent reality, nor even a desire to experience the “thrill” of being in this particular psychological state. Rather Joy is a desire for a transcendent object of satisfaction. As Thomas Chalmers wrote: “There is in man . . . an interminable longing after nobler and higher things . . . which never is appeased by all that the world has to offer” (Chalmers, p. 233, quoted by Hammond, 1943, p. 56). The Preface to Lewis’s novel The Pilgrim’s Regress describes “a particular recurrent experience which dominated my childhood and adolescence and which I hastily called ‘Romantic’ because inanimate nature and marvellous literature were among the things that evoked it” (Lewis, 1977a., p. 7). The power of fairy tales (undoubtedly part of what Lewis means by “marvelous literature”) lies in their ability to transport us into a world imbued with “romantic” Sehnsucht. Consider this passage from a book Lewis loved, The Wind in the Willows:

Pro: A Defense of C. S. Lewis’s Argument from Desire 29

A bird piped up suddenly, and was still; and a light breeze sprang up and set the reeds and bulrushes rustling. Rat, who was in the stern of the boat, while Mole sculled, sat up suddenly and listened with a passionate intentness. Mole, who with gentle strokes was just keeping the boat moving while he scanned the banks with care, looked at him with curiosity. “It’s gone!” sighed the Rat, sinking back in his seat again. “So beautiful and strange and new! Since it was to end so soon, I almost wish I never had it. For it has roused a longing in me that is pain, and nothing seems worth while but just to hear that sound once more and go on listening to it forever” (Grahame, 1993, pp. 126–127).

In The Pilgrim’s Regress, Lewis describes the “Romantic” experience of Joy (or “Desire,” as he then called it) as follows:

The experience is one of intense longing. It is distinguished from other longings by two things. In the first place, though the sense of want is acute and even painful, yet the mere wanting is felt to be somehow a delight. Other desires are felt as pleasures only if satisfaction is expected in the near future: hunger is pleasant only when we know (or believe) that we are soon going to eat. But this desire, even when there is no hope of possible satisfaction, continues to be prized, and even to be preferred to anything else in the world, by those who have once felt it. This hunger is better than any other fullness; this poverty better than all other wealth. And thus it comes about, that if the desire is long absent, it may itself be desired, and that new desiring becomes a new instance of the original desire. . . . In the second place, there is a peculiar mystery about the object of this Desire. Inexperienced people (and inattention leaves some inexperienced all their lives) suppose, when they feel it, that they know what they are desiring. Thus if it comes to a child while he is looking at a far off hillside he at once thinks “if only I were there”; if it comes when he is remembering some event in the past, he thinks “if only I could go back to those days.” . . . But every one of these impressions is wrong. . . . I know them to be wrong not by intelligence but by experience. . . . For I have myself been deluded by every one of these false answers in turn, and have contemplated each of them earnestly enough to discover the cheat (Lewis, 1977a, pp. 7-8).

The “cheat” is that “every one of these supposed objects of Joy is inadequate to it. An easy experiment will show that by going to the far hillside you will get either nothing, or else a recurrence of the same desire which sent you thither” (Lewis, 1977a, p. 9). Not only is each of the supposed earthly 30 PETER S. WILLIAMS objects of the “Romantic” desire inadequate to it, but it is a defining feature of the desire that it is occasioned by all manner of earthly objects (occasions that may vary from person to person), which may thus be mistaken as the object of the desire they occasion, but which are discovered by bitter experience not to be the object of the desire they occasion:

Lust can be gratified. Another personality can become to us “our America, our New-found-land.” A happy marriage can be achieved. But what has any of the three, or any mixture of the three, to do with that unnameable something, desire for which pierces us like a rapier at the smell of a bonfire, the sound of wild ducks flying overhead, the title of The Well at the World’s End, the opening lines of Kubla Khan, the morning in late summer, or the noise of falling waves? (Lewis, 1977a, pp. 9-10).

Richard Dawkins notes: “It is often said that there is a God-shaped gap in the brain which needs to be filled” (Dawkins, 2006, p. 347). It was of course Pascal who famously said that there exists in humans an “infinite abyss [that] can only be filled by an infinite and immutable object, that is to say, only by God Himself” (Pascal, 1941, p. 135). As a rejoinder to this line of thought, Dawkins poses a rhetorical question: “[C]ould it be that God clutters up a gap that we’d be better off filling with something else? Science, perhaps? Art? Human friendship? Humanism? Love of this life in the real world, giving no credence to other lives beyond the grave?” (Dawkins, 2006, p. 347). Dawkins’s rejoinder is too shallow because it is in the very appreciation of science, art, human friendship and so on that we discover an unsatisfied desire for something “deeper” and “more.” Indeed, is it not often those who have the most in worldly terms who find life most hollow? (cf. Ecclesiastes 2:10-11). In his sermon “The Weight of Glory,” Lewis argues that “if we are made for heaven, the desire for our proper place will be already in us, but not yet attached to the true object” (Lewis, 2002, p. 98), thereby leaving us with “a desire for something that has never actually appeared in our experience” (Lewis, 2002, p. 98). Lewis suggests that we do indeed find such a desire within ourselves and that our “commonest expedient is to call it beauty and behave as if that had settled the matter. . . . But all that is a cheat. . . . The books or the music in which we thought the beauty was located will betray us if we trust them; it was not in them, it only came through them, and what came through them was longing” (Lewis, 2002, p. 98). As C. E. M. Joad mused, “aesthetic emotion is at once the most satisfying and the most unsatisfying of all the emotions known to us; satisfying because of what it gives, unsatisfying because it gives so briefly, and, in the act of giving hints at greater gifts withheld” (Joad, 1946, p. 354). Thus Lewis affirms “we remain conscious of a desire which no natural happiness will satisfy” (Lewis, 2002, p. 98). In Mere Christianity, Lewis concludes: Pro: A Defense of C. S. Lewis’s Argument from Desire 31

Most people, if they had really learned to look into their own hearts, would know that they do want, and want acutely, something that cannot be had in this world. There are all sorts of things in this world that offer to give it to you, but they never quite keep their promise. The longings which arise in us when we first fall in love, or first think of some foreign country, or first take up some subject that excites us, are longings which no marriage, no travel, no learning, can really satisfy. There was something we grasped at, in the first moments of longing, which just faded away in the reality. . . . [S]omething has evaded us (Lewis, 1997, p. 112)..

2. Taking Joy at Face Value

Joy is the tantalising experience of feeling deeply drawn to a transcendent and innately desirable “something more” that lies beyond one’s grasp. This experience appears to be natural to humanity, being spontaneously occasioned but not satisfied by various worldly “triggers.” These “triggers” are somewhat person-dependent, but often have to do with experiences of beauty or natural grandeur (what the Romantics would call “the sublime”). As Elaine Scarry writes: “Something beautiful fills the mind yet invites the search for something beyond itself, something larger” (Scarry, 1999, p. 29). In other words: “Human beings are brought into this world with a planted seed that sprouts a desire for ultimate meaning and happiness that just will not go away” (Puckett Jr., 2013, p. xiv). Alister McGrath agrees that “poignant desire for something agonizingly elusive can be documented from all ages, and is not limited to individuals who might be deemed to be “religious” or “spiritual” (McGrath, 2014, p. 106). Atheist Bruce Sheiman admits:

I want to believe that the reason we finite beings reach out to an ineffable and unfathomable Absolute is because we are Imago Dei. I want to believe that our timeless quest for goodness and transcendence has its Omega Point in God. . . . [E]ven though I cannot believe in God, I still feel the need for God (Sheiman, 2009, p. ix).

Although the “unfathomable Absolute” is not given much metaphysical specificity in our enjoyment of Joy, contemplation of this experience does not leave us with a mere blank, a felt absence of “something we know not what.” Joy is more like a smell arousing our appetite for some exotic food than the arousal of a desire for a wholly mysterious “X.” The object of Joy transcends the worldly triggers that occasion our desire for it. Ockham’s Razor—the principle that simpler explanations should 32 PETER S. WILLIAMS be preferred over otherwise equally satisfactory complex ones—encourages us to think in terms of a common object of desire behind our varied experiences of Joy. This transcendent object appears to be not merely desired, but innately desirable. If the mere desire for this transcendent object “is itself more desirable than any other [worldly] satisfaction” (Lewis, 1998, p. 12), then the transcendent object itself can hardly fail to be more desirable than any other object, whether within this world or beyond it. Joy hints at the tantalizing possibility that the human condition (in at least its essential elements) might be brought to some sort of fulfilled flourishing in relationship to, or participation in, the innately desirable transcendent “other” vaguely glimpsed through Joy. Joy points beyond the ontological monism of either metaphysical naturalism or pantheism. Naturalism cannot accommodate this transcendent reality. To interpret Joy as a longing for some sort of pantheistic nirvana beyond all distinctions would be to collapse the distinction between the unfulfilled desire and the fulfilment of desire. Contemplation of Joy thus suggests a worldview that is theistic or idealist. Samuel Alexander’s Gifford Lectures on Space, Time and Deity introduced Lewis to the distinction between “Enjoyment” and “Contemplation,” a distinction Lewis himself would later illustrate in terms of “the difference between looking at and looking along” a beam of light (Lewis, 1985, p. 50). Applying this distinction to his experience of Joy convinced Lewis that he could not “Contemplate” (look at) Joy while simultaneously “Enjoying” (looking along) it: “You cannot hope and also think about hoping at the same time; for in hope we look to hope’s object and we interrupt this by (so to speak) turning round to look at the hope itself” (Lewis, 1998, p. 170). To look at Joy by focusing upon its worldly triggers in the hope of recovering the emotional thrill attendant upon Joy was to confuse the means and the end, the sign with that to which it pointed. Lewis consequently stopped deliberately pursuing the emotional thrill that accompanied the enjoyment of Joy, “understanding that it was not heavenly desire that was important, but the object towards which it was pointing” (Barkman, 2009, pp. 115-116). The phenomenology of Joy is an inherently teleological experience that points beyond itself to a transcendent and innately desirable “something more.” To take the phenomenology of Joy at face value is to look along it towards “the transcendent other.” As Lewis points out: “As soon as you have grasped this simple distinction [between looking at and looking along], it raises a question. You get one experience of a thing when you look along it and another when you look at it. Which is the ‘true’ or ‘valid’ experience?” (Lewis, 1985, p. 51). Lewis observes:

It has . . . come to be taken for granted that the external account of a thing somehow refutes or “debunks” the account given from inside. “All these moral ideas which look so transcendental and beautiful from Pro: A Defense of C. S. Lewis’s Argument from Desire 33

inside,” says the wiseacre, “are really only a mass of biological instincts and inherited taboos.” And no one plays the game the other way round by replying, “If you will only step inside, the things that look to you like instincts and taboos will suddenly reveal their real and transcendental nature” (Lewis, 1985, p. 52).

Lewis argues that the reductive impulse must be resisted on at least some occasions because its application to every experience becomes incoherent: “you can step outside one experience only by stepping inside another. Therefore, if all inside experiences are misleading, we are always misled” (Lewis, 1985, p. 54). Lewis concludes that “we must take each case on its merits. But we must start with no prejudice for or against either kind of looking” (Lewis, 1985, p. 54). However, contemporary epistemology is much more open to playing the game “the other way round” (consider the rise of Reformed Epistemology). Indeed, Lewis’s own example of discovering that “the inside vision of the savage’s dance to Nyonga may be found deceptive because we find reason to believe that crops and babies are not really affected by it” (Lewis, 1985, p. 52) illustrates the presumption of innocence conferred in the absence of sufficient reason for doubt upon enjoyed (that is, looked- along) experiences by Richard Swinburne’s “principle of credulity.” As Swinburne explains, “it is a basic principle of rationality, which I call the principle of credulity, that we ought to believe that things are as they seem to be (in the epistemic sense) unless and until we have evidence that we are mistaken” (Swinburne, 2010, p. 115). This basic principle of rationality puts the burden of proof on the shoulders of the sceptic who claims that, despite appearances, to look along Joy is to experience a delusion rather than the insight into the nature of reality it seems to be from the inside. The reductive impulse of metaphysical naturalism cannot be allowed to beg the question against the prima facie evidence to the contrary contained within the enjoyment of Joy. As John Cottingham argues, to ignore the deeply rooted human impulse to transcendence is to shoulder the burden of explaining it away and showing it to be inauthentic: “Intellectual integrity requires that we do not ignore our deepest sensibilities and impulses. Either they must be explained away as distortions or seductive illusions, or their authenticity as possible pointers to something real must be acknowledged. There is no middle ground” (Cottingham, 2009, pp. 15-16). Given the epistemological primacy of trust here, it follows that in the absence of sufficient reason to think that “we are indeed without point or purpose,” we should take our longings to be “purposeful in something close to the terms in which they represent themselves” (Haldane, 2007, p. 248). Furthermore, after quoting “Augustine’s famous dictum that “our hearts are restless till they rest in you, O Lord,” Alvin Plantinga argues for the properly basic status of theism evoked by desire:

34 PETER S. WILLIAMS

Perhaps this restlessness without God leads to belief in God; and perhaps God has designed us in this way to impel us to try to get in touch with him . . . the process leading to the formation of the beliefs in question are directed to the truth: the relevant module of the design plan has as its purpose the production of true belief, even if it goes by way of perception of beauty or wish-fulfillment (Plantinga, 2000, p. 307).

3. Beyond the Prima Facie Case

Having concluded that “[o]ur striving and struggling, wishing and wanting, seek completion in something that is itself complete . . . something that made us for itself, not as an act of narcissism but as one of gratuitous generosity, and something that has the power to redeem innocent suffering” (Haldane, 2007, pp. 247–8), John Haldane argues that:

Should our yearnings be without the possibility of completion then we are indeed without point or purpose; but should these longings be purposeful in something close to the terms in which they represent themselves, then we can hope to enter into the eternal company of God. That is the prospect offered by the theism I have been concerned to argue for, and it is, I believe, the best explanation of our heart’s desire (Haldane, 2007, p. 248).

To the prima facie evidence of Joy we can add a cumulative handful of arguments for the conclusion that theism is “the best explanation of our heart’s desire.” I shall discuss four such arguments, which I will call the abductive argument, the inferential argument, the deductive argument, and the reductio argument from absurdity.

A. The Abductive Argument

Therefore he willed that the hearts of Men should seek beyond the world and should find no rest therein.—J. R. R. Tolkien

In Surprised by Joy, Lewis traces his life via a series of lived experiments wherein he pursues Joy and discovers that one object of attention after another occasions but fails to fulfil the desire:

What I like about experience is that it is such an honest thing. You may take any number of wrong turnings; but keep your eyes open and you will not be allowed to go very far before the warning signs appear. You Pro: A Defense of C. S. Lewis’s Argument from Desire 35

may have deceived yourself, but experience is not trying to deceive you. The universe rings true wherever you fairly test it (Lewis, 1998, p. 137).

Lewis describes his spiritual life as a series of “wrong turnings” in pursuit of the reality that would satisfy the desire he calls Joy, and how his eventual admission of God’s existence, and subsequent conversion to Christianity, brought him to the conclusion that Joy was the desire for a proper relationship with God—a destination that cannot be fully attained until heaven, but which (partly by a process of experiential elimination) can be located and glimpsed from within the present world:

But what, in conclusion, about Joy? for that, after all, is what the story has mainly been about. To tell you the truth, the subject has lost nearly all interest for me since I became a Christian. . . . I believe (if the thing were at all worth recording) that the old stab, the old bitter-sweet, has come to me as often and as sharply since my conversion as at any time of my life whatever. But I now know that the experience, considered as a state of my own mind, had never had the kind of importance I once gave it. It was valuable only as a pointer to something other and outer. While that was in doubt, the pointer naturally loomed large in my thoughts. When we are lost in the woods the sight of a signpost is a great matter. He who first sees it cries “Look!” The whole party gathers round and stares. But when we have found the road and are passing signposts every few miles, we shall not stop and stare. They will encourage us and we shall be grateful to the authority that set them up. But we shall not stop and stare, or not much; not on this road, though their pillars are of silver and their lettering of gold. “We would be at Jerusalem” (Lewis, 1998, p. 185).

Alister McGrath points out that “Lewis’s reflections on desire focus on two themes . . . a general sense of longing for something . . . and a Christian affirmation that God alone is the heart’s true desire, the origin and goal of human longings” (McGrath, 2014, p. 106). These two themes form the two prongs of an abductive argument for “Heaven” (that is, an “eternal relationship with God”) as the best explanation of the common experience of Joy:

Lewis saw this line of thought as demonstrating the correlation of faith with experience, exploring the “empirical adequacy” of the Christian way of seeing reality with what we experience within ourselves . . . Christianity, he points out, tells us that this sense of longing for God is exactly what we should expect, since we are created to relate to God. It fits in with a Christian way of thinking, thus providing indirect 36 PETER S. WILLIAMS

confirmation of its reliability. There is a strong resonance between theory and observation—between the theological framework and the realities of our personal experience (McGrath, 2012, pp. 110-111).

Victor Reppert argues that the experience of Joy supports theism because such an experience is less expected within a naturalistic universe than it is within a theistic one:

To see if the Bayesian argument from desire has any weight at all, let’s assume that we have a person who thinks that theism and atheism are equally likely. . . . The next question is how probable is the desire evidence to arise if theism is true. It seems that theism gives us a reason to suppose that these desires would be likely to arise in a theistic universe, especially if that universe were a Christian universe. On Christian theism God’s intention in creating humans is to fit them for eternity in God’s presence. As such, it stands to reason that we should find ourselves dissatisfied with worldly satisfactions. Let’s put the likelihood that we should long for the infinite given theism at 0.9. Now, what is the likelihood that infinite longings should arise on background knowledge alone? This is the hard part. If we don’t know whether theism is true or not, how likely are we to have desires like [the argument from desire] is talking about? I wouldn’t say that such desires couldn’t possibly arise in an atheistic world. Even though such desires seem to have limited evolutionary use, they could well be by-products of features of human existence that do. But how likely would they arise in such a world? So long as the answer is “less likely than in a theistic world,” the presence of these desires confirms theism. Let’s say that, if we don’t know whether theism is true or not, the likelihood that these desires should arise is 0.7. Plugging these values into Bayes’ theorem, we go from 0.5 likelihood that theism is true to a 0.643 likelihood that theism is true. Thus, if these figures are correct, the argument from desire confirms theism (Reppert, 2006).

J. P. Moreland rounds out the abductive argument by appealing to “the findings of empirical psychology” (Moreland, 2009, pp. 94-95). He observes that: “Contrary to what you might think, power (for example, money, education, and success), health, sexual attractiveness, and being youthful are not all that important as factors conducive to happiness and human flourishing. This important finding runs counter to what we should expect if the naturalistic evolutionary story were true” (Moreland, 2009, p. 95). Moreland proceeds to list the factors that are in fact conducive to happiness and human flourishing:

Pro: A Defense of C. S. Lewis’s Argument from Desire 37

At the top of [the] list is living in a constant spirit of gratitude—a sense of thankfulness in life. Next were unselfishly caring for and helping others, learning to give and receive forgiveness, and finding a deep and real sense of meaning and purpose in life by giving oneself to a larger framework than the individual’s own life (Moreland, 2009, p. 95).

Moreland adds:

Two things stand out about this list. First, they are exactly what one would expect if the biblical account is true. So these findings provide some confirming data for the biblical account. Second, several items on the list are absolutely absurd and irrational on the naturalistic evolutionary account, but they make rational sense on the biblical account (Moreland, 2009, p. 95).

If we experience a desire that remains unsatisfied by worldly objects but is partially satisfied by an apparently transcendent object of desire (just as a large thirst is partially satisfied by a small drink)—an object which coherently explains the desire in question, its arousal by worldly triggers, the failure of worldly goods to satisfy it and the partial nature of the satisfaction available here and now—then we are prima facie justified (via the principle of credulity) in concluding that this apparent object is indeed the real object of our desire.

B. The Inferential Argument

In Mere Christianity, Lewis argues:

Creatures are not born with desires unless satisfaction for those desires exists. A baby feels hunger: well, there is such a thing as food. A duckling wants to swim: well, there is such a thing as water. Men feel sexual desire: well, there is such a thing as sex. If I find in myself a desire which no experience in this world can satisfy, the most probable explanation is that I was made for another world. If none of my earthly pleasures satisfy it, that does not prove that the universe is a fraud. Probably earthly pleasures were never meant to satisfy it, but only to arouse it, to suggest the real thing (Lewis, 1997, p. 113).

Lewis begins with the premise that creatures (probably) don’t have natural or innate desires (desires they are “born with”) unless satisfaction for those desires exists. According to Thomas V. Morris:

38 PETER S. WILLIAMS

An innate desire, roughly, is a desire we seem to be born with, or at least born with the natural tendency to develop. An innate desire spontaneously appears in a person without having to be suggested, planted, coached or acquired. Innate desires tend to be universal, unless their development has been somehow impeded. An artificial, or learned desire is not one that is spontaneous, or even remotely universal (Morris, 1999, p. 222).

Peter Kreeft further explains the difference between natural and learned desires:

We naturally desire things like food, drink, sex, knowledge, friendship, and beauty, and we naturally turn away from things like starvation, ignorance, loneliness, and ugliness. We also desire things like Rolls Royces, political offices, flying through the air like Superman, a Red Sox world championship, and lands like OZ. But there are two differences between the two lists. First, we do not always recognize corresponding states of deprivation of the second, as we do the first. And, most important, the first list of desires all come from within, from our nature, while the second come from without, from society, or advertising, or fiction (Kreeft, 1989, p. 202).

John Haldane enumerates “three significant markers” by which we can identify “natural desires”: “1) spontaneity of occurrence; 2) prevalence . . . and, 3) common linguistic identification of types of desire and/or of satisfaction, and/or of their deprivation. . . . By contrast, artificial desires are 1) acquired or inculcated; 2) liable to be non-universal and culturally variant; and 3) not ubiquitously linguistically represented” (Haldane, 2007, p. 49). Haldane concludes:

In these terms the desire for transcendence has a good claim to be a natural desire. . . . [T]he impulse to seek transcendent fulfilment is testified to in a whole range of cultures and correspondingly is detectable in language. . . . Beyond the anthropological and linguistic evidence for this there is also personal testimony and phenomenological reflection (Haldane, 2007, p. 49).

Lewis’s examples give inferential support for the claim that natural desires have corresponding satisfactions. As Thomas Chalmers observed: “For every desire or every faculty, whether in man or in the inferior animals, there seems a counterpart object in external nature. . . . There seems not one affection in the living creature, which is not met by a counterpart” (Chalmers, p. 230, quoted by Hammond, 1943, p. 56). Gary Habermas and J. P. Moreland add: “There is, in fact, a very large correlation between what appear to be Pro: A Defense of C. S. Lewis’s Argument from Desire 39 natural desires and the existence of objects that could potentially satisfy them. . . . [T]he majority of human experience indicates that this principle is true. Our natural desires are, in fact, good indicators that objects really exist to satisfy those desires” (Habermas and Moreland, 1998, p. 34). Thus, Lewis argues that if a natural desire cannot be satisfied by any object in the present world, it probably has a satisfaction obtainable in another world. As Tim Keller asks:

Doesn’t the unfulfillable longing evoked by beauty qualify as an innate desire? We have a longing for joy, love and beauty that no amount or quality of food, sex, friendship or success can satisfy. We want something that nothing in this world can fulfill. Isn’t that at least a clue that this “something” that we want exists? (Keller, 2008, p. 135).

Trent Dougherty likewise presents the AFD as “a defeasible inference [in which] the premises could be true and the conclusion yet false, but they bear prima facie support for the conclusion.” Dougherty formulates the argument thus:

(1) Humans have by nature a desire for the transcendent. (2) Most natural desires are such that there exists some object capable of satisfying them. (3) There is probably something transcendent. (Dougherty)

Dougherty argues that “reasonable acceptance of 1 & 2 is very easy, so that one may reasonably accept 3” (Dougherty). Steven Lovell defines a desire as a longing for some state of affairs that satisfies the longing in question, such that “if the appropriation of some object, event or state of affairs does not satisfy a desire, then the desire was not a desire for that object, event or state” (Lovell, 2003). I would add the need to bear in mind the distinction between satisfying (in the sense of experientially “fitting” or “answering”) and satiating (in the sense of, at least temporarily, extinguishing) a desire. Lovell defines a “natural desire” as a desire that a creature has simply in virtue of being the kind of creature that it is, a desire that creatures of the relevant kind have, all things being equal, in “the usual course of events” (Lovell, 2003). Hence Lovell renders the first premise of the AFD as follows: “For any instantiated kind, K, and for any type of desire natural to that kind, it is consistent with the way the world is (or was at some earlier time) that a creature of kind K should (at some time) have a satisfied desire of that type” (Lovell, 2003). If we add a second premise that “human is an instantiated kind, and that a desire best interpreted as a desire for communion with God is more plausibly than not a desire natural to the human kind,” it follows that it is probably consistent with the way the world is (or was at some earlier time) 40 PETER S. WILLIAMS that a human being should (at some time or other) have a satisfied desire of that type. Since the satisfaction of a desire for communion with God requires the existence of a God with whom to commune, it follows that the proposition “God exists” must be true (at some time or other). Since God is plausibly not the sort of being that exists at one time but not at another, it would seem to follow that the proposition “God exists” is true simpliciter. Humans are certainly “an instantiated kind,” but how might we support the rest of premise 2? Lovell suggests that there are two basic ways of supporting it: “the experimental route and the metaphysical route” (Lovell, 2003). Taking the metaphysical route, we would first of all claim (with philosophers such as Aristotle, Boethius and Aquinas) that all humans naturally desire happiness, “and then attempt to unpack what happiness might (or must) consist in” (Lovell, 2003) to show that it consists in, or depends upon, fellowship with God. The argument of the experimental or empirical route is that “the desire for communion with God is very widespread; the best explanation of this fact is that the desire for God is a natural desire; therefore, the desire for God probably is a natural desire” (Lovell, 2003). The existence of a widespread desire for communion with God is evidence that the desire for such communion is a natural desire because “[i]f things of kind K normally possess [some feature] F simply because they are things of kind K, then one would expect things of kind K to possess F in a way that is largely insensitive to the environment in which they find themselves” (Lovell, 2003). Hence: “We should only infer that the desire for God is natural if humans in a sufficiently wide variety of circumstances exhibit that desire. The truth of this last claim is, of course, a matter for empirical investigation” (Lovell, 2003). Such an investigation reveals that

this desire is experienced by those of a religious persuasion and those of none; by those with and by those without a religious upbringing; by both men and women; by people of all historical periods; by both the healthy and the sick; by the rich and the poor; by people of all ages, races, and cultures; and by the educated and the uneducated alike (Lovell, 2003).

C. The Deductive Argument

In The Pilgrim’s Regress, Lewis says that the winding path of his own spiritual journey was determined by his pursuit of the proper object of his “Romantic” longing:

Pro: A Defense of C. S. Lewis’s Argument from Desire 41

It appeared to me therefore that if a man diligently followed this desire, pursuing the false objects until their falsity appeared and then resolutely abandoning them, he must come at last to the clear knowledge that the human soul was made to enjoy some object that is never fully given—nay, cannot even be imagined as given—in our present mode of subjective and spatio-temporal experience. This Desire was, in the soul, as the Siege Perilous in Arthur’s castle—the chair in which only one could sit. And if nature makes nothing in vain, the One who can sit in this chair must exist (Lewis, 1977a, p. 15).

Lewis sets out in the mode of an inductive argument by elimination, but his closing remarks suggest a deductive form of the AFD:

(1) Nature makes nothing (or at least no natural human desire) in vain. (2) Humans have a natural desire (Joy) that would be vain unless some object that is never fully given in my present mode of existence is obtainable by me in some future mode of existence. (3) Therefore, the object of this otherwise vain natural desire must exist and be obtainable in some future mode of existence.

From the premise that all natural desires have possible satisfactions we may conclude, via the premise that humans have a natural desire for communion with God, that God exists: “Once the meaning of these premises is made clear, it will be evident that the argument is a valid one, that if the premises are true then the conclusion must also be true” (Lovell, 2003). Atheist A. C. Grayling affirms that “nature does nothing in vain” (Grayling, 2011, p. 9), and as Peter Kreeft comments: “If nature makes nothing in vain, if you admit the premise, the conclusion necessarily follows. Of course, one who wants to refuse to admit the conclusion at all costs will deny the premise—but at the cost of a meaningful universe, a universe in which desires and satisfactions match” (Kreeft, 1989, p. 209). As Richard Purtill argues: “[I]t is possible to resist the pull towards God in our . . . experience . . . but if we do so we invalidate those experiences themselves” (Purtill, 2004, p. 39).

D. The Reductio Argument from Absurdity

Unless there is a more abundant life before mankind, this scheme of space and time is a bad joke beyond our understanding.—H. G. Wells

In “The Weight of Glory,” Lewis writes:

[T]hough I do not believe (I wish I did) that my desire for paradise proves that I shall enjoy it, I think it a pretty good indication that such a 42 PETER S. WILLIAMS

thing exists and that some men will. A man may love a woman and not win her; but it would be very odd if the phenomenon called “falling in love” occurred in a sexless world (Lewis, 2002, p. 98).

Lewis’s underlying assumption here is that the world is not “very odd” in the sense required by the rejection of the correlation of innate desires and objective satisfactions. As C. Stephen Evans argues:

The fact that people in general have a need for water is strong evidence that there is such a thing as water. . . . In a similar manner, the fact that we have a deep need to believe in and find God strongly suggests that God is real. . . . It would be very odd indeed if we had a fundamental need for something that did not exist (Evans, 1996, pp. 57–8).

Likewise, Leslie Weatherhead comments:

If I assert the fact of physical hunger—a desire for food—and deduce that, in a rational universe, that fact points to the fact of food, does my concept of food mean that food is necessarily a figment of my imagination and cannot exist? Assuredly not. If, then, I assert the fact of spiritual hunger—a desire for a fatherly God—and dream that there is a Satisfaction for it . . . does my dream deny the reality? As assuredly not. . . . In a rational universe, rather the reverse (Weatherhead, 1955, p. 404).

A universe in which natural desires and satisfactions do not match would not be a “rational universe” but a “very odd” or “absurd” one. According to the French existentialist Albert Camus, “absurdity springs from a comparison . . . between a bare fact and a certain reality, between an action and the world that transcends it. The absurd is essentially a divorce. It lies in neither of the elements compared; it is born of their confrontation” (Camus, 1959, pp. 22-23). An absurd universe is one in which there is a divorce between the natural desires of human beings and what the universe has to offer by way of satisfaction for those desires:

[I]f natural desires did not have correlating objects, there would be something fundamentally wrong, awry, disjointed, illogical, unfair, twisted, fraudulent, or out of kilter about reality. . . . We hardly need say that many people are indeed disillusioned or living under illusions and that this on its own does not make life absurd. However, life would be absurd if reality had “set us up” to be in this condition (Lovell, 2003).

As John Cottingham argues: Pro: A Defense of C. S. Lewis’s Argument from Desire 43

A natural yearning, implanted deep with us, may of course be vain, or fruitless, or have no object that can satisfy it; but that would be a tragic and ironic—perhaps even an absurdist—account of the human predicament. Such a view . . . was eloquently articulated by philosophers such as Jean-Paul Sartre and Albert Camus. The image of the human condition that Camus left us with is that of Sisyphus. . . . But the very structure of the story, the very labeling of his plight as “absurd,” implies a deep yearning for something more (Cottingham, 2009, p. 15).

We can deny that natural desires have corresponding possible satisfactions, but only “at the cost of a meaningful universe, a universe in which desires and satisfactions match” (Kreeft, 1989, p. 209). It seems to me that the belief that the universe is “meaningful” and not “irrational” or “absurd” is a properly basic belief and that anyone who professes to reject this belief will enter into a state of constant self-contradiction. As C. E. M. Joad argued:

[I]f the objects that satisfy religious need did not exist, it is hard to see how the existence of the need itself is to be explained. Granted [the Freudian theory] that man has made the belief in the objects affirmed by religion because he needs it, he did not create the circumstances in which he needs the belief, the circumstances which obliged him to invent it. Hence, the demand for a religious faith by human beings must be based on the same real nature of things as that which produced the human beings who made the demand. It is hard to credit in practice and it leads to self-contradiction in theory to suppose that nature has constituted man in such a way that he can only survive and prosper if he holds a belief in something which is not (Joad, 1952, p. 91).

One can treat the “argument from absurdity” as an auxiliary syllogism defending the first premise of the inferential or deductive versions of the AFD as follows:

(1) Either it is true that for any instantiated kind, K, and for any type of desire natural to that kind, it is (probably) consistent with the way the world is that a creature of kind K should (at some time) have a satisfied desire of that type, or life is ultimately irrational or absurd. (2) Life is (probably) not ultimately irrational or absurd. (3) Therefore, it is (probably) true that for any instantiated kind, K, and for any type of desire natural to that kind, it is (probably) consistent with the way the world is that a creature of kind K should (at some time) have a satisfied desire of that type. 44 PETER S. WILLIAMS

And so on. However, one can also deploy a reductio AFD as a theistic argument in its own right as follows:

(1) Life is ultimately irrational or absurd unless it is such that, for any instantiated kind K, and for any type of desire natural to that kind, it is consistent with the way the world is that a creature of kind K should (at some time) have a satisfied desire of that type. (2) Humans are an instantiated kind, K, that naturally desire communion with God. (3) Therefore, unless God exists, human life is ultimately irrational or absurd. (4) However, human life is not ultimately irrational or absurd. (5) Therefore, God exists (cf. Lovell, 2003).

Some may declare themselves willing to embrace cosmic absurdity; but that is not an easy affirmation to make or to sustain.

4. Conclusion

With Mark D. Linville, “I believe that the Argument from Desire has a rightful place within a comprehensive “cumulative case” argument for theism” (Linville, quoted in Puckett Jr., 2013, p. xii). The experience of Joy, described so powerfully by C. S. Lewis, can be seen as a natural sign that grounds the properly basic intuition of an innately desirable transcendent reality. Contemplation of Joy points us towards a theistic worldview. To reject the intuition of transcendence contradicts the properly basic belief that the world is not absurd. The existence of Joy abductively supports theism because it is more expected on a theistic than on a naturalistic worldview. Inductive and deductive forms of the AFD round out a cumulative case for the existence of a God in which our ultimate fulfillment lies. Here the AFD points to a plethora of desires the fulfillment of which plausibly require God’s existence (for example, desires for an afterlife, for objective meaning and purpose, the triumph of good over evil, for forgiveness and sanctification). In this way, as McGrath writes, “the deep and bitter-sweet longing for something that will satisfy us, points beyond the finite objects and finite persons . . . points through these objects, and persons towards their real goal and fulfillment in God himself” (McGrath, 1992, p. 55).

Two

CON: QUENCHING THE ARGUMENT FROM DESIRE

Gregory Bassham

I applaud Peter Williams’s perspicuous defense of Lewis’s argument from desire. Although Lewis is the best-known defender of the argument, he nowhere expounds the argument in detail, and he sometimes formulates it in different ways. Williams does a fine job teasing out various ways in which the argument might be developed. However, I shall argue that none of the versions Williams sets forth is successful. Williams claims that the argument from desire succeeds as part of a larger cumulative-case argument for the existence of God. Williams does not explain what he means by a cumulative-case argument, so it might be helpful to begin with a few clarifications. The term “cumulative-case argument” is widely used in contemporary apologetics to refer to inductive arguments that involve a series of evidence- building arguments that, taken together, are intended to make a conclusion rationally convincing. Cumulative-case arguments can be compared to a rope. In well-constructed cumulative-case arguments, each individual argument adds to the probability of the conclusion, just as each individual strand of a rope adds to the strength of the whole. Williams claims that Lewis’s argument from desire has “a rightful place” within a larger cumulative-case argument for theism, but he does not state how large that place is. Does the argument significantly contribute to the probability of theism, or only contribute a little? If the contribution is significant, how great is it? Does the argument by itself render theism probable? That is, is it a successful stand-alone argument for God’s existence? If so, does it make God’s existence barely probable (say 51 percent), or extremely probable (say ninety-five percent)? Moreover, given that Williams formulates and defends four different versions of Lewis’s argument from desire, how do those arguments relate to one another? Are they mutually consistent, such that each can figure in the same cumulative-case argument for theism, or must one pick and choose? Is each version equally forceful, or do some support God’s existence more strongly than the others? Perhaps Williams can clarify his position in his response.

46 GREGORY BASSHAM

1. The Prima Facie Argument

Williams develops his defense of Lewis’s argument from desire in five parts. First he lays out what he calls “the prima facie case” for taking the experience of Joy at face value. Then he sets forth four different versions of the argument from desire, suggesting that each provides some unspecified (but substantial?) degree of support for theism. Let us begin with the prima facie argument. Williams claims that there is an initial presumption in favor of treating Joy as a reliable pointer to God, and thus that the burden of proof lies on the skeptic to prove otherwise. This is supposed to follow from Richard Swinburne’s so-called principle of credulity. According to Swinburne, it is a basic principle of rationality that we ought to believe that things are as they seem to be unless we have positive reason for thinking that we are mistaken (Swinburne, 1979, pp. 254-271). Since Joy seems to point to a transcendent object of fulfillment, we should believe that such an object exists unless we have adequate evidence that such an object does not exist. Presumably, Williams calls this a “prima facie” argument from desire because it does not claim that Joy provides convincing evidence for the existence of God. Rather, it claims only that the burden of proof lies on the doubter to show that Joy is not what it appears. Two things should be said about this argument. First, it is not clear that the principle of credulity is true. Skeptics of various stripes would deny it. So, too, would devotees of religions such as Hinduism and some strands of Buddhism, who hold that phenomenal reality is maya, or illusion. But even if the principle of credulity is true, it cannot be invoked to support the argument from desire. As Swinburne notes, the principle of credulity applies only to “perceptual claims” (Swinburne, 1979, p. 260). What it claims is that “what one seems to perceive is probably so” (Swinburne, 1979, p. 254; emphasis added). As we shall see, what Lewis calls Joy is a certain kind of desire or emotion, not a perceptual experience. With Joy there is no apparent perception of God or Heaven, merely an emotionally-charged longing for what Lewis calls an “unnameable something” (Lewis, 1958, p. 9). Indeed, it is a defining feature of Joy that it has no clear or determinate object, but is easily confused with “false Florimels,” that is, delusive objects (Lewis, 1958, p. 8). If Joy is, in some secondary or extended sense, a perceptual experience, it is not an experience of a Being who “probably is” as “He seems.” Lewis would not claim that we have anything like an adequate grasp of God’s nature or essence. Our experience of God is of what Rudolf Otto called a tremendum mysterium (Lewis, 2001a, pp. 5-10). For these reasons, Swinburne’s principle of credulity does not support Williams’s view.

Con: Quenching the Argument from Desire 47

2. The Abductive Argument

We turn next to the first of four versions of the argument from desire that Williams presents. The arguments differ in terms of their argumentative rigor. One argument claims that its conclusion follows with strict deductive validity. The other three are more modest. The most modest is what Williams calls the abductive argument. Abductive arguments are arguments to the best explanation. Alister McGrath has recently claimed that this may be the best way to understand Lewis’s argument from desire (McGrath, 2014, p. 118). According to McGrath, Lewis’s argument (if it is an argument at all) is not intended to show that God’s existence is certain or even probable. Rather, it seeks to “confirm” theism in the sense of providing positive evidence that theism is true. The argument does this by showing that Joy is more likely to occur given theism than atheism. Joy “fits” better with a theistic picture of reality rather than with a materialistic one. The argument is thus “suppositional.” Its form is: Suppose there is such a being as God. Does not Joy fit in well with such a supposition? “And is not this resonance of supposal and observation indicative of the truth of the supposal?” (McGrath, 2014, p. 118). I see two problems with the abductive argument from desire. One is whether it is faithful to Lewis’s intent. The abductive argument, as McGrath formulates it, is extremely modest. It claims only that Joy provides some evidence for the truth of theism. Lewis’s formulations of the argument seem to indicate stronger conclusions. In one version, Lewis claims that the human experience of Joy provides “clear knowledge that the human soul was made to enjoy some object that is never fully given—nay, cannot even be imagined as given—in our present mode of subjective and spatio-temporal existence” (Lewis, 1958, p. 10; emphasis added). In other passages, Joy is presented as showing that God’s existence is inductively probable, not merely suppositionally “confirmed” (Lewis, 2001b, pp. 136-137; 1949, p. 6). McGrath is concerned to portray Lewis’s argument in the best possible light, but it is difficult to square his interpretation with the relevant texts. The second concern is whether the abductive argument can successfully show that theism is the best explanation of Joy. Inferences to the best explanation are standardly evaluated in terms of various “criteria of adequacy.” These include “simplicity, the number of assumptions made by a hypothesis; scope, the amount of diverse phenomena explained by the hypothesis; conservatism, how well the hypothesis fits with what we already know; and fruitfulness, the ability of a hypothesis to successfully predict novel phenomena” (Schick, 2005, p. 165). Lewis’s theistic hypothesis would seem to score poorly in terms of at least some of these criteria, notably simplicity, the number of assumptions required, and the fruitfulness of invoking supernatural explanations to explain psychological phenomena. 48 GREGORY BASSHAM

Competing naturalistic hypotheses would need to be considered and shown to be inadequate. What sorts of naturalistic hypotheses? Let me mention just one. Erik Wielenberg has suggested that Joy might be explained in terms of evolutionary psychology. Wielenberg notes two features of Joy—the restlessness it induces and the nebulousness of its object—that might explain its adaptive advantage. Early humans favored with a chronic, ill-defined restlessness of heart might have outcompeted other humans who were naturally more sedentary and complacent (and therefore less likely to take risks to seek greener pastures, to innovate, or to seek solutions to their problems). Second, Joy’s indeterminateness, its lack of a clear intentional object, might have led early humans down Lewisian “false paths,” such as the pursuit of sex, power, and adventure, that did have direct fitness advantages (Wielenberg, 2008, pp. 116-117). Wielenberg concedes that these possible explanations are highly conjectural. Nevertheless, they point to a real weakness in Lewis’s argument: his failure to consider seriously the possibility of alternative, naturalistic explanations of Joy. In sum, it is doubtful that Lewis offered an abductive version of the argument from desire, and it is unclear whether any supernaturalistic inference to the best explanation can succeed.

3. The Inferential Argument

Williams’s inferential argument is a version of what John Beversluis more aptly calls the inductive argument from desire (Beversluis, 2007, p. 42). (Inferential arguments can be either inductive or deductive.) According to the inferential argument, Joy provides a basis for a good (“cogent”) inductive argument for the conclusion that God probably exists. Williams quotes Trent Dougherty’s formulation of the argument, which runs as follows:

(1) Humans have by nature a desire for the transcendent. (2) Most natural desires are such that there exists some object capable of satisfying them. (3) Therefore, there is probably something transcendent. (Dougherty)

I want to raise several questions about this argument, to wit:

 Is Joy, in fact, a desire?  If Joy is a desire, can it be equated with “a desire for transcendence”?  Is Joy a natural desire?  Does Joy show that God probably exists, or at best only that “something transcendent” probably exists? Con: Quenching the Argument from Desire 49

 Can plausible alternative inductive arguments be formulated that suggest that Joy cannot be satisfied?

First, how clear is it that what Lewis calls Joy is a form of desire? As John Beversluis points out, according to Lewis Joy has no determinate object, is both intensely desirable and achingly painful, is experienced as a kind of bittersweet “stab” or “thrill,” and is accompanied by the same physiological sensations that go with feelings of “great and sudden anguish” (Beversluis, 2007, pp. 37-38). None of these are features usually associated with desires. In moments of intense aesthetic experience, it is common to feel a “stab” of soaring emotion (Lewis, 1984, p. 72, or an inconsolable “sob in the throat” (Lewis, 1958, p. 129), even in contexts where there is no apparent object of desire. Is it possible, Beversluis asks, that what Lewis calls a desire is better described as an undirected state of mind, like elation or anxiety (Beversluis, 2007, pp., 50-51)? In other words, might Joy be primarily an emotion or sensation and only secondarily, if at all, a desire? That is an issue that deserves more discussion than it receives. Even if Joy is at least sometimes a desire, is it necessarily a desire for something otherworldly? Nearly all commentators on Lewis’s argument from desire assume that Joy can be equated with various forms of “religious” desire—“spiritual longings,” “desires for transcendence,” “heavenly desires,” “desires for eternal happiness with God,” and so forth. Most traditional arguments from desire are framed in terms of such otherworldly desires (cf. Aquinas, 1956, p. 255; Hooker, 1845, pp. 201-202). But Lewis’s argument is different. Lewis is not talking about spiritual or transcendent longings in general. He is talking about one very specific form of emotion or desire that he later came to identify, rightly or wrongly, as a particular form of spiritual longing. Most otherworldly longings, such as a desire for eternal happiness with God in heaven, have relatively specific objects and are not brief, bittersweet “stabs” of intensely desirable, but also anguished, longings. Defenders of Lewis often spend a good many pages arguing that it is natural for humans to experience spiritual longings or desires for transcendence (cf. Puckett, 2012; Williams 2013, pp. 62-80). This may be true, but it misses the point. What needs to be shown is that the relatively rare, fleeting, and in some ways quite unusual sensations/emotions/desires that Lewis calls Joy are both natural and directed at a supernatural object of fulfillment. This is much more difficult to do than it is with generalized “spiritual longings” or “desires for transcendence.” Assuming that Joy is a desire, is it a natural desire in a sense necessary for the argument to succeed? Noting the particular examples Lewis cites (hunger, sex, a duckling’s desire to swim), John Beversluis argues that “natural desires,” for Lewis, are “biological, instinctive desires,” especially those that are “evolutionary adaptations that trigger appropriate responses to external stimuli and whose satisfaction is necessary for the survival of 50 GREGORY BASSHAM individual organisms and of the species of which they are members” (Beversluis, 2007, p. 45). Beversluis then proceeds to argue at length (1) that Joy is not an instinctive biological desire in human beings, since many people (including many “normal” people) never experience it (Beversluis, 2007, pp. 45-46, 52-54), and (2) that Joy is so different from the genuinely biological, instinctive desires Lewis cites (for example, hunger and sex,) that the argument commits the fallacy of weak analogy (Beversluis, 2007, pp. 44-45). This, in my judgment, is an effective criticism, and it exposes a real weakness in Lewis’s argument. In order for the inductive version of the argument from desire to succeed, Lewis needs to show, at minimum, that Joy is relevantly similar to other sorts of “natural” desires which we have good reason to believe have objects that can satisfy them. But clearly, as Beversluis notes, there are enormous differences between Joy and biological, instinctive desires such as hunger, sex, or a duckling’s desire to swim. Unlike the latter, Joy is neither (a) found in all normal members of the species, nor (b) readily explainable in naturalistic, evolutionary terms, nor (c) necessary for the survival of the organism and species, nor (d) straightforwardly “instinctive” or “innate,” nor (as Lewis claims) (e) satisfiable in the natural world. All of this, as Beversluis argues, makes for a weak argument from analogy. Defenders of Lewis might respond that the examples of hunger, sex, and so forth were purely illustrative, and that Lewis does not, in fact, limit “natural” desires to biological, instinctive urges. Perhaps, as Steven Lovell suggests, Lewis is thinking of “natural” human traits in a broader and hard-to- formulate sense—perhaps something like the following:

A certain feature, F, is natural in humans, if and only if, all normal humans possess F simply in virtue of being human (Lovell, 2003, p. 122).

Perhaps so. But then the argument fails on other grounds. Not only is it false that all “normal” humans experience Joy, but the definition is now so loose that all sorts of obviously unsatisfiable desires become natural. These include what we might call “fantasy” desires, that is, desires that, while not necessarily instinctive or innate, are so common or that arise so spontaneously in normal human beings as to be endemic to our species. Examples of natural fantasy desires include a desire to possess superhuman or magical powers, to travel to distant worlds, to know the future, to commune with nonhuman living things, to protect one’s children from heartache and harm, to remain youthful and unaffected by the ravages of time, to find an adoring and awesomely attractive mate or spouse, to be universally loved and admired, and so forth. All of these are “natural” desires in Lovell’s broad sense, shared by pretty much all normal human beings. So, too, might be the wishful desire to achieve ultimate felicity and eternal union with the summum bonum, God— or what William James calls “the craving of our nature for an ultimate peace Con: Quenching the Argument from Desire 51 behind all tempests, for a blue zenith above all clouds” (James 1967, p. 608). But there is no reason to think that all, or even most, natural fantasy desires have objects that can satisfy them. Why, then, should we think that our desire for ultimate happiness has an object that can satisfy it? Assume for a moment that these difficulties can be met and that Lewis’s argument succeeds. What precisely does it show? Many defenders of the argument claim that it provides good reason to affirm the existence of a theistic God and a Heavenly realm of perfect and eternal happiness. This conclusion, however, outruns the evidence. If humans have a natural desire for transcendence, and most natural desires have possible satisfactions, what follows is simply that humans’ desire for transcendence probably has a possible satisfaction. What form this transcendence takes is unclear. Might not Nirvana or absorption in a pantheistic One or even life in a Platonic realm of Forms do the trick? Only if we assume that Joy is a desire for perfect happiness with a perfect being (God) does the argument yield a theistic conclusion. To show this requires a further and perhaps more difficult argument—what’s more, one that Lewis does not provide. Finally, there are inherent drawbacks to inductive arguments like Williams’s inferential argument. As Erik Wielenberg notes, inductive arguments can support very different conclusions, depending on one’s background knowledge and the population sample one considers to be relevant. Joy, he notes, falls not only into the broad category of “natural desires,” but also into the narrower category of “human desires for things that are not part of the known natural universe.” In order for Lewis’s argument to succeed, Wielenberg contends, Lewis must show that all or most desires in this second class have objects that can satisfy them. And this cannot be done, Wielenberg notes, because “it is clear that the vast majority of those desires involve objects [for example, Zeus, the Elysian Fields, and Tolkien’s Valinor] that do not exist.” For all these reasons, I am skeptical that the inferential argument from desire can succeed.

4. The Deductive Argument

In Pilgrim’s Regress, Lewis offers what appears to be a deductive version of the argument from desire. Williams formulates the argument as follows:

(1) Nature makes nothing (or at least no natural human desire) in vain. (2) Humans have a natural desire (Joy) that would be vain unless some object that is never fully given in my present mode of existence is obtainable by me in some future mode of existence. (3) Therefore, the object of this otherwise vain natural desire must exist and be obtainable in some future mode of existence. 52 GREGORY BASSHAM

This argument is strongly reminiscent of Aquinas’s classic deductive version of the argument from desire. Aquinas writes:

[I]t is impossible for natural desire to be unfulfilled, since “nature does nothing in vain.” Now, natural desire would be in vain if it could never be fulfilled. Therefore, man’s natural desire is capable of fulfillment, but not in this life . . . . So, it must be fulfilled after this life. Therefore, man’s ultimate felicity comes after this life (Aquinas, 1956, p. 166).

In this argument, the Aristotelian dictum that nature does nothing in vain (Aristotle, 1948, p. 7) is offered as support for the claim that every natural desire must be capable of fulfillment. Williams says almost nothing in defense of Aristotle’s principle, other than to suggest that existence would be absurd or meaningless if it is false. I shall address this bogeyman of absurdity below. How else might Aristotle’s principle be defended? Peter Kreeft has argued that the principle that nature does nothing in vain is rationally intuitable from the concept of nature (Kreeft, 1989, p. 269). As we shall see, however, there is ample evidence in contemporary biology that nature does sometimes act in vain. Moreover, Kreeft’s view would turn “nature does nothing in vain” into a conceptual truth. Since conceptual truths are purely formal, it could provide no evidence for the presumably empirical claim that all natural desires are capable of fulfillment. Williams hints at another way that the principle that nature does nothing in vain might be defended. Might it be, as Kreeft has elsewhere suggested (Kreeft and Tacelli, 1994, p. 78), a well-confirmed universal generalization, a kind of Humean law of nature, supported by a multitude of uniform and exceptionless observations? The problem with this strategy is that it runs head on into modern science. Evolutionary biology, for example, reveals to us a blind, jerry-rigged world of junk DNA, vestigial organs, repurposed and ill-functioning design structures (think of the Panda’s thumb or the human back), as well as myriads of creatures whose natural instincts are doomed to frustration (sterile mules, flightless birds that retain their instinctive desire to fly, innate desires resulting from mutations that are retained long after changing environments have made them unsatisfiable, and so on). From the vantage point of modern science, nature frequently acts “in vain.” The key premise of the Williams’s deductive argument is that nature makes nothing (or at least no natural human desire) in vain. Since there seems to be no convincing way to argue for this claim, and there are ample reasons for being skeptical of it, the deductive argument is unconvincing.

Con: Quenching the Argument from Desire 53

5. The Reductio Argument

The last version of Lewis’s argument that Williams presents is dubbed the reductio argument. This version is indebted to Steve Lovell, a talented British philosopher who defends the argument in his unpublished 2003 PhD dissertation, Philosophical Themes in C. S. Lewis. The gist of the argument is that the universe would be meaningless or absurd if some natural desires were unfulfillable. But clearly the universe is not meaningless or absurd. In fact, Williams argues, the belief that the universe is meaningful is a properly basic belief. That is, it is a foundational belief that we are rationally warranted in accepting without any evidence or reasons at all. This properly basic belief entails that no natural beliefs lack possible satisfactions. Since the former is rationally warranted, and since the latter follows by self-evident entailment from the former, it follows that the latter is also rationally warranted. Q.E.D. This argument is faulty for two reasons. First, it does not follow from the fact that some natural desires are unfulfillable that the universe is meaningless or absurd. Camus famously argued that human existence is absurd because we find ourselves in a Godless universe that systematically frustrates basic human “needs,” such as our desire for immortality and ultimate meaning (Camus, 1959, pp. 12-24). As many writers have pointed out, Camus’s rhetoric is greatly overblown (Nagel, 1970, p. 226; Feinberg, 1994, pp. 328-329). Even if God does not exist, life is not aptly described as “absurd” because our “needs” for cosmic purpose and eternal happiness are doomed to frustration. Nor is it plausible to claim, as Williams does, that life is meaningless if God does not exist. As I have argued elsewhere, meaning is essentially connected to the notions of value, significance, and purpose (Bassham, 2013, p. 7). Each of these things can still be found in a Godless universe. To suggest, as Williams apparently does, that the universe would be meaningless and absurd if even a single animal species had an unfulfillable natural desire strikes me as seriously hyperbolical. Second, belief that the universe is “meaningful,” in Williams’s loaded sense of that term, is not a viable candidate for being a properly basic belief. As Plantinga notes, a belief can be shown to be not properly basic if there is a “defeater” that shows the belief to be irrational and hence unwarranted (Plantinga, 2000, p. 358). As we have seen, biology reveals many instances in which natural instincts and desires are, or can be, doomed to frustration. These serve as defeaters for William’s putatively properly basic belief. Williams’ appeal to Plantinga’s account of proper basicality leads to a circularity in his reasoning. For Plantinga, epistemic warrant is a matter of proper cognitive functioning, and our minds function properly when they work as God has designed them to work. This means that belief in God is plausibly regarded as properly basic only if God exists. Williams overlooks this and tries to argue for God’s existence by employing premises that 54 GREGORY BASSHAM implicitly presuppose God’s existence. The circularity can be brought out by means of this imaginary dialogue:

W: God exists. A: Why think that? W: Because humans have a natural desire for transcendence, and no natural desires can be unfulfillable. A: Why think that no natural desires can be unfulfillable? W: Because the universe would be meaningless if natural desires could be unfulfilled, and the universe is not meaningless. A: How do you know that the universe is not meaningless? W: Because belief that the universe is meaningful is properly basic. A: Why think that it is properly basic? W: Because a belief is properly basic if it is accepted as basic in a properly functioning mind, and our minds function properly when they accept as basic the belief that the universe is meaningful. A: Why think that our minds function properly when they accept such a belief? W: Because God has designed our minds to function in such a way.

The circularity of this logic is patent. Williams seeks to prove God’s existence by means of premises that no one would accept who did not already believe the conclusion. Williams gestures at another possible way of arguing that belief in the meaningfulness of life is properly basic. Those who refuse to accept the meaningfulness of existence face “constant self-contradiction.” Williams does not elaborate on this claim, so it is difficult to know what he means. It is true, as R. M. Hare has pointed out, that those who claim that “nothing matters” almost inevitably reveal in their actions that they do not truly believe this (Hare, 1972, p. 48). Anyone who gets out of bed in the morning presumably believes that doing so “matters,” that is, has at least some small value. But this is very different from Williams’s extremely robust conception of what a meaningful life entails. On his view, life lacks meaning and is “irrational” and “absurd” if God does not exist or if even one type of natural desire is impossible to fulfill. There is no reason to think that belief in this ultimate form of meaningfulness must not only be believed, but taken as properly basic. In short, the reductio argument fails for two reasons. First, it is a serious exaggeration to say that the universe is “absurd” or “meaningless” if not all natural desires can be fulfilled. Second, the argument will be convincing only to those who already accept its conclusion, and so is question-begging.

Con: Quenching the Argument from Desire 55

6. Conclusion

I have argued that Williams’s prima facie case for Joy fails, as do his four versions of the argument from desire. The best argument of the lot is the abductive argument, which in its most modest form claims only that Joy provides some confirming evidence for theism. This may be true, but is not anything to write home about. After all, many features of this confusing world might be understood as pointing to the existence of God, though many do not. Lewis’s argument from desire may contribute a small, brightly colored stone to this large and complex mosaic. But it is, at best, a small stone.

Three

REPLY TO GREGORY BASSHAM

Peter S. Williams

I thank Dr. Bassham for his paper and will respond to his objections while reviewing the sub-arguments of the argument from desire (AFD) in what I take to be a rough order of ascending strength.

1. The Prima Facie Argument

Clark Pinnock writes that “our deep desire to find fulfilment in a realm beyond the material . . . is a basic human intuition” (Pinnock, 1980, p. 53). The prima facie AFD justifies trusting this intuition by appealing to the wisdom of treating basic beliefs as properly basic (that is, rational) in the absence of a sufficient defeater. On this point I invoked Swinburne’s epistemic “principle of credulity” (PC): “that we ought to believe that things are as they seem to be (in the epistemic sense) unless and until we have evidence that we are mistaken” (Swinburne, 2010, p. 115). Bassham objects that “it is not clear the principle of credulity is true”; but rather than arguing the point he merely reports that “[s]keptics of various stripes would deny it.” If these skeptics have any reasons for their skepticism, Bassham does not say. Instead, he mentions that Hindus think phenomenal reality illusory. To claim that all phenomenal reality is illusory is to admit it has the prima facie appearance of reality. Swinburne’s PC allows that particular appearances of reality might be overwhelmed by sufficient evidence to the contrary, but since this necessarily means trusting some other particular appearances of reality, the rejection of phenomenal reality per se is incompatible with the PC. However, Bassham does not argue for the blanket rejection of phenomenal reality, whereas Swinburne argues that the PC is a basic principle of rationality: “If you say the contrary—never trust appearances until it is proved that they are reliable—you will never have any beliefs at all. For what would show that appearances are reliable, except more appearances?” (Swinburne, 2010, p. 115). Swinburne concludes: “It is basic to human knowledge of the world that we believe things are as they seem to be in the absence of positive evidence to the contrary” (Swinburne, 2010, p. 116). Bassham claims the PC cannot support the AFD because it only applies “to perceptual claims” and Joy is “a certain kind of desire or emotion, not a perceptual experience.” While the claim that a desire cannot be a perceptual 58 PETER S. WILLIAMS experience is controversial (cf. Oddie, 2005), Swinburne does not restrict application of the PC to perceptual claims. He explains that reference to “the epistemic sense” in the PC “describes how we are inclined to believe that things are” (p. 114). The PC is a general principle affirming the rational priority of trust. Swinburne notes that the principle applies to epistemic appearances besides those gathered via “your ordinary senses (for instance, your sense of sight)” (Swinburne, 2010, p. 115). For example, he applies the principle to “relying on your own memory” (Swinburne, 2010, p. 115). Moreover, Swinburne applies the PC to trusting inclinations towards religious beliefs prompted by a wide variety of experiences: “some people become apparently aware of God by hearing a voice or feeling a strange feeling, or indeed just seeing the night sky. But occasionally perceptions do not involve any sensory element at all . . . one just becomes aware that something is so” (Swinburne, 2010, pp. 119-120). Many philosophers acknowledge the wisdom of a general bias towards basic beliefs. C. Stephen Layman argues for what he calls the Starting Principle: “It is rational to accept what seems to be so unless special reasons apply” (Layman, 2007, p. 45). Rik Peels defends the principle of Prima Facie Rationality (PFR): “A cognitive subject S’s basic belief that p is prima facie rational iff it seems possible for all we know that S’s belief that p is produced by a reliable cognitive mechanism that aims at truth” (Peels, 2014, p. 3). He notes that via PFR “virtually all of our basic beliefs are prima facie rational” (Peels, 2014, p. 3). In other words, if a basic belief seems to one to be within one’s intellectual rights to accept, then one is prima facie justified in accepting it as a properly basic belief. Such prima facie rationality attaches to Joy. Bassham thinks the PC cannot support the AFD because “Lewis would not claim that we have anything like an adequate grasp of God’s nature or essence. Our experience of God is of what Rudolf Otto called a tremendum mysterium.” Again: “With Joy there is no apparent perception of God or Heaven, merely an emotionally charged longing.” Bassham’s objection seems to be that the PC cannot attach to propositions lacking some minimum specificity that the concept of God lacks if we limit ourselves to the information provided by religious experience. It is true that Joy is not the sort of direct, positive and specific experience of God or Heaven that can feature in direct-perception forms of the argument from religious experience (cf. Alston, 1991; Moreland, 1987; Swinburne, 2010). Nevertheless, Joy is more specific than Bassham’s “emotionally charged longing for what Lewis calls an ‘unnameable something.’” The “unnameable” is not necessarily the wholly unspecified. While Lewis would agree with Rudolf Otto about God being a mysterium tremendum et fascinans, and while he would deny we have a comprehensive grasp of God’s nature, he clearly thought we had a theological understanding “adequate” for reference to, argument about and relationship with God (cf. Lewis, 1977b). As Moreland notes: “One can know the truth Reply to Gregory Bassham 59 about something without having exhaustive knowledge of that thing” (Moreland and Nielson, 1993, p. 239). Nor is the object of Joy essentially “unnameable.” After his conversion Lewis associated Joy with the heavenly new “Jerusalem,” having discovered that his positive religious experience met with and partially satisfied (and thus confirmed) his Romantic longings (cf. Lewis, 1998a, p. 185). Joy is a by-product occasioned by various immanent objects of experience without being a desire for any of those objects, which can thus be temporarily confused with the desire’s transcendent object by the “inexperienced” (Lewis, 1977a, p. 8). As Lewis writes:

[T]he Desire itself contains the corrective of all these errors. The only fatal error was to pretend that you had passed from desire to fruition, when, in reality, you had found either nothing, or desire itself, or the satisfaction of some different desire. The dialectic of Desire, faithfully followed, would retrieve all mistakes, head you off from all false paths, and force you not to propound, but to live through, a sort of ontological proof (Lewis, 1977a, p. 10).

As the tantalizing experience of feeling drawn to an innately desirable object of desire that transcends all the immanent objects that occasion longing for it, Joy is more than a desire for, or feeling of “something, we know not what.” To take the phenomenology of Joy at face value is to look along Joy towards the inherently desirable, world-transcending other. Thus I deny that Joy lacks the minimum specificity required for being susceptible to the PC. After all, Swinburne himself applies the PC to the vague “general awareness of a power beyond ourselves guiding our lives (though not so much in a more detailed awareness of the exact nature of God and of his particular purposes)” (Swinburne, 2010, p. 115). The experience of Joy may not provide much metaphysical specificity, but it need not do so to contribute to the cumulative case for theism. Moreover, the contemplation of Joy steers us in a theistic direction. As Douglas Groothuis argues: “the argument from yearning renders credible some transcendent source of human satisfaction beyond the material world. . . . [I]t points towards a theistic worldview since it is based on the claim that humans desire a transcendent reality that can satisfy the human person” (Groothuis, 2011, p. 370). Finally, the prima facie AFD appeals to a range of natural desires that point towards God. Lewis counted “the irrepressible thirst for immortality” as being among the contents of “natural desire” (Lewis, 1937, p. 409). Charles Taylor notes “our aspiration to separate ourselves from evil and chaos” (Taylor, 2007, p. 711). Anthony O’Hear speaks of “the perfection we long for in some other world” (O’Hear, 2001, p. 144). Roger Scruton notes “those ancient and ineradicable yearnings for something else—for a homecoming to our true community” (Scruton, 1999, p. 27), and argues that beauty “points us 60 PETER S. WILLIAMS beyond this world, to a ‘kingdom of ends’ in which our immortal longings . . . are finally answered” (Scruton, 2009, p. 175).

2. The Abductive Argument

Bassham complains that Alister McGrath’s “extremely modest” formulation of the abductive AFD, which claims only “that Joy provides some evidence for the truth of theism,” is not “faithful to Lewis’ intent,” since “Lewis’s formulations of the argument seem to indicate stronger conclusions.” In “The Weight of Glory,” Lewis claims the AFD provides at least “a pretty good indication” that Paradise is obtainable (Lewis, 2002, p. 99), whereas McGrath states that the AFD he formulates is merely “indicative of the truth of the supposal” (McGrath, 2014, p. 118), so Bassham appears to have a point. How significant the point is depends on whether one takes the task at hand as being (a) articulating and defending Lewis’s formulations of the AFD or (b) developing and defending formulations of the AFD inspired by Lewis. McGrath could fail in the former while succeeding in the latter. That said, McGrath writes elsewhere that the abductive AFD highlights “a strong resonance between theory and observation” (McGrath, 2012, p. 111), which sounds more like Lewis’s “pretty good indication.” That point aside, McGrath did not dominate my discussion of the abductive AFD. I also presented Victor Reppert’s Baysean treatment and J. P. Moreland’s discussion of flourishing. Bassham trades engagement with Reppert and Moreland for a discussion of various “criteria of adequacy” by which “[c]ompeting naturalistic hypotheses would need to be considered and shown to be inadequate.” The most important question here is whether naturalistic hypotheses surpass supernatural ones in terms of explanatory adequacy. Bassham chides Lewis for failing “to consider seriously the possibility of alternative, naturalistic explanations of Joy.” However, the attempt to explain away Joy naturalistically means taking “the Way of the Disillusioned ‘Sensible Man’” (Lewis, 1997, p. 112). Lewis might respond that naturalistic explanations for Joy, like naturalistic explanations of consciousness in general, are attempts to reductively explain away the prima facie phenomenological appearance of reality. He would likely caution us about the self-defeating nature of such a project taken to its logical terminus (cf. Lewis, 1991), and would point to the explanatory inadequacy of naturalism to account for human rationality (cf. Craig and Moreland, 2000; Reppert, 2003; Menuge, 2004) to counter any doubts about “the fruitfulness of invoking supernatural explanations to explain psychological phenomena.” Notwithstanding Reppert and Moreland, Bassham reckons that “Joy might be explained in terms of evolutionary psychology.” Before critiquing specifics, it is worth making some general points about naturalistic evolutionary psychology (NEP). First, NEP is a controversial research Reply to Gregory Bassham 61 program (cf. Menuge, 2004; Moreland, 2001 and 2002; Tallis, 2011). Stephen M. Downes reports: “There is a broad consensus among philosophers of science that evolutionary psychology is a deeply flawed enterprise” (Downes, 2014). Second, NEP presupposes the neo-Darwinian account of human evolution, although this is an increasingly embattled hypothesis (cf. Behe, 2006 and 2007; Gauger et al, 2012; Meyer, 2013; Nagel, 2012), and one that Lewis doubted (West, 2012). Third, NEP assumes metaphysical naturalism, a worldview whose simplicity begets explanatory inadequacy (cf. Craig and Moreland, 2009; Moreland and Craig, 2000). Indeed, Moreland argues that “a number of apparent facts about the nature of human persons are recalcitrant facts for naturalistic versions of evolutionary psychology” (Moreland, 2007, p. 115). For a specific NEP account of Joy, Bassham turns to Erik Wielenberg. However, Wielenberg’s explanation lacks explanatory scope, as it only engages with “two features of Joy—the restlessness it induces and the nebulousness of its object.” Wielenberg suggests the former feature “might” be advantageous if Joy arose: “Early humans favored with a chronic, ill- defined restlessness of heart might have outcompeted other humans who were naturally more sedentary and complacent.” Wielenberg’s “might” indicates a low degree of explanatory power. It “might” be the case that early humans afflicted with “a chronic, ill-defined restlessness of heart” would be out- competed by humans free from such existential ennui! Again, Wielenberg suggests the somewhat nebulous nature of Joy “might” be an advantageous if Joy arose: “Joy’s . . . lack of a clear intentional object, might have led early humans down Lewisian ‘false paths,’ such as the pursuit of sex, power, and adventure, that did have direct fitness advantages” (Bassham, summarizing Wielenberg, 2008, pp. 116-17.) Again, Wielenberg’s “might” does not inspire confidence. Indeed, Bassham and Wielenberg concede that “these possible explanations are highly conjectural.” Moreover, Wielenberg’s “false paths” lack simplicity compared to the hypothesis that the “direct fitness advantages” of “sex, power, and adventure” are sufficient unto themselves. When our early humans realized that neither sex, nor power, nor adventure satisfied Joy, would not that lessen the significance of those activities in their minds, thereby constituting an evolutionary disadvantage compared to creatures lacking the Joy desire? Finally, Wielenberg offers no explanation for the appearance of Joy in our gene-pool, only for its natural selection should it appear. As Joe Puckett Jr. writes: “If the only function of natural selection . . . is to blindly remove the species that does not have the needed mechanisms for survival, what has caused humans to look beyond the stars if we can live just fine within them? The fact that we do look beyond them seems . . . to support an “intelligent” (or goal-directed) form of evolution more so than blind (non- directed) evolution” (Puckett, 2013, p. 129). Indeed, when discussing the inductive AFD, Bassham admits Joy is not “readily explainable in naturalistic evolutionary terms.” As Reppert argues, “natural desires that are unfulfillable 62 PETER S. WILLIAMS on earth is precisely what you should expect . . . from the point of view of theism. I seriously doubt that we can do this from the point of view of naturalism, even if a half-way-decent-looking evolutionary explanation of how such desires could arise were forthcoming from the naturalist” (Reppert, 2006).

3. The Inductive Argument

Using criteria from Tom Morris, Peter Kreeft and John Haldane, I argued that Joy is a natural desire (Wielenberg admits this; see also Pruss, 2010). I also argued that the inherent teleology of Joy points beyond the natural world to the inherently desirable transcendent other. While raising many questions about Joy and the desire for the transcendent, Bassham ignores these arguments. Lewis observed that Joy is an unusual desire, which is why the AFD does not use analogy to establish that Joy is a natural desire. Beversluis’s description of Joy highlights some of these unusual qualities (Beversluis, 2007, pp. 37-38), but still sounds much like the desires to love and to be loved. The suggestion that Joy is “an undirected state of mind, like elation or anxiety” (Beversluis, 2007, pp. 50-51) simply ignores the teleological phenomenology of Joy. Bassham’s moments of “intense aesthetic experience” may well be (or include) episodes of Joy. He asks if Joy is not “primarily an emotion or sensation and only secondarily, if at all, a desire”; but even if Joy were only secondarily a desire, it would still be a desire. And while Lewis discusses Joy as an incognisant desire, he also identifies Joy as a desire for the new “Jerusalem” of Heaven (Lewis, 1998a, p. 185), and argues abductively from the “desire for Paradise” (Lewis, 2002, p. 99). Trying to limit natural (that is, innate) desires to “biological, instinctive desires” begs the question against natural desires having transcendent objects. But natural desires are not limited to “biological, instinctive urges.” Consider psychologist Abraham Maslow’s famous “hierarchy of needs.” We have biology-transcending natural desires for such things as objective truth, goodness, beauty, justice, forgiveness, meaning, purpose, libertarian freedom, liberty, love, and rationality (cf. Williams, 1999, 2004, 2013b and 2013c). Lewis clearly does not “limit ‘natural’ desires to biological, instinctive urges.” Lewis casts Joy as a natural desire for a transcendent object; in which case, Bassham says, the AFD “fails on other grounds” (thus abandoning his former objection). Those “other grounds” are the supposed lack of adequate criteria for distinguishing between natural desires and “‘fantasy’ desires . . . that, while not necessarily instinctive or innate, are so common or that arise so spontaneously in normal human beings as to be endemic to our species.” Bassham states that “there is no reason to think that all, or even most . . . fantasy desires have objects that can satisfy them.” As examples he lists desires to (1) possess superhuman or magical powers, (2) travel to distant Reply to Gregory Bassham 63 worlds, (3) know the future, (4) commune with nonhuman living things, (5) protect one’s children from heartache and harm, (6) remain youthful and unaffected by the ravages of time, (7) find an adoring and awesomely attractive mate or spouse, and (8) be universally loved and admired. Bassham elsewhere additionally mentions desires to (9) live to a ripe old age, (10) see all of one’s children live long, happy lives, and (11) be physically attractive. It is unlikely or even impossible that every token of these desires should be fulfilled. However, even naturalists admit it is possible to fulfil desires of type 2, 3, 4, 5, 7, 8, 9, 10 and 11 (men walked on the moon; science can predict the future; apes learn sign-language; it is possible to avoid heartache, to get an adoring and awesomely attractive mate, to be universally admired and to live to a ripe old age; and some folk are physically attractive). Moreover, many non-naturalists think people can possess superhuman and magical powers (cf. Keener, 2011, Appendix B). Hence Bassham’s examples suggest that most types of “fantasy desires” do have objects that can fulfil them, so the AFD need not worry overly about distinguishing between natural and fantasy desires. Wielenberg argues that in addition to the category of “natural desires,” Joy also falls into the “narrower category” of “human desires for things that are not part of the known natural universe.” Wielenberg states: “it is clear that the vast majority of those desires involve objects that do not exist.” Bassham provides the examples of Zeus, the Elysian Fields, and Tolkien’s Valinor. However, far from describing a “narrower category” than “natural desires,” this objection stacks the deck by substituting the category of “sub-fantasy desires for things that do not exist.” As for the sub-category of “natural desires for things that are not part of the known natural universe,” we should give such desires the benefit of the doubt (assuming that their satisfaction is not known to be impossible). After all, examples of such desires with known fulfilments abound. For example, the natural human desire to discover what lies over the next hill is an oft-fulfilled natural desire for something that is not part of the known natural universe. Alternatively, if the objection here actually involves the sub-category of “natural desires for things known not to be part of the natural universe” (as Bassham’s examples suggest), the objection begs the question in favor of naturalism. Besides, many natural (that is, innate) desires have known satisfactions that break the strictures of a naturalistic worldview (for example, desires for objective truth, goodness, beauty, justice, forgiveness, meaning, purpose, libertarian freedom, liberty, love, rationality, and so forth). Bassham notes that the inductive AFD requires a second step to determine more precisely the form that transcendent satisfaction takes. That is why Robert Holyer argues that if we assume that “Joy is a desire . . . for ultimate happiness,” it follows that “since ultimate happiness is not possible without God, Joy is in a sense also a desire for God” (Holyer, 1988, p. 66). Likewise, John Haldane suggests: “God is, or is an essential condition of the 64 PETER S. WILLIAMS satisfaction of the natural desire for transcendent fulfilment” (Haldane, 2007, p. 47). At the very least, both the experience of pursuing Joy through a “lived dialectic” of false worldly objects and the contemplation of Joy point away from naturalism (which cannot admit transcendent reality) and towards theism, although how far the argument will take us in this direction will depend upon which non-naturalistic worldviews we consider “live” options (cf. Geisler, 2013; Williams, 2013a).

4. The Deductive Argument

Lewis’s Preface to the third edition of The Pilgrim’s Regress suggests the following AFD:

(1) Nature makes nothing (or at least no natural human desire) in vain. (2) Humans have a natural desire (Joy) that would be vain unless some object that is never fully given in my present mode of existence is obtainable by me in some future mode of existence. (3) Therefore, the object of this otherwise vain natural desire must exist and be obtainable in some future mode of existence.

Bassham objects that Aristotle’s nothing-in-vain maxim

runs head on into modern science. Evolutionary biology, for example, reveals to us a blind, jerry-rigged world of junk DNA, vestigial organs, repurposed and ill-functioning design structures (think of the panda’s thumb or the human back), as well as myriads of creatures whose natural instincts are doomed to frustration (sterile mules, flightless birds that retain their instinctive desire to fly, innate desires resulting from mutations that are retained long after changing environments have made them unsatisfiable, and so on.) From the vantage point of modern science, nature frequently acts “in vain.”’

However, Aristotle is not as obviously wrong as Bassham thinks. For example, Bassham cites outdated claims about “vestigial organs” and “junk DNA.” J. Budziszewski comments:

Examples of superfluities popular in the previous generation include the appendix and the tonsils. These two are not so often mentioned in the present generation, as word has spread that they have functions after all. . . . But our time has come up with its own chief example: So- called “junk DNA,” nucleic acid sequences that do not code for proteins, which are presented as superfluous on the assumption that coding for proteins is the only thing DNA is for (Budziszewski). Reply to Gregory Bassham 65

Budziszewski reports that “numerous functions have already been discovered for so-called junk DNA; its dismissal as junk turns out to have been embarrassingly premature” (Budziszewski; cf. Wells, 2011; Luskin, 2012.) Bassham employs purported examples of bad design (for example, the panda’s thumb) from the (questionable) argument from dis-teleology (cf. Nelson, 1997) as if they counted against Aristotle’s dictum. However, to achieve something in a manner falling short of some abstract ideal of design does not mean failing to achieve anything. As biochemist Fazale Rana observes, engineers designing complex material systems face trade-offs: “When confronted with trade-offs, the engineer carefully manages them in such a way as to achieve optimal performance for the system as a whole. And this overall efficiency can be accomplished only by intentionally sub- optimizing individual aspects of the system’s design” (Rana, 2008, p. 248). The human back may fall short of some abstract definition of design perfection, but it hardly exists “in vain”! C. S. Lewis had a problem with his thumb joints that prevented him from typing, but he could write with a pen. Ostriches cannot fly, but their wings are used for balance when running and to fan themselves cool. A study of the panda’s “thumb” in Nature concluded: “The radial sesamoid bone and the accessory carpal bone form a double pincer-like apparatus in the medial and lateral sides of the hand, respectively, enabling the panda to manipulate objects with great dexterity” (Endo, 1999, pp. 309-310). Lewis distinguished between the claim that each type of natural desire correlates with a real object of satisfaction and the claim that each token of natural desire will receive satisfaction:

A man’s physical hunger does not prove that that man will get any bread; he may die of starvation on a raft in the Atlantic. But surely a man’s hunger does prove that he comes from a race that repairs its body by eating and inhabits a world where eatable substances exist. In the same way, though I do not believe (I wish I did) that my desire for Paradise proves that I shall enjoy it, I think it a pretty good indication that such a thing exists and that some men will. A man may love a woman and not win her; but it would be very odd if the phenomenon called “falling in love” occurred in a sexless world (Lewis, 1998, p. 92).

Lewis’s appeal to Aristotle should thus be understood as the claim that “nature makes no type of thing in vain.” Indeed, Aristotle can hardly have formulated his maxim in ignorance of the fact that final causes can be frustrated in individual cases. If penguins dream of flying instead of swimming (cf. Elliotta, 2013), that desire truthfully indicates both that flightless birds “come from a race” 66 PETER S. WILLIAMS

(birds) that can fly and that flight is possible per se. It would be odd if the desire for flight occurred in a world without flight, just as it would be odd if eyes occurred in a world without light. Hence Lewis argues that types of natural desire have satisfactions, while acknowledging that tokens of natural desire may go unfulfilled (for instance, starving for want of food). Setting aside divine providence, Lewis’s distinction means it is possible (albeit unlikely) for every token of a natural desire type to go contingently unfulfilled (for example, all life on earth could die of starvation). Hence Lewis is over- reaching when he says that his “desire for Paradise proves that . . . some men will” enjoy it. What he should have said (and which is all he need say) is that the desire for Paradise proves that Paradise is attainable per se (and suggests that some people will enjoy it). Even if one thinks one or more counterexamples can be found to Aristotle’s maxim, one must still grapple with the generalization that “nature makes a minority of things in vain,” which is quite sufficient for securing the ceteris paribus conclusion that the object of Joy probably exists. Finally, as Bassham observes: “The key premise of Williams’s deductive argument is that nature makes nothing (or at least no natural human desire) in vain.” Bassham offers no objection to the premise that “nature makes no natural human desire in vain.” This premise seems well confirmed: “experience indicates that . . . natural desires are, in fact, good indicators that objects really exist to satisfy those desires” (Habermas and Moreland, 1998, p. 34). Thus, the deductive AFD can be restated as follows:

(1) Nature makes no type of natural human desire in vain. (2) Humans have a natural desire that would be vain unless some object that is never fully given in our present mode of existence is obtainable in some future mode of existence. (3) Therefore, the object of this otherwise vain natural desire must exist and be obtainable in some future mode of existence.

(One could obviously re-cast this argument in inferential terms.)

5. The Reductio Argument

Bassham’s point about the “hyperbolical” suggestion that the discovery of “even a single animal species” with “an unfulfillable natural desire” would render the universe meaningless or absurd is nicely put. One might nevertheless feel that in a thoroughly rational universe every type of natural desire would indicate a possible satisfaction; thereby agreeing with Victor Reppert that it is “analytic” to say in that sense that “[e]ither life is absurd or the world is such that all natural desires can be satisfied” (Reppert, 2006). Reply to Gregory Bassham 67

That said, a charitable reading of my former paper would note the repeated appeal to a plurality of existentially relevant human desires (desires for objective value, purpose, meaning, significance, and so forth) that reality either can or cannot satisfy. These desires transcend our “biological, instinctive urges” but underpin the possibility of human flourishing. At the very least, the more of these existential desires we believed to be in vain, the more “odd” (Lewis, 2002, p. 99) or “absurd” we would think our existence. Thus:

(1) Given an instantiated kind K possessing existential desires natural to that kind, the existence of kind K would be absurd to the extent that it is impossible for any member of kind K to have those existential desires satisfied. (2) Humans are an instantiated kind, K, with existential desires natural to their kind that are [probably] impossible to satisfy unless God exists. (3) Therefore, unless God exists, the existence of kind K is [probably] absurd (at least to a substantial extent). (4) However, the existence of kind K is [probably] not absurd (at least, not to any substantial extent). (5) Therefore, God [probably] exists.

I think God is necessary for objective “value, significance and purpose.” There cannot be an objective purpose to human or cosmic existence without a creator (cf. Williams, 2004). Everything, including purpose, is objectively meaningless without objective value. Lewis observes that we desire people to conform themselves to the objective moral law and desire ourselves to be found innocent before that law (Lewis, 1997, pp. 3-7), before arguing that this transcendent law is rooted in God (Lewis, 1997, pp. 18-22). The desire to fulfill moral duty is absurd on naturalism but fits comfortably with theism (cf. Mavrodes, 1986). In the final analysis we must choose between theism and nihilism (cf. Craig, 2008; Craig’s response to Nagel in Craig and Gorra, 2013; Moreland, 1987; Smith’s 2006 response to Feinberg; Williams, 1999 and 2004). Lewis thought belief in naturalism inevitably generates “disharmony between our hearts and Nature” (Lewis, 2013, p. 135). Francis Schaeffer explored the existential “line of despair” generated by reducing “the ‘mannishness’ of man” to the “nonpersonal sources” of the naturalistic worldview (Schaeffer, 1994, pp. 8, 24, 94). As Jeffery Gordon concludes, if the universe is without purpose “man is . . . a creature imbued with passions remarkably inappropriate to the universe is which he is immersed” (Gordon, 2013, p. 151). Faced with this “disharmony,” Tom Morris ponders: “Are our deepest yearnings and desires a good guide to the deepest truths about existence? Or could some of them be, by contrast, totally out of joint with reality?” (Morris, 1999, p. 223). The natural response to his question assumes 68 PETER S. WILLIAMS the priority of epistemic trust. If the satisfaction of our natural existential desires requires God, then the properly basic belief that life is not absurd places the burden of proof on the nihilist. Calling the intuition that life is not absurd “properly basic” does not mean assuming Alvin Plantinga’s account of warrant, let alone his theistic interpretation of that account (cf. Plantinga, 2000, pp. 154 and 351-352). To call a belief “properly basic” is simply to claim it is a rationally held basic belief: “beliefs based on memory, introspection, and perception . . . are normally perfectly rational, even though they are not based on arguments. Such beliefs . . . are properly basic” (Peels, 2014). According to the principle of credulity, if a basic belief seems to one to be within one’s intellectual rights to accept, one is prima facie justified in accepting it as such (that is, as “properly basic”). Moreover, the choice to trust that life is not absurd is the only meaningful and rational choice available (cf. Cottingham, 2003).

6. Conclusion

I think the AFD’s discussed above are all good arguments. The AFD argues “primarily for the existence of Heaven or ‘another world,’ and only secondarily for the existence of God” (McGrath, 2014, p. 105). Some AFDs (the abductive and reductio versions) invoke God more directly than others (the prima facie, inductive and deductive versions), and as with most theistic arguments, the amount of information revealed about God varies from case to case. Hence it is important to note that arguments from desire are mutually consistent, are more powerful when taken together, and most powerful when considered as part of the cumulative theistic case made by both traditional natural theology and a “ramified” (that is, distinctively Christian) natural theology.

Four

REPLY TO PETER S. WILLIAMS

Gregory Bassham

I thank Professor Williams for his thought-provoking reply. Williams’s response is lengthy, and I cannot, in the brief space permitted, reply to all the points he raises. Readers should not assume, however, that silence implies agreement.

1. The Prima Facie and Abductive Arguments

What Williams calls the prima facie version of the argument from desire (AFD) runs essentially as follows: As Richard Swinburne has argued, it is rational to believe that things are as they seem to be absent good reasons to doubt. Swinburne calls this “the principle of credulity.” The experience of Joy is an experience of what seems to us to be true. So the principle of credulity justifies a presumptive trust in the experience of Joy. I want to say three things about this argument. First, it is not clear that the principle of credulity is true. People often think that some mysterious entity is present (for example, ghosts, extraterrestrials, inner voices, and evil spirits) when, in fact, there is no sound reason to draw such a conclusion. Moreover, religious experiences are quite diverse. Some people, for example, experience God as a personal being, while others experience the divine as wholly impersonal. Both sorts of experience cannot be veridical. Finally, what of those, such as atheists, who experience reality as being devoid of any supernatural presences? (cf. Martin, 1986). Reality appears to them to be Godless. Could not the principle of credulity then be used, jiu-jitsu fashion, to argue for a presumption of atheism? Second, William misunderstands the principle of credulity. A careful reading of Swinburne’s various discussions of the principle makes clear that he applies it only to certain sorts of beliefs (namely, beliefs that some entity is present), not to desires. This makes good sense, for the principle is patently implausible if applied to desires. As Buddha noted, human existence is saturated with desire. Some desires are reasonable and realistically achievable, but many are not. Among my “basic” or “natural” desires are desires to travel through time, to remain forever young and vigorous, and to possess miraculous powers of healing and protection. None of these desires deserves any presumption of being obtainable. Indeed, any attempt to apply the principle of credulity to desires would amount to an open and completely indefensible invitation to wishful thinking. 70 GREGORY BASSHAM

Finally, Williams’s attempt to apply the principle of credulity to the experience of Joy depends on a sleight of hand. Joy is covertly treated, not as a desire for Paradise, but as a kind of glimpse or foretaste of it. In other words, Joy is treated as a cognitive religious experience: a way that “reality appears to us.” But then we are no longer talking about the AFD, but instead about a particular version of the argument from religious experience. Whatever cogency the argument has no longer depends on its conative aspect (the alleged fact that we naturally desire Paradise) but on its cognitive side— the fact that reality presents itself to us in a certain way. This is crucially different from Lewis’s version(s) of the AFD. Williams’s defense of the abductive version of the AFD is no more convincing. The abductive argument seeks to show that God is the best explanation of Joy. In my first response, I complained that Lewis moves too quickly to this conclusion and fails to consider alternative explanations, such as Wielenberg’s appeal to evolutionary psychology. Williams responds by denouncing evolutionary psychology as a fraud, and casting a dubious eye on evolutionary theory as a whole. This is not the place, obviously, to debate the merits of evolutionary theory. If Williams has serious doubts about evolution, all I can say is that we have more fundamental disagreements than I supposed. As for evolutionary psychology, nobody views it as a rock-hard science. It is, unavoidably, more speculative than most other branches of the sciences. But the theoretical foundations of evolutionary psychology are reasonably secure. If humans have innate desires, those desires must have an evolutionary basis. Somehow they must have conferred an adaptive advantage on our remote ancestors. Of course, we cannot travel into the past to verify with scientific exactitude what those advantages were. But it does not take a rocket scientist to understand how, say, instinctive aversions to snakes, high places, and incest would help our ancestors survive and successfully reproduce healthy offspring. Wielenberg’s proposed evolutionary explanations of Joy are, as he admits, highly speculative. My point in discussing them was simply to highlight a glaring gap in Lewis’s abductive AFD. If you are offering an inference to the best explanation, it is vital that you actually consider competing explanations. This is particularly true when your proposed explanation is supernaturalistic. To jump immediately to a supernaturalistic explanation, while blithely ignoring all competing naturalistic ones, is the very antithesis of the scientific spirit.

2. The Inductive Argument

The inductive (or “inferential”) version of the AFD runs briefly as follows: So far as we can determine, all (or at least the great majority) of natural desires Reply to Peter Williams 71 have possible satisfactions; Joy (a.k.a. our desire for Paradise) is a natural desire; so Joy probably has a possible satisfaction. In my previous response, I challenged this argument by citing examples of natural desires that do not appear to have possible satisfactions, such as a desire to know the future or to possess magical powers. Williams responds by arguing that at least most of my alleged natural desires do have possible satisfactions. For instance, science allows us to predict the future, and God can confer magical or miraculous powers on whomever he chooses. This is not an adequate response. The AFD is an attempt to prove the existence of God and/or heaven. It is blatantly question-begging, therefore, to presume the existence of God in defending the argument. Moreover, Williams twists my examples in ways that are uncharitable. Of course science often allows us to reliably predict the future. When I speak of “knowing the future,” as an example of a natural desire that does not appear to have a possible satisfaction, I mean of course knowing “in ways that are not naturally or scientifically possible.” Likewise with my examples of traveling to distant worlds, communing with nonhuman creatures, remaining youthful, and the like. We all naturally desire a “magical” way of instantly achieving our desires and dealing with life’s inevitable tragedies and adversities. Williams says nothing that addresses this point. Everyone has “fantasy desires” that are just as “natural” as Williams’s supposedly natural desires for ultimate justice, objective goodness, meaning, purpose, final forgiveness, libertarian freedom, and so forth. Yet we know that at least many of those desires cannot be satisfied (absent divine activity). Why should we think that our desire for infinite and eternal happiness (a.k.a. Joy) does not fall in the category of “apparently unsatisfiable fantasy desires”?

3. The Deductive Argument

Williams proposes two versions of the deductive AFD. One starts from the Aristotelian dictum that “nature does nothing in vain.” The other, more modestly, assumes that “nature makes no type of human desire in vain.” I have already, in effect, responded to the more modest version. I have pointed to many types of natural fantasy desires that do exist in “vain,” that is, that have no naturally possible satisfactions. So let me turn to the first version. In my previous response, I argued that Aristotle’s maxim conflicts with modern science. Pace Aristotle, we now know that many things in nature have no purpose or function, but are simply useless holdovers from an evolutionary process. This is true not only of certain body parts and genetic information, but also of some innate or instinctive desires. Nothing about the evolutionary process ensures that animals cannot have both types and tokens of instinctive desires that cannot be satisfied. An ostrich, for instance, might desire to fly (for example, to escape a predator) long after evolution has made this 72 GREGORY BASSHAM impossible. In Aristotle’s hierarchical, goal-directed universe, everything in nature has a purpose, nothing is useless, and everything is rationally ordered toward the good of the whole. The modern scientific view—to adapt William James’s memorable phrase—is that “[i]n the great boarding-house of nature, the cakes and the butter and the syrup seldom come out so even and leave the plates so clean” (James, 1956, p. 22). Williams gamely attempts to defend the older Aristotelian view by citing sources that claim that talk of vestigial organs and junk DNA is outdated. Two of the sources Williams cites (Wells and Luskin) are connected with the pro- creationism Discovery Institute, and the third (Budziszewski) is not a scientist, but a professor of government, who offers his critique of mainstream science in an online commentary on Thomas Aquinas’s thirteenth-century Treatise on Law! Here I can only invite readers to pick up any college biology textbook (for example, Raven and Johnson, 1999, p. 417) to read up on vestigial organs, or to go online and see how actual scientists (Luskin is a lawyer and a creationist) responded to the ENCODE junk DNA study (for example, Gregory, 2012). Serious discussion of Lewis’s merits as an apologist is not well-served by appeal to such sources. As a fallback position, Williams suggests that defenders of the deductive version of the AFD can always retreat to more modest versions of the argument. For instance, he suggests, they can claim either:

(1) Nature makes a minority of things in vain; so the object of Joy probably exists; or (2) Nature makes no type of natural human desire in vain.

Williams appears not to notice that (1) transforms the argument into an inductive (that is, probabilistic) AFD, and so is not a version of the deductive AFD at all. And (2), of course, ignores my claim that some types of natural desires (for example, desires to possess magical powers) have no natural objects that can satisfy them. But there is a deeper problem with (2). Williams claims that natural human desires include desires for things like objective moral values, meaning, and purpose. But these can exist, Williams argues, only if God exists. So it possible to know that no natural human desires are in vain only if one already knows that God exists. But the deductive AFD is an argument for God’s existence. Therefore, on Williams’s own terms, even the modest version of the deductive AFD that features premise (2) cannot be defended without circularity.

4. The Reductio Argument

In his first paper, Williams appeared to offer a version of the reductio AFD that runs briefly as follows: Life would be meaningless and absurd if any Reply to Peter Williams 73 natural desires are in vain. But life is not meaningless and absurd. So, no natural desires are in vain. In my response, I noted that Williams’s argument implies that life would be absurd and meaningless even if a single lowly organism had a single necessarily frustrated natural desire. But such a conclusion is obviously exaggerated and absurd. So Williams’s reductio can itself be demolished by means of a reductio. In his reply, Williams scales back his original argument and restates it so that it applies only to natural human “existential desires,” such as desires for meaning, purpose, significance, and objective moral values. The modified argument can be stated as follows:

(1) Human existence would be (substantially) meaningless and absurd unless our natural existential desires for objective meaning, values, purposes, and so forth can be satisfied. (2) Human existence is not (substantially) meaningless and absurd. (To deny this would itself be absurd.) (3) Objective meanings, values, purposes, and so forth can exist only if God exists. (4) So, God exists.

How does this slimmed-down version of the reductio AFD fare? Each of the three premises can be challenged. Existential nihilists of Camus’s stripe would deny (2). More seriously, substantial skeptical challenges could be mounted to both (1) and (3). In my first paper, I briefly explained why I find (1) and (3) unconvincing. I think it is obvious that, say, feeding a starving child has value and significance—and therefore meaning— even if it turns out that God does not exist. I think the burden of proof is on Williams to show otherwise. But in the short space remaining to me, let me briefly explain why Williams cannot defend (2), as he attempts, by claiming that it is “properly basic.” As Williams has it, for any person S and any belief p, p is prima facie properly basic for S just in case it seems to S that she is within her intellectual rights in accepting p without any supporting evidence. According to Williams, all (normal?) humans have a hard-wired basic “intuition” that life is not absurd. Human existence just “presents” itself to us, so to speak, as rich in objective meanings, values, and the like. But according to the principle of credulity, we are rationally justified in believing that things are as they seem unless we have sufficient reason for doubt. There are no sufficient reasons for thinking that life is absurd. So, belief that life is not absurd is properly basic for everyone. This argument invites scrutiny at many levels. Is it true that all (or most) humans “naturally intuit” reality as objectively meaningful, purposeful, and so forth? Is the principle of credulity true? Are there, in fact, compelling 74 GREGORY BASSHAM

“defeaters” for belief in objective meanings and values? But let me focus on a less obvious challenge to Williams’s argument: Is it true, as Williams claims, that if it “seems” to S that it is within his “intellectual rights” to accept p as properly basic, then it is presumptively rational for S to do so? Like Hamlet’s “perhaps,” Williams’s “seems” must give us pause. As I see it, Williams is much too quick to play the properly basic card. Are there objective truths, values, and beauty? Does life have objective meaning, purpose, and significance? Is there an afterlife in which ultimate happiness, justice, and forgiveness are attainable? Great thinkers have debated these questions for centuries. Williams’s reductio argument allows us to cut through all these thick debates and believe pretty much whatever we wish, without bothering with such nasty, inconvenient things as “facts” or “evidence.” If it “seems” to you that you are within your intellectual rights in believing some seemingly basic “intuition,” then chances are very good that you are. Oh happy day! I think there is a critically important distinction between it seeming that p is properly basic and p’s actually being properly basic. Like Benjamin Franklin’s elegant French lady, who was astonished to find that in any dispute she was always right (Kurland and Lerner, 1987), we all tend to overrate our personal capacity to home in on truth. Like Socrates, I like to think that God smiles when we exercise our brains and search earnestly for the wisdom we lack.

Five

PRO: THE ARGUMENT FROM REASON DEFENDED

Victor Reppert

Over the past century and a half, philosophical naturalism has certainly become more prevalent. It has been frequently hailed as the philosophical perspective that takes science more seriously than its rivals do. Its prevalence, it seems to me, is based to large extent on a certain mental picture. To use Daniel Dennett’s terminology, the idea is that science progresses by replacing skyhooks with cranes (Dennett, 1995, p. 73) Whereas a skyhook is a mind- first explanation, a crane is a bottom-up explanation that is, at its base, nonmental. Thus, rainbows were first thought to be the result of God’s telling Noah that he would not flood the earth anymore (a skyhook), but is now explained as the result of mindless light refraction (a crane). The greatest triumph of naturalistic explanation came with Darwin’s theory of evolution. There, in a domain in which nearly all educated people saw marks of intelligent design, Darwin was able to provide a bottom-up account of speciation through the trial and error of evolution by natural selection. Richard Dawkins famously said that while one could have been an atheist before Darwin, Darwin showed how to be an intellectually fulfilled atheist, freeing atheism from its most serious explanatory problem (Dawkins, 1986, p. 6). The successes of brain science and artificial intelligence show that our mental life can be fully treated in the same way. Instead of thinking of the mind as independent of matter, we can expect the mind to be explained as simply the activity of a part of the physical world we call the brain. The terms sometimes used to describe this type of position are naturalism, materialism, and physicalism. Naturalism is thought to hold that everything is part of nature, and thus rules out such entities as God while perhaps allowing such things as numbers or Platonic forms, whereas materialism or physicalism says that everything is matter, and that therefore there are no non-material entities such as forms. However, this distinction really does not make much of a difference to the argument from reason that I will be presenting here, because that argument concerns states of mind and how those states are caused. A fundamental question today is whether the trend described above is a permanent one, or whether there is a point at which reductive explanations will break down. One point that is often overlooked in all of this is that many successful scientific reductions have involved siphoning off certain features 76 VICTOR REPPERT from the physical to the mental. Consider the reduction of heat to the mean kinetic energy of molecules. As Richard Swinburne notes:

[T]his reduction has been achieved at the price of separating off the phenomenal from its causes, and only explaining the latter. All reduction from one science to another dealing with apparently very disparate properties has been achieved by this device of denying that the apparent properties (i. e. the “secondary qualities” of colour, heat, sound, taste, etc.) with which one science dealt belonged to the physical world at all. It siphoned them off to the world of the mental. But then, when you come to face the problem of the sensations themselves, you cannot do this. If you are to explain the sensations themselves, you cannot distinguish between them and their underlying causes and only explain the latter. In fact the enormous success of science in producing an integrated physico-chemistry has been achieved at the expense of separating off from the physical world colours, smells, and tastes, and regarding them as purely private sensory phenomena. The very success of science in achieving its vast integrations in physics and chemistry is the very thing which has made apparently impossible any final success in integrating the world of mind into the world of physics (Swinburne, 1997, p. 191).

Another important argument supports the claim that the reductive tendency is headed for a stop, and that is the argument from reason. The argument did not originate with C. S. Lewis by any means. Echoes of it occur in Plato, a version can be found in Kant, and it was developed in a form quite similar to that of Lewis by the British politician and philosopher Arthur Balfour. It was Lewis, however, who made the argument popular and who stated it in a new and forceful way. Versions of the argument occur in many of Lewis’s essays, as well as in his book Miracles: A Preliminary Study (1947, 2nd ed. 1960). The essence of the argument is that if naturalism is true, we could not know that naturalism is true, or that anything else is true, as a result of a train of reasoning. But enterprises such as the natural sciences presuppose that we can acquire knowledge through trains of reasoning, so to accept the sciences we have to deny naturalism, in spite of the fact that it is science that is often thought to support naturalism.

1. The Argument from Reason: Then and Now

My own development of Lewis’s argument against naturalism has resulted in some structural and terminological changes to the argument. Those familiar with current debates in philosophy of mind will readily understand why. The way contemporary philosophers use terms like “naturalism” and “valid” Pro: The Argument from Reason Defended 77 processes of thought has shifted over the years. As a consequence, a certain amount of updating is necessary to make Lewis’s central point relevant to philosophical discussion today. Lewis’s argument differs in an important way from other common arguments against naturalism. For example, some people argue that naturalism cannot explain the existence of consciousness or morality. At least one way of responding to these arguments is to suggest that morality has no objective reality, or to say that consciousness, at least as traditionally understood, does not exist. In my estimation the argument from reason has an advantage over those sorts of arguments in dealing with eliminativist or radically revisionist responses, in that common-sense understandings of the reasoning process seem to be presupposed by the scientific enterprise itself, and therefore the existence of reason as ordinarily understood must be explained in naturalistic terms, not denied or massively revised. Lewis frames his version of the argument from reason by distinguishing naturalism from a competing worldview he calls supernaturalism. In Miracles, he defines naturalism by looking at the way the word “natural” is used. He claims that what is natural is what goes of its own accord—“the given, what is there already: the spontaneous, the unintended, the unsolicited” (Lewis, 1978, p. 7). In Mere Christianity, Lewis distinguishes the materialist view, according to which space, time, and matter simply happen to exist, from the “religious” view, according to which “what is behind the universe is more like a mind than anything else we know” (Lewis, 2001, p. 22). These distinctions are helpful, but I believe that the difference between a naturalistic view and a nonnaturalistic one can be developed with even more precision by distinguishing between what we might call mentalistic and nonmentalistic worldviews.

2. Mentalistic and Nonmentalistic Worldviews

To begin, consider how we might explain why a sleeping pill puts me to sleep. To say that it has a “dormative virtue” seems insufficient; a further explanation in terms of its chemical content and how it reacts with the brain seems critical to the explanation. But explanations can go further, presumably to the “basic” level of matter and how fundamental particles behave, and certain basic law of physics are ultimately likely to be invoked. If we continue and ask why the particles of physics are the way they are and not some other way, we are likely to be told that this is an ultimate brute fact. Unless, of course, we think that some intelligent being chose the laws of physics to be as they are. So, either the “ultimate brute fact” is mental in nature, or it is not. Either we can say “In the beginning was the Word,” or “In the beginning, the Word was not.” 78 VICTOR REPPERT

But how do we distinguish between a mentalistic and a nonmentalistic view? There are four characteristics that seem to mark out the mental from other kinds of states. First, mental states are purposive. They, as Lewis puts it, prefer one thing to another. Second, mental states possess intentionality or about-ness. Mental states are ordinarily about something, whereas nonmental states usually are not. Third, some mental states are normative, and this goes along with preferring one thing to another. Finally, mental states have a perspective, a first-person point of view. A genuinely nonmental account of a state of affairs will leave out of account anything that indicates what it is like to be in that state. If the mind is not ultimate, then any explanation that is given in terms of any of these four characteristics must be given a further explanation in which these marks are washed out of the equation. The concepts of naturalism and materialism have evolved considerably since Lewis’s time. Following William Hasker (Hasker, 1999, pp. 58-64), I am convinced that a nonmentalistic, or broadly materialist, view of the world must possess three essential features. First, for a worldview to be materialistic, there must be a mechanistic base level. Now by mechanistic I do not mean necessarily deterministic. There can be brute chance at the basic level of reality in a mechanistic worldview. However, the level of what I will call “basic physics” is free of purpose, free of meaning or intentionality, free of normativity, and free of subjectivity. If one is operating within a materialistic framework, then one cannot attribute purpose to what happens at the basic level. Purpose-talk may be appropriate for macro-systems, but it is a purpose that is ultimately the product of a purposeless basic physics. What something means cannot be an element of reality as it appears at the most basic level. Third, there is nothing normative about basic physics. We can never say that some particle of matter is doing what it is doing because it ought to be doing that. Rocks in an avalanche do not go where they go because it would be a good idea to go there. Finally, basic physics is lacking in subjectivity. The basic elements of the universe have no “points of view” and no subjective experience. Consciousness, if it exists, must be a “macro” feature of basic elements massed together. Second, the basic level must be causally closed. That is, if a physical event has a cause at time t, then it has a physical cause at time t. That physical cause need not be a determining cause. But even if it is not, there cannot be something nonphysical that plays a role in producing a physical event. If you knew everything about the physical level (the laws and the facts) before an event occurred, you could add nothing to your ability to predict where the particles will be in the future by knowing anything outside of basic physics. Third, whatever is not physical, at least if it is in space and time, must supervene on the physical. Given the physical, everything else is a necessary consequence. In short, what the world is at bottom is a mindless system of events at the level of fundamental particles, and everything else that exists Pro: The Argument from Reason Defended 79 must exist in virtue of what is going on at that basic level. This understanding of a broadly materialist worldview is not a tendentiously defined form of reductionism; it is what most people who would regard themselves as being in the broadly materialist camp would agree with, a sort of “minimal materialism.” Not only that, but I maintain that any worldview that could reasonably be called “naturalistic” is going to have these features, and the difficulties that I will be advancing against a “broadly materialist” worldview thus defined will be a difficulty that will exist for any kind of naturalism that I can think of. At the same time, we must be careful. No planets are mentioned in basic physics, yet planets can and do exist as conglomerations of physical particles. In the case of planets, however, once we have enough information about the positions of the basic particles, the question is closed as to whether there is a planet is there. I have not ruled out by definition the possibility that, for example, reason cannot exist at a nonbasic level even though it is missing at the basic level. Instead, I will argue that a naturalistic world cannot contain reason. The argument from reason claims that reasoning, and in particular the kind of reasoning that is necessary for the success of science, cannot exist in a naturalistic world. Reasoning can only exist if some mentalistic worldview is true. Mentalistic views include traditional theism, but they also include such views as absolute idealism, which is the very position that Lewis himself adopted when Owen Barfield convinced him that the argument from reason was sound (Lewis, 1956, pp. 208-209). Thus the argument is not, strictly speaking, an argument for theism, but if it succeeds it makes theism more likely to be true, in virtue of the fact that it refutes some of the most popular alternatives to theism, namely broadly materialist metaphysical systems.

3. Transcendental Implications of Rational Inference

I have delineated the metaphysical commitments of those who deny that the mental is basic to the universe. At the same time, atheists like Richard Dawkins are not philosophical skeptics. They hold that there is genuine knowledge, discovered by science. They are scientific realists who believe that science discovers the truth about the way reality is. That is why they object, for example, to religious believers who hold to theism as true and thus informative about the nature of reality. They think that science has discovered that evolution is true and that creationism is false. They think physicists discover the truth, and that they also make correct mathematical inferences. They think we literally add, subtract, multiply, and divide, and take square roots of numbers. They think that their own statements refer to realities in the natural world. I maintain that if we believe that the sciences are true, then we are making the following nine assumptions. 80 VICTOR REPPERT

(1) States of mind have a relation to the world we call intentionality, or about-ness. (2) Thoughts and beliefs can be either true or false. (3) Human beings can be in the condition of accepting, rejecting, or suspending belief about propositions. (4) Logical laws exist. (5) Human beings are capable of apprehending logical laws. (6) The state of accepting the truth of a proposition plays a crucial causal role in the production of other beliefs, and the propositional content of mental states is relevant to this causal role. (7) The apprehension of logical laws plays a causal role in the acceptance of inferred conclusions as true. (8) In reasoning, the same individual entertains thoughts of the premises and then draws the conclusion. (9) Our processes of reasoning provide us with a systematically reliable way of understanding the world around us (cf. Reppert, 2003, p. 73).

These are some important epistemological commitments of contemporary naturalism, as opposed to its metaphysical commitments. I contend that there is a conflict between these epistemological commitments of naturalism and its metaphysical commitments. Unless all of these statements are true, as Lewis puts it, no science can be true. But how are all of these things even possible on a broadly materialist view?

4. The Basis Structure of Lewis’s Argument

Here is a formalization of Lewis’s argument as it appears in the revised edition of Miracles (1960). I have made an adjustment in it, because Lewis makes the claim that all knowledge, apart from our own sensations, is inferred from those sensations. What he means by this is not completely clear, because he says we do not begin as children by starting from our sensations and then inferring physical objects. However, when our perceptions are challenged, we do need to draw inferences to defend our perceptual beliefs. However, this is not really needed for the argument, since Lewis could just as easily grant that we perceive physical objects directly and still launch his core argument against naturalism. Moreover, the claim that all knowledge is based on inferences is inconsistent with Lewis’s frequently-asserted claim that we know some things (for example, self-evident moral truths) immediately, by a process of rational intuition. Here, then, is the adjusted version of Lewis’s argument from reason:

Pro: The Argument from Reason Defended 81

(1) No belief is rationally inferred if it can be fully explained in terms of nonrational causes. (2) If naturalism is true, then all beliefs can be fully explained in terms of nonrational causes. (3) Therefore, if naturalism is true, then no belief is rationally inferred. (4) If any thesis entails the conclusion that no belief is rationally inferred, then it should be rejected and its denial accepted. (5) Therefore, naturalism should be rejected and its denial accepted.

The main reason I put the argument in terms of inferred belief is that I want to put the focus on a certain type of rational justification that is essential to the methods of the natural sciences. Lewis had earlier put the argument in terms of “valid” thoughts, but this creates confusion in readers trained in contemporary logic. By formulating the argument as I have in terms of rational inference, I am attempting to avoid the problems I have noted and focus, instead, on what I take to be Lewis’s central point: that if rational inference goes down the drain, so too does science.

5. Anscombe’s Central Objection

In her famous Oxford debate with Lewis, Elizabeth Anscombe’s main objection to Lewis’s original argument was that it failed to distinguish between different senses of the terms “why,” “because,” and “explanation.” There are, she suggests, four explanation-types which are used in different contexts.

(1) Naturalistic causal explanations, typically subsuming the event in question under some physical law. (2) Logical explanations, showing the logical relationship between the premises and the conclusion. (3) Psychological explanations, explaining why a person believes what he or she does. (4) Personal history explanations, explaining how, as a matter of someone’s personal history, they came to hold a particular belief (Anscombe, 1981, pp. 224-229).

Anscombe suggests that explanations of different types can be mutually consistent. Thus a naturalistic causal explanation might offer a complete answer why a particular belief was formed, but that explanation might be compatible with an explanation of a different type (Anscombe, 1981, pp. 224- 229). Now what is interesting is that Lewis, in reformulating his own argument, not only draws the distinctions on which Anscombe had insisted; 82 VICTOR REPPERT he actually makes these distinctions the centerpiece of his revised argument. He makes a distinction between cause-and-effect relations on the one hand, and ground-and-consequent relations on the other. Cause-and-effect relations pertain to how a thought was produced, but ground-and-consequent relations indicate how thoughts are related to one another logically. However, to allow for rational inference, there must be a combination of ground-consequent and cause-effect relationships which, Lewis says, cannot exist if the world is as the naturalist says that it is. To claim that a thought has been rationally inferred is to assert something about how that thought was caused. Any face-saving account of how we come to hold beliefs by rational inference must maintain that “[o]ne thought can cause another not by being, but by being seen to be, a ground for it” (Lewis, 1978, p. 17). However, there are a number of features of thoughts as they occur in rational inference that set them apart from other beliefs. Lewis writes:

Acts of thinking are no doubt events; but they are a very special sort of events. They are “about” something other than themselves and can be true or false. Events in general are not “about” anything and cannot be true or false. . . . Hence acts of inference can, and must, be considered in two different lights. On the one hand they are subjective events, items in somebody’s psychological history. On the other hand, they are insights into, or knowings of, something other than themselves (Lewis, 1978, p. 17).

So here we already have three features of acts of thinking as they occur in rational inference. First, those thoughts have to be about something else. Second, they can be true or false. Third, their propositional contents must cause other thoughts to take place. But there is more:

What from the first point of view is the psychological transition from thought A to thought B, at some particular moment in some particular mind, is, from the thinker’s point of view a perception of an implication (if A, then B). When we are adopting the psychological point of view we may use the past tense. “B followed A in my thoughts.” But when we assert the implication we always use the present—“B follows from A.” If it ever “follows from” in the logical sense, it does so always. And we cannot possibly reject the second point of view as a subjective illusion without discrediting all human knowledge (Lewis, 1978, p. 17).

So now, in addition to the three features of thoughts as they occur in rational inference, we can add a fourth, namely, that the act of inference must be subsumed under a logical law. And the logical law according to which one Pro: The Argument from Reason Defended 83 thought follows another thought is true always. It is not local to any particular place or time; indeed laws of logic are true in all possible worlds. Lewis then argues that an act of knowing “is determined, in a sense, solely by what is known; we must know it to be thus because it is thus” (Lewis, 1978, p. 18). P’s being true somehow brings it about that we hold the belief that P is true. Ringing in my ears is a basis for knowing if it is caused by a ringing object; it is not knowledge if it is caused by tinnitus. As Lewis argues,

Anything that professes to explain our reasoning fully without introducing an act of knowing thus solely determined by what it knows, is really a theory that there is no reasoning. But this, as it seems, is what Naturalism is bound to do. It offers what professes to be a full account of our mental behaviour, but this account, on inspection, leaves no room for the acts of knowing or insight on which the whole value of our thinking, as a means to truth, depends (Lewis, 1978, p. 18).

I should note here that although I think there is a certain amount of sense behind the idea of an act of knowing which is determined by what it knows, my inclination is simply to appeal to the epistemological commitments of contemporary naturalism given above, and to argue that naturalism cannot meet those commitments using only the resources allowed by the metaphysical commitments of naturalism.

A. “Full” Explanations of Acts of Reasoning

Anscombe, in her brief response to Lewis’s revised argument, claimed that Lewis did not flesh out the concept of “full explanation” that he was using. Anscombe had expounded a “question relative” conception of what a “full explanation” is; a full explanation gives a person everything they want to know about something. John Beversluis, in the first edition of C. S. Lewis and the Search for Rational Religion, explicates this idea as follows, using the string quartets of Beethoven as his example:

Fully means “exhaustively” only from a particular point of view. Hence the psychologist who claims to have fully explicated the quartets from a psychological point of view is not open to the charge of self- contradiction if he announces his plans to attend a musicologist’s lecture on them. In music, as in psychology, the presence of non-rational causes does not preclude reasons. In fact, there is no limit to the number of explanations, both rational and non-rational, that can be given why Beethoven composed his string quartets. . . . All of these “fully explicate” the composition of his string quartets. But they are not 84 VICTOR REPPERT

mutually exclusive. They are not even in competition (Beversluis, 1985, pp. 73-74).

However, there do seem to be legitimate cases of explanatory exclusion. For example, if we explain the presence of presents under the Christmas tree in terms of the largesse of Santa Claus, we cannot at the same time explain them by saying that Mom and Dad bought them and put them there. Explaining disease in terms of germs excludes explaining them in terms of a voodoo curse. In particular, naturalism imposes limitations concerning what kinds of explanations can be given. Although explanations may satisfy an inquirer concerning how something might have occurred, any other relevant explanation that might be provided would have to employ only naturalistically acceptable ontological resources. In any event, the key elements of the argument I present are the three defining features of naturalistic views presented above. If the physical is closed, then an explanation for anything either has to be itself physical, or at least supervene on the physical. If the state of the physical is necessary and sufficient for all states, including mental states, and if the physical has no rational content, then what role can reasons play in the actual formation of beliefs? Because of the ambiguities connected with the idea of “full explanation,” my own development of the argument from reason has avoided talk about full explanations, but has instead focused on the ideas of mechanism, causal closure, and supervenience. Given these three doctrines, it seems that some kinds of explanations face the prospect of being ruled out. Even the most nonreductive forms of materialism maintain that there can be only one kind of causation in a physicalist world, and that is physical causation. It is not enough simply to point out that we can give different “full” explanations for the same event. Of course, we can. Nevertheless, given the causal closure thesis of materialism, there cannot be causal explanations that require nonmaterialist ontological commitments. The question that is still open is whether the kinds of mental explanations required for rational inference are compatible with the limitations placed on causal explanations by materialism. If not, then we are forced to choose between saying that there are rational inferences and accepting materialism. However, materialism is invariably presented as the logical conclusion of a rational argument. Therefore, we must choose to reject materialism.

6. The Argument from Intentionality

One of the transcendental implications of rational inference is the existence of states of mind that are about other things. If there is to be reasoning, it must be literally true that our thoughts have intentional content. As Lewis argues: Pro: The Argument from Reason Defended 85

We are compelled to admit between the thoughts of a terrestrial astronomer and the behaviour of matter several light-years away that particular relation which we call truth. But this relation has no meaning at all if we try to make it exist between the matter of the star and the astronomer’s brain, considered as a lump of matter. The brain may be in all sorts of relations to the star no doubt: it is in a spatial relation, and a time relation, and a quantitative relation. But to talk of one bit of matter as being true about another bit of matter seems to me to be nonsense (Lewis, 1967, p. 64).

What Lewis is talking about is what philosophers call intentionality, or the about-ness of our mental states. So a naturalist who wants to rebut Lewis’s argument has to show that states of mind that are about other things are possible, given naturalism. Now, all states, according to even minimal naturalism, have to be the necessary implications of physical states. But how is that even possible? Is there anything about the physical that guarantees that the mental states that we all believe ourselves to be in must obtain? It is hard to see what that could be. Can any conglomeration of nonintentional facts entail the existence of an intentional state? Some critics will maintain that this kind of argument invariably commits the fallacy of composition. Just because the mental is missing from the physical level does not mean that it cannot be present as a “system” fact about a set of particles. To see why Lewis’s intentionality argument does not commit the fallacy of composition, consider two sorts of examples about how parts and wholes can be related. Consider, first, the case of the size of a brick wall, based on the positions of the bricks. In the case of the wall, given the state of the bricks, the question is closed as to whether the wall is there, or how tall it is. Even though none of the bricks is six-feet-tall, they can be added up in such a way that the height of the wall is a determinate fact based on the sizes of the bricks, and the sizes of the bricks are determinate facts based on the sizes of the elementary particles that make them up. Contrast this with the case of whether a particular homicide was justified or unjustified. Here we can look at the homicide at every scientific level—the physical, the chemical, the biological, the psychological, and the sociological—and no entailment can be drawn as to whether the homicide was justified or unjustified. Something over and above the physical data must be brought in to make this kind of a judgment. Either there is some nonnatural fact that makes the statement concerning the rightness or wrongness of the homicide justified or unjustified, or the matter is a subjective matter, determined by the preferences of an individual or a society. We might express this difficulty in the following way. Suppose we are given a complete list of physical facts, facts about where all the particles are. 86 VICTOR REPPERT

The information provided is insufficient to determine a unique mental state that a person is in. There is no entailment relation of any kind to the relevant mental state. In virtue of what is some physical state about some other physical state? This is the familiar worry about intentionality, a worry made more difficult by my claim that the kind of intentional states involved in rational inference are states in which the content is understood by the agent and put into a propositional format. Is there a set of necessary and sufficient physical conditions which jointly entail the conclusion that agent A is in the state of believing, or doubting, or desiring, or fearing that proposition p is true? If the fact about what a person’s mental state is about does not follow from the state of the physical, then there is nothing else from which it can possibly follow. In the case of mental states, I do not see how the physical states can possibly “add up” to any determinate mental state. There is a qualitative difference between the physical base and mental content that no amount of investigation can possibly bridge. Surprisingly some influential naturalists, including Daniel Dennett (Dennett, 1987, p. 313), accept this line of argument. Could it be, then, that we can live without determinate mental content? Given naturalism’s commitment to the natural sciences, the naturalist must presuppose the existence of mathematicians as well as scientists. Therefore, some serious consequences follow from the indeterminacy of mental states. It would mean that what Dawkins means by atheism is indeterminate. It means that it is not literally true that Einstein developed his theories of relativity from Maxwell’s equations. In order for there to be rational inference, there has to be a fact of the matter as to what someone believes to be true. If the state of the nonmental substrate fails to guarantee this, then, on the naturalistic views we have been considering, then nothing does, and there can be no such facts. So, unless the state of the physical can guarantee, for example, that Einstein is in the mental state of concluding that relativity is true based on inferences from Maxwell’s equations, science, and therefore naturalism, is undermined. Naturalistic analyses of reasoning seem to commit one of two mistakes. One of those mistakes is giving a description of mental processes which in effect explains the mental away, because it de-mentalizes mental states. Eliminative materialists have that kind of problem. The other is to attribute states of mind to the brain, converting the brain into a person and implying that mentalistic accounts somehow are not mentalistic because they are attributed to a physical object, namely the brain. It is almost as if naturalists think they can say the magic word “brain” and naturalize any teleological function. That will not do. The use of brain-talk does not guarantee that the explanations that use brain-talk are genuinely naturalistic. To these types of arguments I am inclined to reply, “Interesting fellow Mr. Brain. Remarkable what he can do.”

Pro: The Argument from Reason Defended 87

7. The Argument from Mental Causation

For naturalists, the status of intentional states also creates a serious problem for mental causation. Although I have developed the argument from reason in various ways, one of the most important is the argument from mental causation. Recall for a moment Lewis’s discussion of how rationally inferred beliefs must be caused. He says, “One thought can cause another not by being, but by being seen to be, a ground for it” (Lewis, 1978, p. 17). So besides the existence of facts to think about, and our capacity to perceive a self-evident rule that permits the inference, we also must be able to arrange those facts to prove a conclusion, and it must be possible for new beliefs to be brought into existence by this kind of process of reasoning. To those who, like Anscombe, are inclined to think that reasons-explanations are always non-causal in nature, I would like to ask how we are to understand words like “convince” or “persuade.” Presumably, rational persuasion is the goal of argumentative discourse, but if reasons are in no sense causal in nature, that is impossible. Suppose we were to answer Lewis’s question about what logical grounds have to do with the actual occurrence of beliefs as psychological events by saying: "Nothing. Beliefs (if they exist at all given naturalism—of course this is denied by eliminativists) are strictly epiphenomenal. It seems to us that we hold beliefs for good reasons, but if we examine how those beliefs are produced and sustained, we find that reasons have nothing to do with it. We think they do, but this is just one more example of the 'user illusion.'” If we were to say that, it seems to me that the possibility of science as an operation would have to be called into question. As Jerry Fodor once put it:

If it isn't literally true that my wanting is causally responsible for my reaching, and my itching is causally responsible for my scratching, and my believing is causally responsible for my saying . . . if none of that is literally true, then practically everything I believe about anything is false and it's the end of the world (Fodor, 1990, p. 156).

Further, we have to look at just what is involved when we talk about causal transactions. Only some properties of an object are causually relevant to the production of the effect. For example, if I take the baseball that Luis Gonzalez hit to win the 2001 World Series for the Arizona Diamondbacks over the New York Yankees, and throw it at the window, it would break the window only in virtue of the force it applied to the window. It does not break the window in virtue of its having been the ball Gonzo hit against Mariano Rivera. When Lewis says “[o]ne thought can cause another not by being, but by being seen to be, a ground for it,” not only must one mental event cause another mental event, but it must do so in virtue of its propositional content, 88 VICTOR REPPERT and in fact, in virtue of the kinds of logical relationships that exist between the relevant propositions. There are a couple of arguments that have been developed to show that given the causal closure of the physical, rational inference is impossible. In William Hasker’s third chapter of The Emergent Self, entitled “Why the Physical Isn’t Closed,” Hasker uses a counterfactual argument to show that the kinds of counterfactuals involved in mental causation will turn out false if the physical is closed (Hasker, 1999, pp. 58-80). Let us just take what it is to be persuaded by the evidence for some claim. Let us say that Marcia believes that O. J. Simpson is guilty of murder on the basis of the blood evidence, along with other considerations. What this would have to mean is that if there were no evidence in favor of O. J.’s guilt, she would not think him guilty. If it turns out she was hardwired or sufficiently prejudiced to think of African- American former football stars as guilty of murder regardless of the state of the evidence, this would make her claim to believe on the basis of evidence false. So for someone to claim to believe that O. J. is guilty (or innocent) on the basis of evidence, the following conditionals must be true.

(1) If strong evidence supporting O. J.’s guilt exists, then Marcia would believe that O. J. is guilty. (2) If strong evidence supporting O. J.’s innocence exists, then Marcia would believe that O. J. is innocent.

If naturalism is true, then sufficient physical causes for one’s forming the belief that O. J. is guilty must exist if you are to believe that O. J. is guilty. Thus, if the physical conditions exist for you to form the belief that O. J. is guilty, then you will form that belief, and if they do not you will not. Yet, those physical conditions contain nothing about blood evidence or any other kind of evidence. After all, there could be a similar world in which the evidence-thoughts do not occur, but the belief is formed anyway. But does the physical determine the mental? Is it really true that, given the physical state of the world, Marcia’s beliefs are determined? There does not seem to be any contradiction in asserting that Marcia’s beliefs could not be very different from what they are even if the physical state of the world was exactly the same. In fact, a “zombie-world” seems logically possible, in which all the physical states of the universe remain exactly as they are now, but there are no minds or mental states at all. Another important point is that naturalists often appeal to evolution to explain the existence of reason. The idea is that surely evolution would select for good reasoning methods over bad reasoning methods. However, this will work only if the mental states involved in rational inferences are causally effective. As Hasker says: “If we accept the physicalist premises of causal closure and the supervenience of the mental, Darwinist epistemology flunks out completely: it has no ability whatever to explain how any of our conscious Pro: The Argument from Reason Defended 89 mental states have even the most tenuous hold on objective reality” (Hasker, 1999, p. 76). Good mental habits can only be selected for by evolution if they are causally relevant. Nothing makes a causal difference unless it makes a difference as to what happens in the world. So unless mental states are causally relevant, then natural selection cannot be used to explain why we have the mental states that we have. In short, I am persuaded that the argument from reason points to a serious difficulty for any naturalistic view of the world. Careful attention to the nature of mind and matter will, on my view, invariably reveal a logical gap between them that cannot be bridged without fudging categories.

Six

CON: NATURALISM UNDEFEATED

David Kyle Johnson

In his contribution to this volume, Victor Reppert offers a significantly revised version of C. S. Lewis’s so-called “argument from reason.” I have been tasked, however, not only with replying to Reppert, but also with showing why Lewis’s version of the argument fails. So let me begin by summarizing Lewis's argument and showing why it falls short. I will then examine and critique Reppert’s revised version to pinpoint why the argument from reason is ultimately a failure.

1. Lewis’s Argument

Lewis first presented the argument from reason at length in the first edition of his 1947 book, Miracles: A Preliminary Study (Lewis, 1947). By my estimation, this version of the argument was successfully defeated by Elizabeth Anscombe in 1948, at a meeting of the Oxford Socratic Club, where she raised objections based on the ambiguity and Lewis's misuse of the words “validity,” “irrationality” and “because." I will not rehash that debate here. To be charitable to Lewis, let us instead look at the revised version of the argument that Lewis published in response to her objections in the second edition of Miracles (Lewis, 1960). Lewis’s argument from reason is an argument against naturalism, which Lewis defines as the view that nothing supernatural exists and that everything that does exist is either some form of physical "stuff" or is fully explainable in terms of physical stuff. The argument attempts to show that naturalism is not compatible with the existence of reasoning—that is, that reasoning cannot exist if naturalism is true. If successful, the argument shows not only that naturalism is false (because reasoning does exist), but also that naturalism is epistemically self-undermining (because naturalism can be rationally believed only if reasoning is possible). Why, according to Lewis, is naturalism incompatible with the existence of reasoning? In short, because reasoning requires insight into real objective relations of logical support, and if naturalism is true, then all acts of inference can be fully explained in terms of non-rational causes. If an act of inference can be fully explained in terms of non-rational causes, then it does not involve any insight into real objective relations of logical support, and thus is not reasoning. Thus, if naturalism is true, no genuine act of reasoning can ever occur. 92 DAVID KYLE JOHNSON

That is the kernel of Lewis’s argument, but let us now examine it in more detail. Lewis sees reasoning as a mental process that recognizes the "logical relation between beliefs or assertions" (Lewis, 2001, p. 23). In reasoning, we recognize—implicitly or explicitly—that some things we believe or know are reasons for drawing certain conclusions. Reasoning thus presupposes what Lewis calls a “ground-consequent relation” (Lewis, 2001, p. 23). Some things we believe (the ground) are seen as entailing, or providing evidence for, a conclusion (the consequent). And what properly connects the ground with the consequent are principles of logical inference, such as non-contradiction, modus ponens, and the rules that govern abduction (inference to the best explanation). If an act of inference is an act of reasoning, it must involve a perception of such logical connections. When we reason, we must perceive or otherwise cognize logical connections that exist, as Lewis says, “outside our own minds” (Lewis, 2001, p. 23) and draw the conclusion for that reason. If the mind, in drawing an inference, is not moved by perception of real logical laws, then we have no reason to believe that that particular inference was drawn in a reliable manner—that the process by which it was reached is a reliable means of acquiring true beliefs. If we do not see these objectively valid principles of inference, and how they lead from one thing (the ground) to another (the consequent), and accept the latter as a result of this inferential support, then what we call “reason” is not a reliable pointer to reality. In short, if truth-preserving reasoning is to occur, we must draw the conclusion we do because, and ultimately because, we recognize a ground-consequent relation between our premises and our conclusion. But if naturalism is true, Lewis argues, perceived logical connections are not ultimately why we draw the conclusions that we do. Recall that naturalism claims that everything that exists is either some form of physical stuff or is fully explainable in terms of physical stuff. This means that our acts of inference, and the conclusions that we draw, are ultimately explained by the causal interaction of physical stuff. Given what we have learned about the brain, most naturalists would say that such causal interactions happen in our brains. Simply put, they would suggest, we draw the conclusions we draw ultimately because our neurons interact in certain ways, not because of perceived ground-consequent relations. In her critique of Lewis's original argument, Elizabeth Anscombe argued that a conclusion can be drawn because of a physical cause-effect process but also (at the same time) because of a rational inference (the recognition of a ground-consequent relation). Both types of explanation can be an accurate account of why the conclusion was drawn because these two different kinds of explanations are not mutually exclusive. This is where Lewis and Anscombe disagree. Lewis believes that even though both natural and mental processes can exist, one must be ultimate—one must be the real, more fundamental, or “full” explanation. Reasoning can exist only if ground-consequent Con: Naturalism Undefeated 93 explanations can at least sometimes be ultimate. But according to Lewis’s understanding of naturalism, it is physical cause-effect explanations that are always ultimate. This is the case, Lewis argues, because even if ground-consequent explanations exist, naturalism entails that they have nothing to do with why particular conclusions are drawn. Why? Because you could subtract the ground-consequent explanation without affecting anything; the cause-effect relation would still exist and the same conclusion would still be drawn. Lewis writes:

But even if grounds do exist, what exactly have they got to do with the actual occurrence of the belief as a psychological event? If it is an event it must be caused. It must in fact be simply one link in a causal chain which stretches back to the beginning and forward to the end of time. How could such a trifle as a lack of logical grounds prevent the belief’s occurrence or how could the existence of grounds promote it? (Lewis, 1960, pp. 24-25).

So, even if naturalism allows for the existence of ground-consequent reasoning, it is still incompatible with such reasoning being the ultimate explanation for the conclusions we draw, and thus naturalism is incompatible with the existence of reasoning.

2. Clarifying Lewis’s Argument

What I initially found most confusing about Lewis’s argument is that it seems paradoxical. On the one hand, he explicitly argues that reasoning does not exist if naturalism is true. Yet, in his subsequent response to Anscombe, he appears to admit that naturalism can recognize the existence of reasoning (that is, of ground-consequent inferences), but claims that if naturalism is true, we have no reason to believe that such inferences are a reliable means to truth. How can reasoning not exist but also be unreliable? After all, something cannot be unreliable unless it exists. The key to unraveling this paradox is realizing that Lewis is offering what logicians call a “theoretical” definition of reasoning, as opposed to a “lexical” or “reportive” one. Lewis thinks that reasoning can exist only if (a) it involves a special sort of non-physical causality (namely, rational insight into objective logical relations of inferential support) and (b) that non-physical causality is the ultimate explanation for why the conclusion is drawn. Since naturalism, according to Lewis, denies that there can be any sort of ultimate non-physical causality, what the naturalist calls “reasoning” is not really reasoning at all. Naturalism is compatible with the existence of ground- 94 DAVID KYLE JOHNSON consequent inferences, sure. But since these inferences do not have ultimate mental explanations, they are not really examples of reasoning. Contrary to first appearances, this does not commit Lewis to thinking that all reasoning is reliable. (Clearly it is not; “good reasoning” is not redundant, and “bad reasoning” is not an oxymoron.) A conclusion can be ultimately caused by a non-physical insight into objective logical relations— and thus be the result of reasoning—and yet not be justified because the insight was faulty. (For example, it could invoke a logical fallacy.) Lewis is claiming, however, that reliable reasoning is possible only if the mind can grasp, and be ultimately moved by, objective principles of inferential support. Since, according to Lewis, naturalism denies that such causally efficacious “graspings” can be ultimate explanations for the conclusions we draw, naturalism implies that no acts of inference can be reliable. With this clarification in mind, we can summarize the main thread of Lewis’s argument from reason as follows:

(1) No act of inference is reliable unless mental ground-consequent processes exist and those processes reliably produce true belief. (2) Such processes can reliably produce true belief only if the mental recognition of the ground-consequent relation between beliefs is the ultimate cause of the conclusions we draw. (3) But, on naturalism, physical cause-effect processes are the ultimate cause of our conclusions and mental processes are a “trifle”—they are so irrelevant you could subtract them and it would make no difference. (4) If (3) then, on naturalism, the mental recognition of the ground- consequent relation between beliefs can never be the ultimate cause of the conclusions we draw. (5) Thus, reliable inferences do not exist if naturalism is true.

Naturalism can be rationally believed and defended only if reliable inference is possible; so, Lewis concludes, it cannot. Of course, the argument is valid, and thus sound if its premises are true. So let us now consider objections to its premises.

3. Where Lewis Goes Wrong

The most vulnerable premise in this argument is premise (3). It is simply not true on all forms of naturalism that mental processes are a trifle—that subtracting them would make no difference. In fact, according to most naturalists they are not. To those familiar with modern philosophy of mind, this is obvious. Modern neuroscience has solidified the longstanding idea that the existence of mental activity is a direct consequence of the activity of the Con: Naturalism Undefeated 95 brain. The relationship is so strong that it is supervenient—that is, no change in the mental can happen unless there is a change in the physical brain. Thus it is simply false to suggest that, on all forms of naturalism, you could subtract the mental ground-consequent reasoning and nothing would change (for example, the same conclusion would be drawn). If the relationship between the mind and brain is supervenient, you cannot subtract the mental without changing the physical, and thus subtracting or changing the conclusion drawn. To more fully understand why, consider a mosaic of small pictures that, when viewed from a slight distance, reveals a larger one—like the one I had in my undergrad dorm room which consisted of individual frames of the Star Wars movies arranged to reveal Darth Vader. The relationship between Vader and the frames is supervenient; no change in Vader is possible without a change in the frames. As such, it makes no sense to suggest that you could subtract Vader and yet leave all the individual frames of the mosaic alone. If the tiles are arranged just as they are, Vader necessarily exists. Yet this is exactly the kind of relationship that most naturalists suggest exists between the mind and brain. That is what supervenience is. On such naturalistic theories, it makes no sense to suggest that the mental is a trifle because you could subtract the mental without changing the physical, because you cannot, any more than you could subtract Vader without changing the frames of the photomosaic. Consequently, contrary to what Lewis claims, naturalism does not imply that you could subtract a ground-consequent process while leaving the physical world undisturbed. It is true that some philosophers believe that so-called philosophical zombies—beings physically identical to us but without minds—are possible. Thus some philosophers do think you could subtract the mental without changing the physical. But this fact does not help Lewis. At best this means that what Lewis says about the mental being a trifle is true on some varieties of naturalism. But for Lewis’s argument to succeed, he must show that all varieties of naturalism are incompatible with rational inference. He is arguing that naturalism, as such, is essentially incompatible with the existence of rational inference. Since, on many varieties of naturalism, his third premise is false, the argument fails. One might be tempted to drop Lewis’s “trifle” observation to save the argument, thus making the third premise claim simply that naturalism implies that ground-consequent accounts can never be ultimate explanations. But such a premise would also be false. Some varieties of naturalism maintain that mental ground-consequent explanations can be ultimate. Property dualism is perhaps the best example. Property dualists suggest that, although only matter exists, it has two kinds of properties—physical properties and mental properties, both of which have unique causal effects. Clearly, this is a variety of naturalism that is perfectly compatible with ground-consequent explanations being ultimate. Property dualism is a form of naturalism in that it holds that all that exists is either a physical thing or a 96 DAVID KYLE JOHNSON property of a physical thing. Yet it maintains that mental processes, like acts of ground-consequent reasoning, are ultimately causally responsible for our conclusions. Another example can be found in mind-brain identity theory, which suggests that mental events are numerically identical to brain events—that is, that they are one and the same thing. So not only can you not change or subtract the mind without changing or subtracting the brain, but on identity theory the “ultimacy” of ground-consequent explanations and physical cause- effect explanations must go hand in hand. Since they are numerically identical on this naturalist theory, the fact that one is ultimate logically necessitates that the other is as well. Lewis’s argument might be thought to be successful against specific varieties of naturalism that explicitly deny the causal efficacy or the existence of the mind. Epiphenomenalism, for example, suggests that minds exist but are causal danglers—they do not cause anything, including our behavior. Eliminative materialism claims that minds do not exist at all. If Lewis’s second premise is right and our conclusions must ultimately be caused by mental grounds-consequent processes if they are to be reliable, clearly both epiphenomenalism and eliminative materialism are incompatible with the existence of reliable reasoning. Such processes cannot be the ultimate cause of anything if they have no causal powers or do not exist at all. Unfortunately, the second premise of Lewis’s argument is what epiphenomenalists and eliminative materialists explicitly deny. Epiphenomenalists argue that the idea that mental events, like acts of ground- consequent reasoning, cause anything is an illusion. The fact that mental events are constantly correlated with our actions makes us think they cause them, but correlation does not entail causation. Eliminative materialists would argue that Lewis’s entire argument (and especially the second premise) is rooted in a type of folk psychology that can, should, and will eventually be rejected. Thus, as an argument against such versions of naturalism, Lewis’s argument would simply beg the question. For one to believe that the second premise of his argument is true, one would already have to assume that those versions of naturalism are false. And although most scientists probably do not accept such views—and thus would not be bothered by Lewis’s assumption that these views are false—many philosophers do accept these views. In a debate with them, the truth of the second premise would be the issue at hand; but the argument from reason does not establish that premise—it assumes it. Thus, the argument from reason cannot successfully refute either epiphenomenalism or eliminative materialism. (Lewis’s separate, but related, argument about evolution not being compatible with reliable reasoning might be thought of as a defense of the second premise. Reppert briefly mentions but does not develop that argument; I will be happy to take that issue up in the rejoinders if he wishes to defend it.) Con: Naturalism Undefeated 97

Premise (2) is challenged further by the existence of thinking computers, like IBM’s Deep Blue (which beat the world’s best chess player, Garry Kasparov) and Watson (which bested Ken Jennings and Brad Rutter, the best players in the history of Jeopardy!). They show us that mental recognition of ground-consequent relations is not at all necessary for reliable production of true beliefs. The physical cause-effect processes they go through more reliably produce true beliefs about successful chess strategies and correct replies to Jeopardy! clues than anything else on the planet. One might object that such processes do not produce true beliefs because Deep Blue and Watson do not believe anything—they are computers without mentality. However—setting aside replies rooted in the possibility of future artificial intelligence—we can easily avoid this objection by noting that we could theoretically hook a person up to these computers in such a way that they reliably produce true beliefs in that person without that person going through any mental ground-consequent process. Imagine that, after the computer figures out the right answer, it just makes the answer appear in the person’s mind by sending the appropriate signal to his or her brain. Clearly, in such a scenario, true beliefs would be reliably produced without any mental effort at all. And that suggests a challenge to even the first premise of the argument: reasoning may not require the apprehension of mental ground-consequent relations in the first place. Even though Watson does not have a mind, and so does not mentally apprehend ground-consequent relations, there seems to be nothing inaccurate in saying that Watson rationally and reliably infers the correct responses to Jeopardy! clues. And it will be even harder to deny the rationality of androids—robots whose behavior is exactly like ours—if we one day invent them. But we may not even need artificial intelligence to find examples of nonmental rational inferences; we may need only to look to ourselves. Multiple repeated experiments have shown that many decisions we make, which we think are caused by conscious decision making, are actually made unconsciously; only after unconscious areas of our brains have already determined what action we will take is the fact of that decision fed to conscious parts of our brains where the mental experience of making the decision is generated (Dvorsky, 2013). So-called “mental decision making” often involves ground-consequent reasoning; whether the decision is simple or complex, we draw a conclusion regarding how we should behave. But these experiments show us that such reasoning is often (perhaps always) done unconsciously, and only after it is complete does our brain bother to produce the mental experience of recognizing the relationship between grounds and consequent. So mental phenomena, like ground-consequent reasoning, sometimes are causal danglers! But we must consider these nonmental processes rational inferences because, not only are they often reliable, but if anything is rationally inferred it is the conclusions of reasonable human 98 DAVID KYLE JOHNSON beings—and these nonmental processes are the causal explanations for many of the conclusions we draw. This neuroscientific discovery does not disprove the existence of rational inferences, but shows us more accurately what rational inferences are—namely, logical processes that are not necessarily conscious or even mental. In fact, we may not even need grandiose experiments to know that reasoning can be nonmental; we can just look to our own experience. We have all had an answer we were trying to figure out come to us out of the blue, in a eureka moment—that is, without conscious recognition of ground-consequent relations. Of course, such answers are not actually descending on us from some Platonic realm; unconscious parts of our brains are working out the solution for us. Academic accounts of such “unconscious reasoning” exist (see Johnson-Laird, 2008, especially Chapter 5), and such reasoning is valuable even to chess champions, like Kazparov. Kasparov does not always think fifteen moves ahead; sometimes his instincts just recognize that one move puts him in a stronger position than another. Many times, it is explicitly a matter of pattern recognition—again, not classic ground-consequent reasoning. If he has time, he will check these instincts with conscious analysis; but even then, the conclusion was reached, originally, through unconscious reasoning (Kasparov, 2008, especially Chapter 14). Likewise, Ken Jennings (at least if he is anything like me) does not always figure out the right response to Jeopardy! clues by going through a conscious ground- consequent reasoning process—sometimes the answers just come to him. Clearly, it seems, rational inference does not require the recognition of ground-consequent relations; most certainly, the reliable production of true belief does not. Thus, all the major premises of Lewis’s argument face serious objections. For this reason, the soundness of his argument is on shaky ground indeed.

4. Where Reppert Goes Wrong

Reppert does not really defend Lewis’s argument from reason; in fact, he rejects many of Lewis’s stated assumptions. Instead, he offers his own “updated” versions of the argument from reason which he thinks are not subject to the same objections. To this end, instead of couching his argument in terms of full or ultimate explanations, he argues that naturalism has metaphysical commitments that are in conflict with its epistemological commitments. Metaphysically, Reppert suggests, naturalism is committed to (a) the claim that all events, at the base level, can be explained mechanistically (b) the causal closure of the physical, and (c) a principle of supervenience which entails that anything non-physical that exists in space and time must strongly supervene upon the physical—“[g]iven the physical, Con: Naturalism Undefeated 99 everything else is a necessary consequence.” These metaphysical commitments, Reppert argues, are not compatible with what naturalists are committed to epistemically, notably: (1) that rational insight into logical relations plays a causal role in acts of inference, (2) that thoughts possess “intentionality” (aboutness), (3) that propositional content (what the thought is about) is operative in mental causation, and (4) that it is possible to form beliefs on the basis of evidence. I do not have space to deal explicitly with all the metaphysical-vs.- epistemological conflicts that Reppert alleges, but we can see that his claims are doomed to fail by examining the main arguments he offers. Take, for example, Reppert’s argument from mental causation. He suggests that the metaphysical commitment of naturalism to causal closure prevents mental explanations from doing any causal work when it comes to explaining why we draw the conclusions we do. Why? For one thing, the causal closure of the physical entails that mental entities, like beliefs, are causally inert. When speaking for the naturalists, he asserts: “Beliefs (if they exist at all given naturalism—of course this is denied by eliminativists) are strictly epiphenomenal.” Furthermore, if beliefs do have causal efficacy, they must have it in virtue of their propositional content. But, Reppert argues, naturalism rules this out:

There does not seem to be any contradiction in asserting that [a person’s] beliefs could not be very different from what they are even if the physical state of the world was exactly the same. In fact, a “zombie- world” seems logically possible, in which all the physical states of the universe remain exactly as they are now, but there are no minds or mental states at all.

Both of these suggestions are mistaken. The first suggestion is refuted by simply looking at the available naturalistic theories of mind. Pace Reppert, eliminative materialism and epiphenomenalism are only two of many naturalistic philosophies of mind; the causal closure of the physical does not necessarily rule out the causal efficacy of the mental. For example, as we have already seen, property dualism suggests that mental properties are real existing properties that have causal efficacy. The same would be true of identity theory, since it identifies mental events with brain events, and brain events are what cause our actions. Reppert suggests that the difficulties he raises “will exist for any kind of naturalism that [he] can think of.” Since they obviously do not exist for property dualism or identity theory, it seems that he did not think very hard. Of course, Reppert might insist that such theories make the mistake of de-mentalizing the mental by attributing it to the physical brain; he will likely invoke his somewhat confusing quip, “Interesting fellow Mr. Brain. Remarkable what he can do.” There are a few things to say in response. First, 100 DAVID KYLE JOHNSON even if these theories are mistaken, Reppert’s argument still fails. Like Lewis, Reppert is arguing that naturalism is essentially (by its very “nature,” if you will) incompatible with the existence of the mental (and thus with reasoning). If so, then there is no possible naturalistic theory that is compatible with the mental. But even if property dualism and identity theory are false, they are still possible naturalistic theories; by showing that they are compatible with the mental, one refutes Reppert’s argument. In response, Reppert might suggest that by incorporating the mental, such theories are no longer naturalistic theories; but this follows only if we assume that the mental can only be accounted for in non-materialist terms. But, of course, such an assumption is exactly what property dualists and identity theorists argue against. Thus, this assumption would simply beg the question. In addition, such naturalists would argue that they are not de- mentalizing the mental at all—they are simply including it in the physical. And the reason they are doing so is because we have discovered that, indeed, Mr. Brain is remarkable. Through years of careful scientific work and research we have discovered that Mr. Brain is directly responsible for the existence of the mental; the mental does not exist as a separate entity, but is entirely dependent upon Mr. Brain. Since the remarkable Mr. Brain is a physical object, it makes perfect sense to include the mental in the physical world. Reppert is also mistaken in claiming that, for naturalists, the propositional content of beliefs must be causally inert because such content could be changed or subtracted without effect. He claims that “there could be a similar world in which the evidence-thoughts do not occur, but the belief is formed anyway.” Like Lewis, Reppert clearly does not fully appreciate how most naturalists understand the supervenient relationship between the mind and brain: you cannot change (or subtract) the mental without changing (or subtracting) the neurophysical—any more than you can subtract Darth Vader from my photomosaic without changing the individual pictures. Indeed, half of philosophers think that the zombie world Reppert describes is metaphysically impossible (Bourget & Chalmers, 2013). When certain neurophysical states obtain, certain mental events necessarily obtain as well. The mental phenomenon of reasoning is strictly necessary if the conclusion in question is to be drawn—you cannot have one without the other. Therefore, the propositional content is relevant to the conclusion’s derivation. It is worth noting that, even on a weak understanding of supervenience, the neurophysical necessitates the mental in such a way as to refute Reppert’s argument; on such a view, our conclusions still cannot be drawn without specific mental content. But it is true, by definition, that Reppert’s claim about the contingent relationship between the neurophysical and mental is false given the strong understanding of supervenience to which Reppert insists naturalists are metaphysically committed. On such a view, contrary to Reppert’s assertion, there is not even a logically possible world in which two Con: Naturalism Undefeated 101 objects share all neurophysical properties but do not also share all mental properties. (On the distinction between strong and weak forms of supervenience, see McLaughlin and Bennett, Section 4.1.) Reppert reveals the mistake behind his reasoning when he suggests, as we saw above, that the fact that there is no “contradiction” in denying a certain necessary linkage between mind and brain, shows that the relationship between the mind and brain is merely contingent. This invokes a pre-Kripkean linguistic theory of necessary truths which is obviously false (Kripke, 1980). There is no logical contradiction, for example, in asserting “the morning star is not the evening star,” but that does not mean that they actually are distinct objects; once you understand their nature (they are both the planet Venus), you realize one cannot exist without the other. Likewise, once you understand the nature of mind and brain, you realize that you cannot have one without the other. The fact that there is “no contradiction” in denying necessary mind- brain connections does not prove that mental events do not strongly supervene on brain events. Consider, next, Reppert’s argument regarding intentionality or “aboutness.” He insists that naturalism cannot account for such things because he cannot “see” how it could. He points this out with a series of rhetorical questions:

Now, all [intentional] states, according to even minimal naturalism, have to be the necessary implications of physical states. But how is that even possible? Is there anything about the physical that guarantees that the mental states that we all believe ourselves to be in must obtain? It is hard to see what that could be. Can any conglomeration of nonintentional facts entail the existence of an intentional state?

He goes on to clarify his position by suggesting that even a complete list of physical facts (for example, about a person’s brain) is insufficient to determine what mental state a person is in. “In the case of mental states, I do not see how the physical states can possibly ‘add up’ to any determinate mental state.” Reppert’s first mistake here is obvious: he takes his inability to understand how the physical could explain the mental to be a reason to conclude that it cannot, in principle, do so. But, of course, that is not the case. It certainly cannot be a deductive proof of such—that is, unless Reppert is omniscient; only if Reppert knows everything would “his inability to see how something could be done” be proof that “it cannot be done.” If he is not omniscient, his inability could just be a result of his ignorance. But his inability is not good inductive evidence either. On matters such as this, only experts can be confident in such judgments, and Reppert is not an expert in the relevant fields (for example, neuroscience and cognitive psychology). In 102 DAVID KYLE JOHNSON point of fact, experts in the relevant fields generally do not share Reppert’s pessimistic outlook (even if exceptions can be found). What is more, it is not clear that even expert opinion could settle a matter such as this; consequently, it seems that not even an expert’s inability to find a solution to such a problem is any indication that there is none. We might think that Reppert’s point is deeper: it is not merely a lack of understanding; we can positively see that no physical state of affairs can “add up” to determinate mental facts, just like we can see that no set of purely descriptive facts can possibly add up to moral facts—like whether a homicide is morally justified. But, of course, both matters are far from settled. It is commonly said that you cannot get an “ought” from an “is,” but this is disputed. Utilitarians, for example, would argue that facts about how much happiness an act produces can fully determine its moral status. Likewise, the matter of whether mental facts are derivative from physical facts is far from settled. Not only is there not agreement among the experts about whether the hard problem of consciousness will be solved, there is not even agreement about whether it is really a problem to begin with. If the argument from reason hangs on this deeper assumption—that “you cannot get a mind from a brain”—it hangs from a thin thread indeed. Moreover, most naturalists expect that we will one day be able to do exactly what Reppert insists is impossible: be able to look at the physical state of a person’s brain and determine exactly what mental state he is in. In fact, current neuroscience has already made great strides in that direction. And this is what most naturalists expect because they believe that the relationship between the brain and mind is supervenient. Reppert suggests that “[i]f the state of the nonmental substrate fails to guarantee [mental facts], then, on the naturalistic views we have been considering, nothing does, and there can be no such facts.” But on most varieties of naturalism, the nonmental substrate does guarantee mental facts, because the latter supervenes on the former. Thus, such theories suggest that the physical does exactly what Reppert thinks it must do in order for such theories to be viable. Although our knowledge of the brain’s workings is still limited, most naturalists believe that one day we will see how it all works together—just like we can now look at individual frames of Star Wars and figure out that, in a certain arrangement, they will give rise to Darth Vader. Being able to look at a person’s brain and tell what that person is thinking is not, in principle, impossible. Some think that such detailed mental-physical mappings will probably occur. If they do, it will not simply be a matter of finding correlates between mental and brain events; it will be a result of truly understanding how the brain works. Once again, it seems that Reppert’s misunderstanding of what supervenience entails has crippled his argument. Looking beyond Reppert’s main arguments regarding mental causation and intentionality, we see that, like Lewis, Reppert assumes that reasoning and reliable belief-producing processes must be mental. As I have already Con: Naturalism Undefeated 103 shown, however, these assumptions are undercut by Deep Blue, Watson, and recent research into unconscious reasoning processes. Moreover, what Reppert assumes is true of naturalism would in fact be rejected by many naturalists, such as epiphenomenalists and eliminativists. Quite simply, they would deny many of the epistemic notions to which Reppert thinks naturalism is committed. I will be glad to say more about these issues in the replies, but, regardless, notice that we have seen Reppert’s arguments fall prey to all the same objections that haunted Lewis. So Reppert is mistaken in thinking that his updated arguments are a significant improvement over Lewis’s. I do not have space to say more here about the host of other things that Reppert thinks naturalism cannot account for—for example, propositional content, truth and falsity, and the apprehension of (metaphysically existing) logical laws. But the objections to such arguments will boil down to something similar to those I have already expressed. Although accounting for such things is philosophically complicated, there are viable naturalistic accounts that can be offered. And what is more, it is not even clear that such things actually are necessary for the existence of reasoning or reliable inference. On many levels, therefore, the argument from reason—both in its classic (Lewisian) and revised (Reppertian) forms—is a failure.

Seven

REPLY TO DAVID KYLE JOHNSON

Victor Reppert

I am grateful for Kyle Johnson’s critique of my defense of C. S. Lewis’s argument from reason. I believe I need to begin by explaining the difference between Lewis’s argument and mine. While I believe Lewis’s argument to be fundamentally correct, a lot has happened in philosophy over the course of the more than five decades since Lewis wrote his revised argument, and still longer since Miracles was first published in 1947. In particular, I believe that the concept of naturalism can be explicated in far more detail than Lewis could possibly have done in his time. But since Johnson’s criticisms of my argument and Lewis’s are much the same, we can focus on what I take his central point to be, and why I think it does not refute Lewis’s argument. Now a good definition of naturalism, materialism, or physicalism is pretty important here, because if you define naturalism broadly enough, I could qualify as a naturalist. That is, I could say that I believe that all entities are natural, including psychons (which people used to call souls), angelons (which people used to call angels), and the theon (which used to be called God). William Hasker, in The Emergent Self (1999), developed a tripartite definition of minimal materialism which I think also works for naturalism. That is, I do not think any view can be thought to be genuinely naturalistic unless it satisfies these three requirements. It is the view that:

(1) At the basic level, reality is mechanistic. That is, it lacks intentionality, subjectivity, purposiveness, and normativity. None of these items can enter into a description of reality at the basic level of analysis. (2) The basic level of analysis (which we typically call physics) is causally closed. (3) Whatever else exists must supervene on the basic level. It must be the sort of thing that must be the way it is because the physical is the way it is.

Johnson has not objected to the way I have these things defined, so I take it that he finds my definitions acceptable. Notice here that to get an adequate definition of the physical, we have to proceed by defining certain features of 106 VICTOR REPPERT the mental and denying that the physical contains these mental elements at the most basic level of analysis. It is the last of these requirements, supervenience, which comprises the centerpiece of Johnson’s response to me. If the mental supervenes on the physical, then one cannot remove the mental without also changing the physical. Thus if naturalism is true, the state of the mental is necessarily the way it is given the state of the physical. What that does is render the mental explanation and the physical explanation compatible. Both are true, and the physical explanation does not exclude the mental explanation. Johnson explicates this through an analogy to a picture, which, at the micro level, consists of a set of various Star Wars scenes. The mosaic picture, taken as a whole, is a picture of Darth Vader. He writes:

The relationship between Vader and the frames is supervenient; no change in Vader is possible without a change in the frames. As such, it makes no sense to suggest that you could subtract Vader and yet leave all the individual frames of the mosaic alone. If the tiles are arranged just as they are, Vader necessarily exists.

Now, I actually think that this example illustrates the difficulties with the argument from supervenience, and that there are some significant disanalogies between the mosaic case and the mental-physical supervenience case. Most notably:

(1) In the case of the pictures, something can have the property of “being a picture of Darth Vader” by having the color, size, and shape of something that looks like Darth Vader. (2) Nothing in the picture, either of Vader or any of the smaller Star Wars scenes, has anything to do with anything that does not have a particular location in space and time.

Now let us look at what has to supervene on the physical in the mental- physical case. At the base level, there is nothing intentional, subjective or perspectival, nothing purposive, and nothing normative. Yet mental states are all four of these things. What is more, suppose we are talking about a rational inference here. The logical truth that legitimizes this rational inference has no particular location in space and time. Yet it is relevant to whether the inference occurs or does not occur. In virtue of what do mental states supervene on physical states? Is it because the physical states add up to mental states when put together? That seems to be the case with the Darth Vader picture. If the pictures grouped together cause us, under normal conditions, to see Darth Vader, then it really is a picture of Darth Vader. It is more difficult, surely, to say something like that in the case of mental states. For example, it does make sense to say, in the Reply to David Johnson 107 case of such mental states, that even if something had been put together which leaves an outside observer to think that what we have here is a bat, we might still wonder if the set of physical states that we are referring to as a bat experiences what it is like to be a bat. We do not expect a picture of Darth Vader to experience what it is like to be Darth Vader, but if we encountered something that acted like Darth Vader, we would want to know if it really knew what it was like to be Darth Vader, or whether it was simply producing Vader-like behavior. If something supervenes on the physical, we need to know why it supervenes. To say that it is just a brute fact that mental state X supervenes on physical state Y will not do. The reason is not hard to seek. If supervenience theory is true, then everything supervenes on the physical, and that would have to include the supervenience relation itself. If everything supervenes on the physical plus a supervenience relation, then it is not quite true that everything supervenes on the physical. So we need an account of why the supervenience relationship obtains to avoid a reflexivity problem. Johnson presents two types of naturalistic theories which, he says, allow us to say that the ground-consequent relations and the perceptions of them that are essential for rational inference can indeed be ultimate explanations. Now I must admit to being at least somewhat puzzled as to what he means by ultimate explanations. I would have thought ultimate explanations were explanations at the most basic level of analysis—that is, explanations at the level of basic physics. Now if the basic level is mechanistic, then mental explanations cannot possibly be ultimate explanations, since basic physics is by definition mechanistic, and therefore lacking in intentionality, purpose, subjectivity, and normativity. Perhaps he means something else by “ultimate,” but if he does, it would be nice to know what he does mean. I am myself inclined to agree with naturalist Andrew Melnyk when he says that “[n]aturalism claims that nothing has a fundamental purposeful explanation. . . . Naturalism says that whenever an occurrence has a purposeful explanation, it has that explanation in virtue of certain nonpurposeful (e.g., merely causal) facts” (Melnyk, 2007). Property dualism is the first of the naturalistic theories Johnson discusses. He writes: “Property dualists suggest that, although only matter exists, it has two kinds of properties—physical properties and mental properties, both of which have unique causal effects.” But here Johnson is forgetting how we defined “naturalism.” If there are nonphysical properties of physical objects that have unique causal effects, then one of the basic defined tenets of naturalism is violated: the causal closure of the physical. Let us take a pair of electrons, Eric and Ed. Eric the electron is an electron in the body of Richard Dawkins. Ed is an electron inside the body of the Archbishop of Canterbury. Knowing this, we might know that Ed is far more likely to be inside an Anglican Church this Sunday morning than Eric is. But if the physical is causally closed, then where Eric 108 VICTOR REPPERT and Ed will be on Sunday morning is either fully determined by the present state of the world at the physical level plus the laws of physics (assuming determinism), or it is governed by quantum-mechanical indeterminism. But this indeterminism, given causal closure, is brute chance and nothing more. It is not a “window” for intentional states to affect the physical. Imagine a being that is omniscient with respect to the state of the physical world. Such a being could learn nothing useful about where Eric and Ed will be if it learns simply that Eric is in the body of an atheist, while Ed is in the body of a Christian. The electrons will go where the laws say they must go, or where chance places them. To say otherwise would be to deny the causal closure of the physical world. To say that the mental properties have unique causal effects means that matter, where these properties are instantiated, behaves differently from the way it does where those properties are not instantiated. That is something that cries out for explanation, yet the explanation cannot be found within the system of nature itself. Why do the fundamental laws covering the behavior of matter suddenly change when they are configured into a brain? The appeal of naturalism is the promise that, in the final analysis, reality can be explained by a single set of laws and facts. A property dualist must either accept epiphenomenalism about mental states, or else reject the causal closure of the physical. According to Edward Feser:

Property dualism seems if anything to have a worse problem with epiphenomenalism than does Cartesian dualism. Recall that the Cartesian dualist who opts for epiphenomenalism seems to be committed to the absurd consequence that we cannot so much as talk about our mental states, because if epiphenomenalism is true, those mental states have no effect at all on our bodies, including our larynxes, tongues and lips. But as Daniel Dennett has pointed out, the property dualist seems committed to something even more absurd: the conclusion that we cannot even think about our mental states, or at least about our qualia! For if your beliefs—including your belief that you have qualia—are physical states of your brain, and qualia can have no effects on anything physical, then whether you have qualia has nothing to do with whether you believe that you have them. The experience of pain you have in your back has absolutely no connection to your belief that you have an experience of pain in your back; for, being incapable of having any causal influence on the physical world, it cannot be what caused you to have beliefs about it (Feser, 2005).

So property dualism does not render mental states causally effective. Quite the opposite. If the physical is causally closed, that means that only physical properties are causally relevant. So, if there are properties that are not physical, they are causally irrelevant, and so it can never be the case that, for Reply to David Johnson 109 example, Jones believes that atheism is true because he sees that the evidence from evil supports atheism. The second naturalistic theory Johnson presents is identity theory. He writes:

Another example can be found in mind-brain identity theory, which suggests that mental events are numerically identical to brain events— that they are one and the same thing. So not only can you not change or subtract the mind without changing or subtracting the brain, but on identity theory the “ultimacy” of ground-consequent explanations and physical cause-effect explanations must go hand in hand. Since they are numerically identical on this naturalist theory, the fact that one is ultimate logically necessitates that the other is as well.

Well, no. Identity theories come in two varieties, type-type identity theories and token-token identity theories. In explaining mental states, a type- type identity theory might actually offer some hope of reconciling the ground- consequent and cause-effect relationships if, for example, there was some discernible type of physical state that, for instance, everyone is in when they draw a modus ponens inference—that is, an inference of the form “If A then B; A; so B.” But this claim seems implausibly strong. The version of identity theory that is typically defended is a token-token identity theory, but that in no way guarantees that every characteristic of the object in question is going to be relevant to the causal transaction in question. And even type-identity theory might not be strong enough. For explanations to be identical, the properties would need to be logically equivalent; mere empirical correlation would not be sufficient. To provide an example here that I have employed before: Suppose I purchased the baseball the Luis Gonzalez hit when he hit the single that gave the Arizona Diamondbacks their world championship in 2001 in their victory over Mariano Rivera and the New York Yankees. Now suppose I (foolishly) let some kids play a baseball game with the Gonzo baseball, and it strikes and breaks my dining room window. It would then be true that the baseball Gonzo hit was the baseball that broke my window, but the baseball’s having been hit by Gonzo in the World Series would not make it any more or less capable of breaking my window than any other baseball. The size, velocity, and density of the ball would affect that, but its having been the ball used in the World Series would have nothing to do with it. Hence we will be able to say that the ball Gonzo hit broke the window, but it cannot possibly break the window in virtue of its being the ball Gonzo hit. It is one thing to say that prior brain states identical to the thoughts “All men are mortal” and “Socrates is a man” caused me to have the subsequent thought that “Socrates is mortal.” It is another to say that the ground-consequent relation between the first premises and the conclusion was relevant to the actual forming of my belief that the 110 VICTOR REPPERT conclusion is true. The laws of physics, not the laws of logic, govern the actions of the physical world. Nor, as Johnson claims, does the existence of unconscious reasoning undermine my argument. There are certainly cases where we draw conclusions, by force of habit, or in virtue of processes other than explicit, conscious reasoning. All I need for my argument is the claim that, in order for science to be what it is, there are occasions when we perform explicit rational inferences. There are times when we go through a process of inference in which the steps in the process can be explicitly spelled out, so that someone else with the same inferential capacity can replicate the process and see how we got our answer. In chess not everything has to be calculated exactly, but in many endgame positions, this is exactly what is required. Similarly, when we learn math, we are expected to show our work. It is not enough to get the right answer; we are expected to show that we have performed the precise conscious reasoning steps so that someone can trace those steps and get the same result we did. The teachers who checked our work in math needed to know whether we perceived the ground-consequent relations that we were supposed to perceive in order to demonstrate the relevant math skills. Johnson is also wrong in thinking that computers prove that reasoning can be done by purely mechanistic systems. Sure, a computer is a machine, but it is a machine that has derived intentionality instead of original intentionality. The meaning of everything it does has to be programmed in. In the last analysis, as Karl Popper pointed out, a computer is just a glorified pencil. The argument from reason says that reason cannot emerge from a closed, mechanistic system. The computer is, narrowly speaking, a mechanistic system, and it does “follow” rational rules. But not only was the computer made by humans, the framework of meaning that makes the computer’s actions intelligible is supplied by humans. As a set of physical events, the actions of a computer are just as subject as anything else to the indeterminacy of the physical. If a chess-playing computer plays the move Rf6, and we see it on the screen, it is our perception and understanding that gives that move a definite meaning. In fact, the move has no meaning to the computer itself; it only means something to persons playing and watching the game. Suppose we lived in a world without chess, and two computers were to magically materialize in the middle of the Gobi desert and go through all the physical states that those computers went through the last time they played a game. If that were true they would not be playing a chess-game at all, since there would be no humans around to impose the context that made those physical processes a chess-game and not something else. Hence, Johnson’s computer objection is a red herring. In addition, Johnson mentions two other naturalistic theories which he thinks can respond adequately to the argument from reason: eliminativism and epiphenomenalism. Here he mentions the standard eliminativist response that the self-refutation argument against it begs the question because it Reply to David Johnson 111 presupposes folk psychology. But what the self-refutation argument assumes is not that folk psychology is true; what it argues for is that the truth of propositional-attitude psychology is necessary for, for example, scientific knowledge. And that could be true even if eliminativism is true. That is, there is a possible world in which there are no propositional attitudes, but also no scientific knowledge. Of course, scientistic philosophers like the Churchlands will certainly avoid such a view like the plague, but their critics’ argument does not assume that it is false, and therefore they are not begging the question. Similarly, some philosophers may be epiphenomenalists, but the charge of epiphenomenalism has been an important element in arguments against non-reductive materialism, and for good reason. Epiphenomenalism does imply that prior mental states do not cause subsequent mental states, and that implies that the kind of rational inference on which the natural sciences rely does not occur. This is certainly not logically impossible, but again, it is hardly a position that someone who thinks highly of the sciences can reasonably accept. In short, Johnson simply supposes that since naturalistic theories have been proposed that are thought by some philosophers to allow for rational inference, they must all work. It takes more than this to refute the argument from reason. When we examine such theories closely, it becomes apparent that fatal problems remain, and often those problems are pointed out by defenders of other naturalistic theories. In a good deal of philosophy of mind over the last five decades or so, physicalism seems to have been taken as a kind of absolute presupposition. A good example would be Daniel Dennett, who once remarked: “[B]efore I could trust any of my intuitions about the mind, I had to figure out how the brain could possibly accomplish the mind’s work” (Dennett, 1994, p. 236). This leads him to treat the brain as a “syntactic engine” that mimics the competence of semantic engines (though where Dennett thinks semantic engines can be found is, to say the least, very unclear). This strikes me as dogmatic, and leads me to think that, for the most part, materialist philosophers have not solved the problems posed by anti-materialist arguments such as Lewis’s argument from reason, but rather have presupposed that there has to be a materialistic solution to such problems. But what if those assumptions are questioned? If they are, then the problems posed by arguments of the kind I have presented seem to me to expose a deep incoherence in naturalism. I do not think Johnson’s responses really show us how to resolve this incoherence.

Eight

REPLY TO VICTOR REPPERT

David Kyle Johnson

Let me begin by expressing appreciation to Victor Reppert for his reply to my comments on Lewis’s argument from reason. Here in my final remarks, to show why the argument from reason is ultimately a failure, I would like to identify six reasons why Reppert’s defense of the argument from reason against my objections is unsuccessful.

1. Reppert Dropped My Argument against Lewis

Although Lewis was certainly a non-naturalist, and he was attempting to derail naturalism, the central conclusion of Lewis’s argument from reason is not that naturalism is false, but that it is self-refuting in a certain sort of way. In a nutshell, Lewis’s argument from reason runs as follows: “If naturalism is true then the process by which the conclusion ‘naturalism is true’ is derived is unreliable, and thus the conclusion is unjustified. So a naturalist cannot be justified in believing that naturalism is true.” Consequently, to defend Lewis’s argument from reason—or even the notion that it is “fundamentally correct,” as Reppert claims—one must ultimately defend the notion that naturalism is epistemically self-defeating. Reppert, however, does not do this. Instead, he simply argues directly that naturalism is false (because it is incoherent, since its ontological commitments conflict with its epistemological commitments). If he accomplishes that goal, he will have achieved something that he deems desirable, but he will not have defended Lewis’s argument from reason. The fact that naturalism is false does not entail that it is self-defeating. We can see this mistake at the very beginning of Reppert’s reply. In my first response, I observed that according to some naturalist theories the relationship between the mind and brain is supervenient, and showed why this would entail that the mental recognition of ground-consequent relations is not a trifle, contrary to Lewis’s claim. It thus follows, I showed, by the logic of Lewis’s own argument, that on such naturalist theories, ground-consequent reasoning is necessary for and ultimately explains derived conclusions—and this is just what Lewis suggests naturalist theories must entail to allow for (reliable) reasoning to occur. To counter this argument, Reppert simply argues that the relationship between the brain and mind is not actually supervenient. Notice that, even if true, this would not show that the naturalistic theories in 114 DAVID KYLE JOHNSON question are epistemically self-refuting, only that they are false. They would be wrong about the relationship between the mind and brain, but it would still be the case that, if they were true, then the recognition of ground-consequent relations would not be a trifle (and in fact would be the ultimate explanation for derived conclusions). So, in what I think is now an obvious way, Reppert’s reply regarding supervenience is not relevant to whether Lewis’s argument from reason is sound. Reppert suggests that, in general, my reply simply amounts to this: “[S]ince naturalistic theories have been proposed that are thought by some philosophers to allow for rational inference, they must all work.” This is not accurate, and I am unclear what he means by “work.” Clearly, they cannot all be true; they are mutually exclusive theories. But none of them even needs to be true to refute Lewis’s argument from reason; just one of them needs to be epistemically consistent. Since, by Lewis’s own argument, naturalistic theories that maintain a supervenient relationship between mind and brain are epistemically consistent, Lewis’s argument fails. And since its soundness is our primary concern here, the main debate is settled.

2. Reppert Misunderstands Supervenience

But let us now look at Reppert’s denial of mind-brain supervenience. Reppert suggests that a mosaic of frames from the Star Wars movies can have the supervening property of looking like Darth Vader because a picture of Vader is something that has a certain kind of size, color, and shape, and these properties are not lacking in the mosaic’s parts (the individual frames). But mental events have properties that no physical part of the brain has— intentionality, subjectivity, and so forth. Thus, Reppert claims, the mental cannot supervene on the physical. This simply does not follow; on many naturalistic accounts, the mental supervenes on the brain in virtue of being an emergent property. But an emergent property simply is a property had by a whole in virtue of its parts despite the fact that the parts themselves do not have that property. The fact that none of the individual frames looks like Vader would not prevent the whole mosaic from having that property. Likewise, the fact that no part of the brain has mental properties would not prevent the whole from having such properties. Reppert simply commits the fallacy of composition; the fact that the parts of a whole lack a certain property does not necessarily entail that the whole lacks that property. If the property is emergent—as some naturalists suggest is the case with mentality—then the whole can have a property the parts lack. Reppert also claims that naturalists must explain how mind-brain supervenience works; they cannot simply assert that the mind supervenes on the brain as a brute fact, because that generates a reflexivity problem. Here, Rely to Victor Reppert 115 three mistakes must be noted. (1) One need not explain a supervenient relationship to know that it exists. Simply looking is enough to know that Vader supervenes on the mosaic; I need not be able to explain how each individual tile contributes to the image. Likewise, neuroscience has revealed that the mental is a result of the operations of the physical brain, and no change in the mental is possible without a change in the brain. We may not yet know how the brain produces supervenient mental properties, but we can know that it does—and that is all we need. (2) There is nothing inherently wrong with accounting for such a relationship by appealing to brute facts; all explanations bottom out somewhere. Proposing a fundamental law—“when neurons are organized in certain ways, they generate emergent supervenient mental properties”—may be our best explanation of consciousness. (Such a proposal is certainly more explanatory than substance dualism.) (3) There is no reflexivity problem. While it is true, on naturalism, that everything supervenes on the material world, relationships (including supervenience) are not things in the same way that objects and events are—things that, naturalists suggest, do supervene on the material world. Reppert is engaging in a bit of verbal sleight-of-hand here by equivocating on the word “thing.”

3. Property Dualism Really is a Naturalist Theory

Reppert claims that I was unclear what I meant by an “ultimate explanation.” But it could not be clearer that I meant exactly what he means: “explanations at the most basic level of analysis.” My main argument was that both identity theory and property dualism, if true, entail that mental explanations exist at the most basic level of analysis; property dualism, for example, does so by suggesting that the mental is a causally efficacious property of the material world. And this is true, despite the fact that some naturalists (like Andrew Melnyk) are not property dualists. Now, Reppert does argue that property dualism is not a naturalistic theory, because that would violate his second tenet of naturalism, causal closure, by suggesting that mental properties have causal powers. If true, this undermines my point that property dualism is an example of an epistemically consistent naturalist theory; and I will grant that some philosophers understand property dualism as a non-naturalist theory. But this is not the proper way to understand it; after all, its most common contemporary varieties are emergent materialism and non-reductive physicalism. The property dualist can consistently be a naturalist because the naturalist commitment to causal closure simply implies that nothing outside the material world causes anything to happen in the material world (because nothing beyond the material world exists). Of course, if a naturalist thinks that the material world only has physical properties, then that naturalist will think mental causation violates causal closure. But by including mental properties 116 DAVID KYLE JOHNSON among the kinds of properties that material objects have, the property dualist’s suggestion that the mental has causal powers in no way violates causal closure, because it does not entail that something from outside the universe is causing anything to happen. If both physical properties and mental properties are part of the material world, something mental (for example, a decision) causing an event does not violate causal closure any more than something physical (for instance, an electron’s charge) does. The property dualist can even claim, as Reppert suggests that the naturalist must, that ultimately everything will be explicable in terms of a single set of laws; it is just that, for the property dualist, those laws will govern both kinds of properties that matter has. In reply, Reppert might insist that property dualism is not a naturalistic theory because it also violates his first tenet of naturalism by including mental properties at the basic level of analysis. But this simply reveals that he has chosen a (non-standard) definition of naturalism to load the dice in his favor—to preclude naturalism from doing the one thing he says it must do. In reality, this tenet does not actually express a necessary component of naturalism (which just is the view that nothing exists beyond the natural world). Not only is naturalism not necessarily mechanistic (a possible world consisting of only chaotic matter would still be naturalistic), but neither a mechanistic world nor a naturalistic world would preclude mentality from existing at the basic level of analysis—as property dualism and identity theory reveal. Simply put, the mental is not necessarily supernatural, as Reppert seems to assume. Property dualism is not committed to epiphenomenalism either. In fact, most property dualists like the theory because it allows them to reject epiphenomenalism and embrace mental causality. If you read Reppert’s quote from Feser carefully, you will see that Feser is not arguing that property dualism entails epiphenomenalism, only that property dualists who also embrace epiphenomenalism are committed to a counterintuitive thesis.

4. Reppert Drew Irrelevant Distinctions and Offered a Bad Analogy

Reppert challenges my claim that, on identity theory, mental explanations are ultimate, since, on identity theory, mental explanations are numerically identical to physical brain explanations (which are ultimate explanations). Reppert apparently misunderstood either my argument or identity theory, because he thinks this claim is easier to defend on type-identity theory than on token-identity theory. But the distinction between these two theories is irrelevant to my point. Regardless of whether each type of mental event is identical to a certain single type of brain event, or can be identical to multiple types of brain events, it is still the case—on both theories—that each instance of a mental event is identical to some instance of a brain event; thus, on both Rely to Victor Reppert 117 theories, every explanation in terms of a brain event is, ipso facto, an explanation in terms of a mental event. So, both the token and type theory of identity are theories in which mental explanations are ultimate. Reppert’s clarifying analogy is also flawed. Consider the baseball that broke his window. Of course, the fact that it is numerically identical to a baseball that was once hit by Luis Gonzalez is not what caused it to break the window. But that is a nonessential accidental property of the ball; its physical properties would not change if it was not hit by Gonzalez. This is not true of the mind and brain’s relationship on identity theory. Because of their identity, a mental event cannot be different, or lacking, without the corresponding physical brain state being different than it is, which would alter its causal powers.

5. Reppert Missed the Relevance of Nonmental Forms of Reasoning

Reppert thinks the existence of unconscious human reasoning is irrelevant; all his argument “needs” (requires? assumes?) is the fact that we occasionally go through explicitly mental reasoning processes. But that is simply false; his argument does not merely require or assume that we sometimes consciously reason. His argument explicitly claims that purely physical, nonmental, processes—the kind that he says naturalism insists are ultimately explanatory—cannot reliably produce true beliefs. The fact that reliable unconscious (that is, nonmental) belief-forming processes occur in the physical brain is direct empirical evidence that this claim is false. I made a similar point regarding the unconscious reasoning processes of computers to which Reppert also objected. Ironically, however, Reppert seems to have overlooked the fact that I had already anticipated and responded to his objection. I wrote:

One might object that such processes do not produce true beliefs because Deep Blue and Watson do not believe anything—they are computers without mentality. We can easily avoid this objection by noting that we could theoretically hook a person up to these computers in such a way that they reliably produce true beliefs in that person without that person going through any mental ground-consequent process. . . . Clearly, in such a scenario, true beliefs would be reliably produced without any mental effort at all.

But perhaps that did not make my point clear enough. So consider Descartes’s claim that a machine “could never use words, or put together other signs, as we do in order to declare our thoughts to others'' (Descartes, 1985, p. 140). For Descartes, decisions regarding how to use language happen in the soul, an immaterial entity able to carry out processes that cannot be 118 DAVID KYLE JOHNSON performed by a (soulless) physical mechanism. Clearly, a machine that unequivocally passed the Turing Test would falsify Descartes’s claim. Even if you did not think such a machine “understood language” (because it lacked mentality), it is still using words in meaningful ways. Indeed, even what computers can do with language today should be impossible if the use of language could only be accomplished by a soul. Likewise, a computer that reliably generates true conclusions falsifies the claim that mechanistic processes cannot do so. As with Descartes, it is Lewis’s substance dualism that drives his suggestion that mechanistic processes cannot reliably generate true conclusions. For that, you need mentality—a soul. But although we have yet to produce a machine that can reason exactly like a human, even what we have accomplished so far with computers should be impossible on the theory of dualism that lies behind Lewis’s claim. If reasoning truly is a non-mechanical process, it should not be possible to program a computer to figure out anything—especially something as complicated as chess moves or Jeopardy! responses. Even if you do not think such computers have mentality, they are still generating conclusions. Consider a future in which a computer-brained android is created that duplicates human behavior in every respect, including human reasoning. Let us call him Data and suppose that he even recounts ground-consequent reasoning as he explains how he arrived at his (extremely reliable) conclusions. Of course, we would know that Data only goes through a mechanical process to arrive at his conclusions; we designed him. But will Reppert still insist that only mental operations can reliably produce true conclusions? If so, Reppert’s position is irrationally unfalsifiable—nothing could count as evidence against it. After all, Data is doing exactly what you do to make me think that you reliably produce true conclusions. Yet if such an argument should be applied to Deep Blue and Watson, it should be applied to Data.

6. Reppert Assumes the Truth of What He is Trying to Prove

I claimed that, as a refutation of eliminativism, the argument from reason would simply beg the question by assuming the truth of what eliminativism rejects: folk psychology. Reppert suggests, however, that the argument from reason does not assume folk psychology; instead it argues that “the truth of propositional-attitude psychology [that is, folk psychology] is necessary for, for example, scientific knowledge.” Of course, that is true by definition; knowledge is justified true belief, and beliefs do not exist unless folk psychology is true. So, unless folk psychology is true, there can be no scientific knowledge. But one can find this conclusion troublesome only if one thinks that scientific knowledge is possible; and one can think scientific knowledge is possible only if one already embraces folk psychology. So Rely to Victor Reppert 119

Reppert proves with his own words that, indeed, his argument assumes folk psychology and thus begs the question against eliminativism. A similar point applies to Reppert’s comments on epiphenomenalism. Epiphenomenalism suggests that mental causation is an illusion, and thus that rational inference—the notion that our drawing of conclusions is the result of the mental recognition of logical relationships—is an illusion. They explicitly deny that the natural sciences are based on rational inference; that is a part of the theory. To object to the theory because it “implies that the kind of rational inference on which the natural sciences rely does not occur” simply begs the question. To think that the sciences rely on rational inference, or to be bothered by the fact that they do not rely on it, one would already have to reject epiphenomenalism. By convincing someone that they believe in mental causation (or folk psychology), you can convince them that they are not an epiphenomenalist (or eliminativist); but you cannot defeat such a theory with an argument that merely assumes what it denies.

7. Conclusion

I have shown that Reppert made six different mistakes, each of which is sufficient to render his defense of the argument from reason a failure. Reppert defends the argument from reason so vehemently, I believe, because he is a supernaturalist—more specifically a substance dualist (or at least, so it seems). As such, he thinks naturalism unfairly enjoys “a kind of absolute presupposition” that leads naturalists to dogmatically assume that there are materialistic solutions to the “fatal problems” that dualistic arguments encounter. Of course, this is false. Although the vast majority of philosophers accept one naturalistic theory of the mind or another—only 27% or so even lean towards non-physical interpretations of the mind (Bourget & Chalmers, 2013)—materialism is not merely presumed; it is accepted because of overwhelming evidence in its favor. What is more, such philosophers do not just assume that there is a satisfactory materialistic theory of the mind; they seek one because non-materialistic theories of mind have failed so completely (Johnson, 2013). Substance dualism fails as a philosophical theory, we might say, just as the argument from reason fails in its attempt to defend it.

Nine

PRO: THE MORAL ARGUMENT IS CONVINCING

David Baggett

In Book I of Mere Christianity, C. S. Lewis offers a version of the moral argument for God’s existence. Moral arguments for God come in many variations, and have been offered by a number of major thinkers, ranging from Kant to Newman to Robert Adams. Lewis’s variant was a popularized version, owing to its original presentation as BBC radio addresses during the Second World War. Here I will lay out Lewis’s argument, amplifying the discussion just a bit at a few points as needed, and then briefly address three major challenges to the argument. Perhaps it would be useful to say at the outset that if one is looking for a strict proof of God’s existence, Lewis’s argument is bound to disappoint. We can think of a proof as a rigorous deductive argument featuring premises we know or all reasonable people would affirm to be true. Few arguments in philosophy are able to satisfy such stringent requirements, and that is why philosophical arguments rarely constitute proofs. Proofs tend to be associated with certainty, something that most philosophers have realized is elusive. Even if the argument is not a proof, however, it might still be a good argument, in the sense of providing rationally persuasive evidence in favor of the conclusion. I shall argue that Lewis’s moral argument is a good argument in this latter sense.

1. The Law of Human Nature

Lewis begins by identifying what he calls the “law of human nature,” which he thinks is on clear display whenever people quarrel about proper behavior. In such situations, people invariably find themselves saying things like “How’d you like it if anyone did the same to you?” or “That’s my seat, I was there first” (Lewis, 1952, p. 17). Educated and uneducated, young and old alike, speak in such terms, and clearly are not simply saying that the other person’s behavior does not happen to please him. Rather, they are appealing to an objective standard of behavior that both parties are presumed to accept. Rarely, Lewis notes, does a person accused of ethical misbehavior say, “To hell with your standard.” Nearly always, the person “tries to make out that 122 DAVID BAGGETT what he has been doing does not really go against the standard, or that if it does there is some special excuse” (Lewis, 1952, p. 17). So Lewis observes that both parties in ethical quarrels seem to have in mind some kind of Law or Rule of fair play or decent behavior or morality about which they really agree. What distinguishes humans quarreling, in fact, from animals fighting, is that “quarrelling means trying to show that the other man is in the wrong,” which presupposes some measure of agreement about what is right and wrong (Lewis, 1952, p. 17). This Law of Right and Wrong, as Lewis calls it, is what he means by the “Law of Human Nature.” Our bodies are governed by the law of gravitation, but we are also governed by the law of human nature, except that the latter, unlike the former, can be disregarded. Like all other things in the natural world, we have no choice about “obeying” gravitation, but we can choose to obey or disobey the moral law. Lewis attributes the old label for the fundamental moral law—“Law of Nature”—to the fact that people long thought that everyone knew it by nature and did not need to be taught it. Many ethicists have thought that the human idea of decent behavior is obvious to everyone. And Lewis is inclined to concur, insisting that otherwise much that was said about Nazi wrongdoing would have been nonsense—and Lewis was on pretty solid ground in those radio talks in assuming that most of his audience entertained rather strong convictions about Nazi culpability. Lewis’s first point, then, is that most everyone assumes that there is an objective, binding, authoritative moral law. Some actions (and presumably other things too, like motives, but we will focus on actions) are right, and some are wrong, objectively speaking. At this point Lewis anticipates an objection from the ranks of relativists, those who wish to relativize morality to culture. Usually relativists begin with the diversity thesis, the claim that moral beliefs and practices have varied widely throughout the world and human history. Ethical relativism, however, is the idea not merely that moral beliefs and practices vary from one culture to another, but that actual moral truth itself is relative. Something may be ethically right in Culture A but not in Culture B. How diversity of ethical belief can be thought to lead to ethical relativism depends on the sort of relativist we consider. Some might wish to argue that the best explanation of the diversity thesis is ethical relativism; others might think that diversity in moral belief causes diversity in moral truth; yet others might think it is a semantic or logical connection, trading on an equivocation of the term “moral” as meaning both moral belief on the one hand and moral truth on the other. Lewis’s main response to the relativist challenge, however, goes after the diversity thesis itself, which many might think is the least vulnerable part of the challenge. Lewis’s point is not that no such diversity exists; rather his suggestion seems to be that there is not nearly so much diversity as is Pro: The Moral Argument is Convincing 123 sometimes thought. In a famous passage he reflects on the issue of moral diversity among different civilizations and ages like this:

There have been differences between their moralities, but these have never amounted to anything like a total difference. If anyone will take the trouble to compare the moral teaching of, say, the ancient Egyptians, Babylonians, Hindus, Chinese, Greeks and Romans, what will really strike him will be how very like they are to each other and to our own. Some of the evidence for this I have put together in the appendix of another book called The Abolition of Man; but for our present purpose I need only ask the reader to think what a totally different morality would mean. Think of a country where people were admired for running away in battle, or where a man felt proud of [double-crossing] all the people who had been kindest to him. You might just as well try to imagine a country where two and two made five. Men have differed as regards what people you ought to be unselfish to—whether it was only your own family, or your fellow countrymen, or everyone. But they have always agreed that you ought not to put yourself first. Selfishness has never been admired. Men have differed as to whether you should have one wife or four. But they have always agreed you must not simply have any woman you liked (Lewis, 1952, p. 19).

Lewis is arguing that we should not miss the forest for the trees, getting so caught up in the real moral differences there are between cultures and civilizations that we overlook the real moral points of resonance and deep agreement that exist. There are ever so many other things that could be said against the thesis of ethical relativism, but having argued that the diversity thesis on which relativism tends to be predicated is usually overstated, Lewis is content to move on. It is worth noting, perhaps, that Lewis may be wrong to suggest that selfishness has never been admired. Some ethical egoists seem to extol regard for self-interest as the very essence of morality, and although self-interest and selfishness can be distinguished, many of the things that could be said in defense of enlightened self-interest as a legitimate moral motivation could apply to selfishness, broadly construed. At any rate, what strikes Lewis as most remarkable in this discussion is that true moral skeptics are rare. “Whenever you find a man who says he does not believe in a real Right and Wrong, you will find the same man going back on this a moment later” (Lewis, 1952, p. 19). This brings to mind the occasional teacher who, to demonstrate this lesson, assigns a failing grade to an earnest student who has written a paper defending ethical relativism. When asked for an explanation, the professor then matter-of-factly explains the grade as his personal preference requiring no further explanation or 124 DAVID BAGGETT justification. It is the rare student in such a situation who is content with so relativistic an explanation. Lewis’s argument against relativism, then, seems to be a rejection of the main argument in its favor—an argument based in one way or another on the diversity thesis. He seems to be right that such arguments in favor of ethical relativism are weak, but he could be pushed to say more about what makes ethical relativism false. That the main argument in its favor fails does not show that it is false. This is where arguments actually against ethical relativism can come in handy, one of which will be discussed shortly, but for now we will leave ethical relativism behind. One reason for doing so is that my interlocutor in this volume, Erik Wielenberg, also rejects ethical relativism, and it would thus be more fruitful to move on to other topics. This brings us to the second part of Lewis’s argument. Having argued for the reality of an objective moral law, Lewis then suggests that none of us is really keeping it. His point is not a judgmental one that applies to others; he is quite willing to admit that he belongs among the moral lawbreakers. He thus identifies what he considers to be a rather obvious fact: that “this year, or this month, or, more likely, this very day, we have failed to practice ourselves the kind of behavior we expect from other people.” He recognizes that all of us have long lists of excuses for our moral failings, and he sets aside whether such excuses are good or not. “The point is that they are one more proof of how deeply, whether we like it or not, we believe in the Law of Nature” (Lewis, 1952, p. 21). No, Lewis says, we believe in it profoundly, feeling its force and authority, and trying to hide our shame for failing to live up to it. He concludes: “These, then, are the two points I wanted to make. First, that human beings, all over the earth, have this curious idea that they ought to behave in a certain way, and cannot really get rid of it. Second, that they do not in fact behave in that way. They know the law of Nature; they break it. These two facts are the foundation of all clear thinking about ourselves and the universe we live in” (Lewis, 1952, p. 21).

2. Instincts and Social Conventions

Lewis pauses at this point to answer a few objections and to clarify some things about the law of human nature, beginning with the issue of whether what Lewis calls the moral law is simply herd instinct. Lewis acknowledges that we find ourselves on occasion with strong desires to act in certain ways— including helpful or altruistic ways—and that these desires may be rooted in basic human instincts; but he distinguishes a desire to act, on the one hand, from the sense or feeling that one ought to help whether one wants to or not. The desires describe various impulses we may feel, some other-regarding and doubtless others self-regarding; but the moral law adjudicates between these Pro: The Moral Argument is Convincing 125 sometimes conflicting desires and confers authority on some over others, thereby prescribing how it is that we should behave. As Lewis puts it: “The Moral Law tells us the tune we have to play; our instincts are merely the keys” (Lewis, 1952, p. 22). Lewis adduces as an additional piece of evidence that the law of human nature is more than instinct the fact that the moral law often counsels that we heed the weaker of two impulses and strive to make the weaker, other- regarding impulses and instincts within us stronger than they are. He takes this as evidence that whatever is doing the work in such cases is not merely our natural instincts themselves. We are not acting from instinct when we try to make an instinct stronger than it is, Lewis says, adding: “The thing that tells you which note on the piano needs to be played louder cannot itself be that note” (Lewis, 1952, p. 23). For Lewis, a third reason to think the moral law is not one of our instincts is that there is no one instinct we have that is always good and in accord with the moral law. Every universal impulse, we might say, is sometimes approved in its expression by the moral law and sometimes disapproved. Sexual impulses, patriotism, familial bonds—each of these can in some circumstances be good and in others bad. Extending the musical analogy once more, Lewis writes:

Think once again of a piano. It has not got two kinds of notes on it, the ‘right’ notes and the ‘wrong’ ones. Every single note is right at one time and wrong at another. The Moral Law is not any one instinct or any set of instincts: it is something which makes a kind of tune (the tune we call goodness or right conduct) by directing the instincts (Lewis, 1952, p. 23).

Lewis accentuates the insight by pointing out that giving pride of place to any single impulse, however noble, can in the end become treacherous; deeds done in the name of love of humanity, for example, but without justice are bound to become cruel and hideous. If morality is something more than merely natural instinct, though, might it be simply an echo of what our parents and teachers and peers have taught us—that is, merely a learned social convention? Here Lewis suspects we may be misled by a false assumption about what sorts of things can be learned. Among things that we learn, some are merely invented and conventional, like which side of the road to drive on, but others are rooted in unchangeable objective reality, like the multiplication tables. So which category does the moral law fit in? Is it mere convention or, as Lewis puts it, a “real truth”? He gives two reasons to think morality is more like objective mathematics than conventional driving rules. First, he harkens back to the earlier point that there is a great deal of resonance when it comes to moral rules throughout the world and human history, exceptions notwithstanding. The exceptions—the moral differences 126 DAVID BAGGETT we find—can potentially be explained by a range of phenomena, such as moral immaturity, mistaken worldviews, wishful thinking, human obstinacy, and irrationality. What is more astounding about morality, though, are the vast swaths of convergence of perspective, especially on what we might call the core moral convictions of which we find evidence almost universally. Sociologist Tony Campolo once noted how amazed he was at the degree of moral consensus evident at United Nations meetings and in the United Nations’ Declaration of Human Rights—that despite vast differences in politics, culture, and history, a shared moral framework was invoked to support basic human rights and to condemn such evils as racism, sexism, and dehumanization. Perhaps, he suggests, cultural anthropologists have become so enamored with exotically peculiar cultures that they fail to see what is generally true for the human race. Second, Lewis mentions moral progress, and how moral theory needs to carve out room for so important a category. Earlier we pointed out that rejecting the main argument in favor of ethical relativism is not the same as refuting it. More argument is needed, and here the inability of ethical relativism to make room for important categories like moral progress can be enlisted. This example can do double duty, serving as evidence against both ethical relativism and, as Lewis intends it here, the notion of morality as a mere social convention as well. For, even if someone were to insist that morality can be kept relativized at the level of whole cultures, this would mean that no culture is ever wrong. When a majority of Americans believed that the most egregious and dehumanizing forms of human slavery were morally permissible, then that was “true for them,” as relativists are wont to say. The category of moral progress would be lost on this analysis, though, and surely one of the most important functions of moral theory is to make appropriate room for so crucial a category. Since the United States was “right” to endorse slavery and “right” later on to denounce it, it was right all along and never wrong, so where was the moral progress? Most people take it as axiomatic that the rejection of slavery did mark moral progress, and take such moral recognition as patently obvious, which likely shows that there is something radically wrong with moral relativism. Similarly, suppose moral convictions are reduced to social conventions. Then, too, moral progress is impossible. As Lewis notes: “The moment you say that one set of values can be better than another, you are, in fact, measuring them both by a standard, saying that one of them conforms to that standard more nearly than the other. But the standard that measures two things is something different from either” (Lewis, 1952, p. 25). Lewis concludes this part of his discussion by pointing out that some apparent moral disagreements are really just disagreements about facts. The illustration he uses is the fact that we no longer put witches to death. Lewis suggests that the real difference between us and our medieval forbears on this issue is that we believe that there are no genuine witches (that is, people who Pro: The Moral Argument is Convincing 127 have made a pact with the devil and possess malignant super-powers as a result) and they believed that there are. As Lewis notes, “It may be a great advance in knowledge not to believe in witches: there is no moral advance in not executing them when you do not think they are there. You would not call a man humane for ceasing to set mousetraps if he did so because he believed there were no mice in the house” (Lewis, 1952, p. 26).

3. The Reality and Foundation of the Moral Law

Why does Lewis characterize the moral law to which we fail to comply as odd? In one sense of course it is the most natural thing in the world to recognize that human beings are imperfect and fall short of moral ideals. Lewis’s intent, though, is not to fix blame. Rather, he considers it evidentially significant that such a thing as the moral law would exist in the first place. The very possibility of moral imperfection, of our not being what we ought to be, strikes Lewis as interesting and telling. Here’s why. When it comes to the physical laws of nature, they may really not be laws at all, but simply a description of how the natural world operates. It is not as if a falling stone decides to obey the law of gravitation. Such laws are rather descriptions of what nature in fact does. In contrast, however, the moral law assuredly does not amount to what human beings in fact do, for we fall short of the moral law all the time. All too often the moral law prescribes that we ought to do something that we fail to do. So unlike inanimate objects acting in complete conformity with the laws of nature, the moral picture involves the descriptive matter of what we actually do and the prescriptive matter of what we ought to do. How we behave is not the whole story, for another important part of the picture is how we ought to behave. This is indeed peculiar. In fact, modern language of moral obligation and the moral law, G. E. M. Anscombe once argued, is rather different from Aristotle’s teleological understanding of morality; and she attributed the difference to the influence of Christianity and its notion of morality rooted in the Torah. This accounted, she argued in her classic “Modern Moral Philosophy,” for the law-like connotations associated with moral oughtness and obligation. She further saw that naturalistic ethicists had tried to retain the atmosphere of such locutions but without the theological foundations that gave them substance, leading to all manner of mischief. Despairing over the prospect of a return to a law-based ethic, she thought modern ethicists should eschew such obligation talk and return to doing ethics in the spirit of Aristotle (Anscombe, 1958). Moral language today is so peculiar, in fact, that Lewis suggests that this is why many people try to explain it away. Some attempt to reduce moral 128 DAVID BAGGETT impropriety to an instrumental matter—as we do with a tree that, for our purposes, does not shade us well and is, for this reason and in this sense, a “bad tree.” But moral misbehavior is not a matter of personal inconvenience. There is a world of difference between two behaviors, each equally inconvenient to us, but one morally wrong and the other morally permissible. This shows that personal inconvenience is not sufficient for an action to be immoral. Nor is it necessary. A wartime traitor on the other side can prove most convenient, but this hardly exonerates him from being “human vermin” (Lewis, 1952, p. 28). Not infrequently morality requires us to forego all sorts of behaviors we might enjoy or to endure ones we despise. Does morality, more generally perhaps, pay the human race as a whole, even if it fails to pay each individual? Morality conduces to social harmony, after all, and it would seem that it functions well as a kind of social glue holding a society together. Lewis notes that as an explanation of the moral law, this misses the point. For why do what is good for society? What makes that something to do? Morality requires that we care about social harmony, true enough, but now we are just arguing in a circle. What explains the moral law? That it conduces to social harmony? But why care about social harmony? Because it is the moral thing to do? As Lewis puts it, we are saying what is true, but not getting any further (Lewis, 1952, p. 29). Caring about society and trying to benefit it is part of what the moral law consists in. So Lewis thinks reflection on our moral experience provides good reason to believe that there is something above and beyond the ordinary facts of human behavior, a real law not of our own creation that presses on us, wielding a binding, prescriptive power over us. What is still needed is an explanation of the foundation of the moral law. What lies behind the law? If morality provides evidence, the question becomes: evidence for what? To that question Lewis now turns. Lewis considers morality to have something to offer in our effort to figure out the nature of reality and the meaning of it all. He begins by distinguishing between two important worldviews: a materialistic or naturalistic worldview on the one hand, and a theistic or religious worldview on the other. A materialist, Lewis says, takes matter and space to have always existed for reasons beyond our ken. Matter, then, behaving in certain fixed ways, ultimately produces creatures like us able to think and reason (Lewis, 1952, p. 31). The religious vision of reality, in contrast, affirms that there is something more like a mind than anything else behind the universe— something conscious, with purposes and preferences. This mind was responsible for creating the universe, partly for reasons beyond our understanding, but partly for producing creatures like itself—creatures with minds (Lewis, 1952, p. 32). Ever since human beings have pondered ultimate questions, both views have turned up. Moreover, Lewis argues that science alone cannot adjudicate the conflict. For the question of whether there is anything outside the purview Pro: The Moral Argument is Convincing 129 of science is not itself a scientific inquiry. Science by its nature confines itself to an examination of the physical universe; whether there is anything beyond the universe must be determined by other means. Exhaustive scientific knowledge still leaves questions unanswered like, “Why is there a universe?” or “Has it any meaning?” Hope for answering such questions, though, comes from, of all things, ourselves, because we are the one and only thing in the whole universe that we know about from the “inside,” as it were. When we look at our moral experience, we apprehend a moral law not of our own creation that we recognize we ought to obey. So in our quest to know whether the universe is all that exists, on the one hand, or whether there is a power behind it that makes it what it is, on the other, we have to ask if there is reason to believe there is something more than the facts in principle observable within the universe. Morality reveals that there is. Perhaps Lewis overstates it when he says: “The only way in which we could expect it to show itself would be inside ourselves as an influence or a command trying to get us to behave in a certain way” (Lewis, 1952, p. 33). But his point can go through even if this is not the only way. An objective moral law that tells us how we ought to behave reveals, in this one case where we have direct, inside information, that there is something more than the observable facts. For this reason, Lewis takes morality to be better explained by theism than by naturalism. He cautions against moving too fast here, even though plenty of people, with some justification, suggest that this is a pitfall to which he himself has succumbed, a charge we will have occasion to discuss momentarily. But Lewis suggests anyway that the tentative conclusion he feels justified in drawing is that morality provides evidence for something like a mind behind the universe that explains the moral law. “I am not yet within a hundred miles of the God of Christian theology. All I have got to is a Something which is directing the universe, and which appears in me as a law urging me to do right and making me feel responsible and uncomfortable when I do wrong” (Lewis, 1952, p. 34). Why a mind? Because the notion of mere matter giving authoritative, prescriptively binding moral direction strains credulity. Perhaps the most notable recent attempt to come to terms with and address this challenge from an atheistic perspective is Thomas Nagel’s Mind and Cosmos (2012). Interestingly Nagel echoes J. P. Moreland’s reservations about naturalism while resisting theism despite its explanatory power for reasons Lewis presciently anticipated and addressed in Miracles (Baggett, 2013).

4. Old Time Religion?

Lewis now anticipates someone crying foul by saying that he has been doing something that looks like philosophy by identifying interesting and important 130 DAVID BAGGETT phenomena in need of explanation, and ending up with something that conspicuously sounds like religion, which has been tried and found wanting. To this Lewis replies in three ways. First, he acknowledges we all want progress, but progress means getting nearer our destination. If we have taken a wrong turn, the most progressive among us will be the first to turn around and make the needed course correction. In general, though, Lewis thinks talk of being progressive too often conceals a priority other than truth. His second point is that his analysis so far is still quite a distance from the God of any actual religion, still less of Christianity. He has simply been trying to see what can be accomplished “on our own steam.” What he thinks we have discovered are two pieces of evidence: The first is the universe itself, which would show the Somebody or Something behind the moral law to be a consummate artist, but potentially not very friendly at all. The other piece of evidence is the moral law itself, which provides some inside information. In a real sense, Lewis thinks that this is the better evidence, revealing to us more about this mind (or something like a mind) behind reality. “You find out more about God from the Moral Law than from the universe in general just as you find out more about a man by listening to his conversation than by looking at a house he has built.” What we find out is that the Being behind the universe is “intensely interested in right conduct—in fair play, unselfishness, courage, good faith, honesty, and truthfulness” (Lewis, 1952, p. 37). C. Stephen Evans has argued something similar in recent years:

The most well-known use of the [moral] argument is found in C. S. Lewis’s amazingly popular Mere Christianity. . . . [I]n my view, the two theistic natural signs that lie in our moral experience [the intrinsic value of human beings and moral accountability], and that help make moral arguments for God plausible, not only point to God in a more powerful way [than do cosmological and design arguments], but also offer more insight into God’s nature. For they point to a God who is essentially good, and who desires a relationship with human persons. Someone who comes to know God as the ground of moral obligation does not know merely an abstract metaphysical reality but a personal reality who makes a claim on the life of humans. Someone who knows God as the source of the inherent value that is present in human persons also knows God as a personal being, and as the one who pre-eminently possesses the worth and value we see in human beings (Evans, 2010, p. 109).

As the source of the moral law, God is Himself good. A. C. Ewing argued similarly—that the source of morality must itself be moral. But Lewis does not mean by “good” something soft and sentimental, but rather something hard as nails. “There is nothing indulgent about the Moral Law. . . . It tells you to do the straight thing and it does not seem to care how painful, or dangerous, or difficult it is to do. If God is like the Moral Law, then He is not soft” Pro: The Moral Argument is Convincing 131

(Lewis, 1952, p. 37). Lewis says that some people talk as if meeting the gaze of absolute goodness would be fun, but they should think again. “Goodness is either the great safety or the great danger—according to the way you react to it. And we have reacted the wrong way” (Lewis, 1952, p. 38). Although the argument at this stage, in Lewis’s view, is still a hundred miles from Christian theology, it functions, to his thinking, as a useful prolegomenon to Christianity, which makes no sense until we have faced the sort of facts Lewis has been describing. Christianity makes much of repentance and forgiveness, which make little sense to those unaware of a need to repent or the need for forgiveness. We only seek a doctor when we realize we are sick. Only those who have attempted to meet the demands of the moral law and consistently fallen short come to realize that they lack the resources to meet the standard imposed by morality. This is where Christianity comes to answer the questions that arise. The Good News of Christianity can only come after the bad news that the moral law cannot be obeyed on our own. Lewis concludes the argument with this reflection:

I quite agree that the Christian religion is, in the long run, a thing of unspeakable comfort. But it does not begin in comfort; it begins in the dismay I have been describing, and it is no use at all trying to go on to that comfort without first going through that dismay. In religion, as in war and everything else, comfort is the one thing you cannot get by looking for it. If you look for truth, you may find comfort in the end: If you look for comfort you will not get either comfort or truth—only soft soap and wishful thinking to begin and, in the end, despair (Lewis, 1952, p. 39).

5. Assessing the Argument

In light of the original context of Lewis’s argument, there are bound to be plenty of junctures at which it is vulnerable to criticism. Setting aside vulnerabilities avoidable by charitable interpretations, what are the most substantive challenges that can be raised against the argument? To get a sense of what they are, perhaps it would be useful to identify the overall framework of the argument. Erik Wielenberg, my interlocutor in this volume, has usefully delineated a logical structure of Lewis’s argument. Eschewing John Beversluis’s uncharitable depiction of the argument as a piece of patently flawed deductive reasoning, Wielenberg casts the argument as abductive in nature—an inference to the best explanation. This is a pattern of reasoning widely used in history, science, philosophy, and other disciplines in which explanations are sought by means of nondeductive reasoning. In abductive reasoning, a principled set of criteria is applied to a pool of potential explanation candidates in an attempt to arrive at the best explanation of a 132 DAVID BAGGETT particular phenomenon. Here is how Wielenberg hammers Lewis’s case into discursive format:

(1) Lewisian moral phenomena exist. (2) The best explanation of the existence of Lewisian moral phenomena is the existence of a Higher Power that created the universe. (3) So: There is a Higher Power that created the universe (from 1 and 2). (4) The Higher Power issues instructions and wants us to engage in morally right conduct. (5) If (4), then there is a good, mindlike Higher Power that created the universe. (6) Therefore, there is a good, mindlike Higher Power that created the universe (from 4 and 5) (Wielenberg, 2008, p. 63).

What is meant by Lewisian moral phenomena? Wielenberg deftly plumbs Lewis’s discussion to identify three items in the list: (1) the Law of Nature— the moral law, namely, universal, objective moral truths; (2) moral knowledge, including instances of a priori moral knowledge—with both the psychological component of our moral beliefs and the normative component of warrant or justification obtaining for at least some of those beliefs; and (3) a cluster of moral emotions associated with our knowledge of moral facts, including moral guilt and a sense of moral obligation. As we saw above, Lewis argues that a theistic picture of reality better explains such phenomena than does a naturalistic one. An argument of this nature has at least three large tasks to accomplish if it is going to work. First, it has to argue for, or at least reasonably assume, a picture of moral realism. If morality were entirely a subjectivist, noncognitivist, or anti-realist matter, then the heart of the Lewisian moral phenomena on which the argument is based would be gutted. Second, it has to argue that the theistic picture makes a great deal of sense and constitutes a good explanation. This requires not only that the theistic account be defended against various objections, but that a positive case for it be made. Third, defenders of the argument need to make the case that no alternative explanation is as good, which is, frankly, a formidable job indeed. An abductive approach, by taking on all comers, needs to avoid giving short shrift to any legitimate moral theory in contention. This argumentative strategy requires quite a bit of patience to articulate carefully and does not lend itself to a hasty approach. Wielenberg himself will raise challenges against Lewis’s argument, so allow me to wrap up this first part of the discussion by saying just a bit about each of these three stages of the argument. What I think is rather clear is that Lewis at best sketches but the first steps in all of these areas. The challenge of Pro: The Moral Argument is Convincing 133 moral anti-realism, for example, is an important one, and a challenge increasingly in vogue. The stance among atheists of denying objective moral facts remains, if anecdotal evidence carries much weight, a minority position. More typically when we ask atheists if they think torturing children for fun is wrong, they will say, “Of course! And I don’t need God to tell me it is.” Then they try to back up their moral convictions with some secular ethical theory or other. But if Nietzsche was right, the number of such moral deniers will rise, which we may be witnessing already. Such moral deniers typically claim to see with great clarity that without God, as Dostoyevsky put it, everything is permissible; they are also willing to bite the bullet and follow the implications of their view. In their own way they are trying to follow the evidence where they think it leads, and there is something laudable about that. One of the big reasons why such thinkers—from Richard Joyce to Joel Marks—are leaning in this direction of moral denial is an evolutionary challenge to ethics. It goes something like this: Evolution, especially evolutionary moral psychology, offers a good explanation of our moral beliefs and feelings that requires no commitment to moral truth. They then say that since there is no other evidence for moral truth than our moral beliefs and feelings, which can be explained without appeal to moral truth, we should give up belief in moral truth and become moral skeptics. The way evolutionary moral psychology goes about explaining the formation of moral beliefs gets complicated, but it uses game theory and evolutionarily stable strategies and all sorts of things to make the case that through evolution we have formed our basic moral responses and convictions. It is always possible that our moral convictions somehow track moral truth, but that seems unlikely. And so, again, this leads to moral skepticism and in fact to moral denial. How might one respond to this evolutionary challenge to ethics? Let us suppose, as Lewis did, that evolution is true. In addition, suppose that evolutionary moral psychology in fact can be shown to be highly effective at explaining the formation of our moral beliefs. Does this provide evidence to suggest that there is no moral truth? I will argue that the answer is no, for two reasons. The first point to emphasize is what I call the “primacy argument.” The moral deniers are investing tremendous confidence in a complicated narrative that suggests that morality is not true, but just an illusion. Even the atheist Thomas Nagel has suggested that such elaborate stories are too filled with guesswork to invest us with much confidence in their conclusions. The atheist Louise Antony has admitted as much, saying that any argument for moral skepticism will be based on premises that are less obvious than the existence of objective moral values themselves. Second, the moral skeptic’s argument is self-defeating. Skepticism is a greedy Mistress, and she does not rest content with spoiling our morals. Much 134 DAVID BAGGETT of our entire knowledge base would be undermined, not just our moral claims, because the very conclusions of our deliberations are the result of events that are inevitably caused to happen in our brains by the inexorable operation of the laws of the physical world. So there is something ironic and self-defeating about the moral skeptics: they claim to see clearly the implications of naturalism when it comes to morality, but they fail to see its implications for consciousness, cognition, free will, and rationality. Their argument shows too much; if it is right, we lack grounds for believing it. It is thus self-defeating. Other reasons against moral skepticism could be adduced, but that is enough for now. I do not pretend that this is a decisive argument against moral anti-realism. In the limited space available, though, it is a start.

6. Fleshing Out the Theistic Story

The second and perhaps stiffer challenge for the moral apologist is fleshing out the theistic alternative, defending it against objections, and making a positive case for it. Since this task is far bigger than anything that we can undertake here, let us focus on Lewis’s particular formulation and a potential strength and weakness or two. Wielenberg identifies an objection to the moral argument that came from Bertrand Russell’s 1927 piece “Why I Am Not a Christian.” Wielenberg identifies a central Russellian claim (RC) to be this:

(RC) The only way a being (even God) can be good is by conforming its actions to a moral law of which it is not the author (Wielenberg, 2008, p. 65).

Since Lewis, in Mere Christianity, maintains that God is good and is the ultimate source of objective rightness and wrongness, Russell’s objection strikes at the heart of Lewis’s analysis. Wielenberg notes that in his 1943 essay “The Poison of Subjectivism,” Lewis describes the view that God is “the mere executor of a law somehow external and antecedent to His own being” as “intolerable” (Wielenberg, 2008, p. 79). This means, Wielenberg adds, that “it is incumbent upon Lewis to provide an alternative account of the nature of God’s goodness: How can God be good if not by conforming to a moral law of which He is not the author?” (Wielenberg, 2008, p. 65). Wielenberg thinks that Lewis’s writings suggest three distinct answers to this question. In “The Poison of Subjectivism,” Lewis wrote that “God neither obeys nor creates the moral law. The good is uncreated; it never could have been otherwise; it has in it no shadow of contingency; it lies, as Plato said, on the other side of existence” (Wielenberg, 2008, p. 80). Wielenberg takes Lewis to be suggesting, among other things, that “the good” and “the moral law” are one and the same. Wielenberg rejects Steve Lovell’s inference that this requires an identity relation between God and an abstract object, opting Pro: The Moral Argument is Convincing 135 instead for the idea that Lewis means to equate the good with the moral law. So the first way Lewis suggests that God might be good other than by conforming to an independent moral law is this:

(LA 1) Being identical to the moral law is a way of being good.

Wielenberg identifies two other ways Lewis suggests in his writings that God’s goodness can be understood:

(LA 2) Loving love, fair play, unselfishness, courage, good faith, honesty, and truthfulness is a way of being good, and, from The Problem of Pain,

(LA 3) Desiring that human beings attain genuine happiness (that they freely love God and strive to become Christlike) is a way of being good (Wielenberg, 2008, pp. 66-68).

Wielenberg argues that (LA 2) and (LA 3) are inconsistent with Lewis’s argument against Dualism—namely, that Dualism is internally inconsistent, because if there are two supposedly Ultimate Powers, one good and one evil, there must be an even higher Power that serves as the standard by which these two lesser Powers are judged. Assuming Wielenberg is right, this leaves Lewis with (LA 1). In “The Poison of Subjectivism,” Lewis’s argument for (LA 1) goes like this: “[W]hat lies beyond existence, what admits no contingency, what lends divinity to all else, what is the ground of all existence, is not simply a law but also a begetting love, a love begotten. . . . God is not merely good, but goodness; goodness is not merely divine, but God” (quoted in Wielenberg, 2008, p. 66). What is the problem with (LA 1)? Wielenberg suggests that the proposal that God is the moral law is “roughly as puzzling as the proposal that God is the property of goodness. Insofar as I can understand the claim that God is the moral law, it seems to be the claim that God is identical to a conjunction of ethical facts, facts like: It is morally wrong to torture the innocent just for fun. It is hard to see how a conjunction of such facts could be a mindlike Higher Power, much less the personal God of Christianity. The obscurity of Lewis’s proposal here seems to me to be a serious strike against it” (Wielenberg, 2008, p. 67). Wielenberg also thinks that (LA 1) makes way for an atheistic alternative to theistic ethics, a claim we will take up in the next (and last) section. What should we say about Wielenberg’s charge that (LA 1) is objectionably obscure? An equation of God with the moral law does indeed tread the verge of incoherence, but that reading of Lewis seems unnecessary. What is the ground of all existence, Lewis wrote, is “not simply a law” 136 DAVID BAGGETT

(emphasis added). The suggestion pretty clearly seems to be that God is more than the moral law, even if somehow or other the moral law is bound up in Him. Moreover, saying that God is not merely good, but goodness itself, seems to suggest that we can not only attribute the property of moral goodness to God (at least in some extended analogical sense, and in addition to other forms of goodness, like ontological completeness and the like), but that there is an operative “is of identity.” God is the good, or goodness itself, in some ultimate sense. Although Lewis may not have entirely seen the distinction between the axiological category of the good, on the one hand, and the deontic matter of the right (especially moral obligations), on the other, this can prove a useful distinction in this context. When it comes to moral goodness or the good, God’s very nature may be its locus; when it comes to moral obligations, something else about God (perhaps His commands, or His will, or His desires) may prove its source. It is clear from Lewis’s nonvoluntarist answer to Plato’s Euthyphro Dilemma that he believed strongly that God’s commands are rooted in what is already moral. For example, he wrote, “It has sometimes been asked whether God commands certain things because they are right, or whether certain things are right because God commands them. With Hooker, and against Dr. Johnson, I emphatically embrace the first alternative” (Lewis, 2002, p. 409). But since not everything that is good is obligatory, Lewis and other theists need some effective way of delimiting from among what is good what is also more specifically morally obligatory. The claim is not that Lewis was a divine command theorist; in one sense he clearly was not, but there is more to say. Suppose, following C. Stephen Evans, we distinguish between the discretion thesis and the modal status thesis; even if Lewis rejected the former—the claim that God has prerogatives about which commands to issue—he likely would affirm the latter—that God’s commands would confer on actions a new moral status in some significant sense. If the modal status thesis, but not the discretion thesis, is essential to divine command theory, then Lewis, arguably like Aquinas, could have represented the limiting case of a divine command theory on the nonvoluntarist end of the spectrum. On this view, all of God’s commands could be necessary, reflections of His perfect and loving nature, and yet in virtue of God’s commands the resulting moral law would have a modal status and a prescriptive power and authority it would otherwise lack. Much needs to be said to defend this—and has been—but this would be one account of how the moral law could depend on God and have God for its foundation, but requires nothing like equating a person with a property, a rule, or an abstract object. Interestingly, Brian Leftow suggests that, although God is a person, he may nonetheless fulfill functions typically reserved for abstract objects (Leftow, 1990). Saying God is goodness might seem to elicit the charge of excessive obscurity again. For how is that possible? Advocates of divine simplicity would offer their account, as would neo-Platonists, but for a more recent Pro: The Moral Argument is Convincing 137 account consider Robert Adams’s Finite and Infinite Goods. Wielenberg is familiar with Adams’s work and has interacted with it, but since it came after Lewis he thinks we have little idea what Lewis would think of it. Fair enough, but it is worth pointing out that the thesis Adams defends in rigorous fashion is that God is the Good—goodness itself, as Lewis might put it. Whatever Lewis might think of some of the particulars of Adams’s account, the conclusion is certainly one with which Lewis resonated. That Adams can provide such an account using the rigor of analytic philosophy and leading insights from philosophers of language is good evidence to suggest that affirming an identity relation between God and goodness, whether it is true or not, is at least coherent. I remain unconvinced, therefore, by Wielenberg’s claim that Lewis’s suggestion is objectionably obscure.

7. The Competition

Even if the theistic alternative can be fleshed out and articulated persuasively, the final task for the moral apologist is to argue that no alternative moral theory explains the Lewisian moral phenomena as well as theism does. This is, to my thinking, the biggest challenge of all. It is exactly here that Lewis’s argument is most in need of further development. For an abductive approach requires more than a broad-brushed dismissal of “naturalistic ethics.” Naturalistic attempts to explain morality come in a dizzying array of forms, and there is no substitute for discussing each one carefully and in turn. This cannot be done hastily or in short compass. What I propose to do here, instead, is to follow Wielenberg’s suggestion for a promising direction to look to generate a naturalistic alternative to a theistic ethic, a direction suggested by what we have been discussing: the idea that the moral law is unchanging, featuring necessary truths, truths that simply could not have been false. “The notion that some ethical truths are necessarily true provides the foundation for one kind of atheistic ethical realism,” Wielenberg puts it (Wielenberg, 2008, p. 88). By Wielenberg’s lights, the notion of ethical truths that are necessarily true and are not identical to God is more plausible than the notion that such truths are identical to God, because the former view “lacks the obscurity and possible incoherence of the second view” (Wielenberg, 2008, p. 88). To assess Wielenberg’s analysis, I will confine my treatment to the Lewisian phenomena of moral truth and knowledge, and we will try to pass at least a preliminary verdict on his critique. Once more, saying God is goodness can seem initially obscure, however intuitive it seems to many theists, but perhaps one way to understand it—and I can only begin to sketch out this approach—can come from thinking some more about the nature of necessary truths. Alvin Plantinga, Robert Adams, Thomas Morris, and others have suggested that theism provides a better 138 DAVID BAGGETT account of necessary truths than naturalism does. Consider, for example, the fact that propositions—the bearers of truth values—are the sorts of things that are thought about and can be known. “The idea,” Plantinga writes, “that there are or could be truths quite beyond the best methods of apprehension seems peculiar and outre and somehow outrageous. What would account for such truths? How would they get there? Where would they come from? How could the . . . propositions . . . exist in serene and majestic independence of persons and their means of apprehension? How could there be propositions no one has ever so much as grasped or thought of? It can seem just crazy to suppose that propositions could exist independent of minds or persons or judging beings” (Plantinga, 1982, pp. 67-68). Plantinga continues:

So what we really have here is a sort of antinomy. On the one hand there is a deep impulse towards anti-realism; there can’t really be truths independent of noetic activity. On the other hand, there is the disquieting fact that anti-realism . . . seems incoherent and otherwise objectionable. We have here a paradox seeking resolution, a thesis and antithesis seeking synthesis. And what is by my lights the correct synthesis was suggested long before Hegel. This synthesis was suggested by Augustine, endorsed by most of the theistic tradition, and given succinct statement by Thomas Aquinas: Even if there were no human intellects, there could be truths because of their relation to the divine intellect. But if, per impossibile, there were no intellects at all, but things continued to exist, then there would be no such reality as truth. The thesis, then, is that truth cannot be independent of noetic activity on the part of persons. The antithesis is that it must be independent of our noetic activity. And the synthesis is that truth is independent of our intellectual activity but not of God’s (Plantinga, 1982, p. 68).

Necessary truths, on this account, are those propositions God believes in this and all possible worlds, owing to their being that deeply grounded in reality. Theism provides a compelling reason to think that reality itself contains such truths, and cares about good and evil, right and wrong. The necessary truths, on this view, potentially provide a window into God’s very nature. An atheistic Platonic account, even if it is conceivable in some sense, is not nearly so plausible or principled. The idea that free-floating ontological truths, inert abstract objects, are just sort of out there—without explanation, without foundation—seems by my lights considerably more obscure than a theistic understanding of such truths obtaining because of the noetic activity of God. Adams, moreover, thinks such an account best explains our knowledge of necessary truths: Pro: The Moral Argument is Convincing 139

If God of his very nature knows the necessary truths, and if he has created us, he could have constructed us in such a way that we would at least commonly recognize necessary truths as necessary. . . . This theory is not new. It is Augustinian, and something like it was widely accepted in the medieval and early modern periods. I think it provides the best explanation available to us for our knowledge of necessary truths. I also think that that fact constitutes an argument for the existence of God. Not a demonstration; it is a mistake to expect conclusive demonstrations in such matters. But it is a theoretical advantage of theistic belief that it provides attractive explanations of things otherwise hard to explain (Adams, 1983, p. 751).

An affirmation of necessary moral truths, at the end of the day, strikes me as less a naturalistic view than a quasi-theistic one at the least. As George Mavrodes puts it, a Platonic-sounding world replete with necessary truths seems congenial (especially congenial when compared to some other philosophical views) to a religious understanding of the world (Mavrodes, 1995, p. 587). Lewis noticed the moral law was odd; Mavrodes called binding moral obligations queer, especially in a naturalistic world. If we are talking about the best explanation of such truths, not merely possible ones, theism has the upper hand over naturalism in their estimation. Binding duties and an authoritative moral law prescribing our behavior—while affirming and undergirding intrinsic human values, dignity, equality, and worth—fit into a theistic world more naturally than they do into a materialistic one. Much more would have to be said, of course. But at any rate, affirming the existence of necessary truths reflective of God’s perfect and unchanging nature is a coherent way a theist can flesh out a Lewisian analysis according to which, in a clear sense, God himself is the locus of perfect goodness. Once more, even if it is not true, or even particularly plausible, the charge of incoherence does not stick. The naturalistic alternative by comparison almost seems like a remote promissory note at best, and problematically obscure in the extreme at worst. As John Rist observes:

Plato’s account of the ‘Forms’ (including the Good) as moral exemplars leaves them in metaphysical limbo. They would exist as essentially intelligible ideas even if there were no mind, human or divine, to recognize them: as objects of thought, not mere constructs or concepts. But, as Augustine learned, and as the Greek Neoplatonists had asserted, the notion of an eternal object of thought (and thus for Plato a cause of thought) without a ceaseless thinking subject is unintelligible. Intelligible Forms, never proposed as mere concepts, cannot be proposed as Plato originally proposed them, as free-floating metaphysical items (Rist, 2002, p. 40). 140 DAVID BAGGETT

Obviously, too much has been left unsaid here to draw bigger conclusions about the overall viability of Lewis’s moral argument. But in terms of our ability to flesh it out, fill in the details, defend it against objections, provide positive reasons to take it seriously, and argue for theism’s superior power to explain various moral phenomena, the argument seems to have great potential.

Ten

CON: A CRITIQUE OF THE MORAL ARGUMENT

Erik J. Wielenberg

In light of the admirable accuracy and clarity of David Baggett’s exposition of both Lewis’s moral argument and certain elements of my critique of that argument, I can turn directly to substantive philosophical issues in this first response. Baggett engages with my critique of Lewis’s moral argument (in Wielenberg 2008) in two main ways. First, he responds to my charge that Lewis’s proposal that God is the moral law is obscure by pointing out that contemporary philosophers, most notably Robert Adams, have developed coherent understandings of the claim that God is the Good. Second, he argues that theism can better explain the reality of necessary moral truths than a naturalistic view like mine can. I will address each of these contentions in turn.

1. God as the Good

As Baggett points out, one important element of a defense of a theistic moral argument is making the case that theism provides plausible explanations of central moral phenomena—for example, the existence of moral obligations, or, in Lewis’s terminology, the existence of the Law of Human Nature (or the moral law). Lewis grappled with the relationship between God and the moral law in various writings. His fullest discussion of that relationship is found in his essay “The Poison of Subjectivism.” It is there that Lewis asserts that “God neither obeys nor creates the moral law” (1995, p. 80), and endorses a third option. I continue to believe that the third option that Lewis actually put forward is obscure. In defense of that claim I would point to the fact that Lewis himself says that his proposal is one that we cannot really comprehend, likening our attempt to grasp the nature of the relationship between God and the moral law both to our attempt to grasp the nature of the Trinity and a Flatlander’s attempt to grasp the nature of a cube (1995, 79-80). The problem in all three cases according to Lewis is that we (or the Flatlanders) lack the conceptual resources to represent the relevant aspect of reality both completely and accurately; as he puts it, “our categories betray us” (1995, p. 80). 142 ERIK J. WIELENBERG

Still, I agree with Baggett that Robert Adams’s development of the idea that God is the Good is at least Lewisian in spirit. More importantly, if Adams’s account provides a plausible theistic explanation of human moral obligations it can be used to defend a theistic moral argument of the sort Lewis advanced. Accordingly, I turn now to an examination of Adams’s view. Adams offers theistic accounts of the nature of three central moral entities: (i) the Good, which is a transcendent paradigm of goodness, (ii) the goodness of finite things, and (iii) moral obligation. According to Adams, the Good = God. Adams’s claim that the Good = God is modeled after another identity claim: that water = H2O. One interesting feature of the latter identity claim is that it is not true by definition; the meaning of the word “water” includes nothing about water’s chemical composition, as shown by the fact that people used the word “water” competently before the rise of modern chemistry. One lesson to be drawn from this example is that the meaning of a given term does not always reveal the full nature of the thing to which the term refers (Adams, 1999, pp. 15-16). Adams similarly argues that although the meaning of the word “good” includes nothing about God, it is nevertheless the case that the Good = God. So Adams’s claim that the Good is God is a claim about the nature of the Good rather than the meaning of the word “good” (or the word “God”). Adams also maintains that finite goodness = resemblance to the necessarily existing divine nature, and that moral obligation = being commanded by God; again, these are claims not about the meaning of the words “good” or “obligatory” but rather about the nature of finite goodness and moral obligation. Adams’s identification of the Good with a necessarily existing God yields Lewis’s claim that “[t]he good is uncreated; it could never have been otherwise; it has in it no shadow of contingency” (Lewis, 1995, p. 80). If successful, Adams’s account also explains “how the moral law could depend on God and have God for its foundation, but requires nothing like equating a person with a property, a rule, or an abstract object.” In this way, Adams’s account holds out the promise of avoiding the obscurity of Lewis’s remarks on the relation between God and the moral law while at the same time yielding conclusions that are broadly Lewisian. However, it seems to me that Adams’s theistic account of moral obligation has a serious problem. To see the problem, note first that in developing his divine command theory, Adams relies on the idea that there are some important similarities between human interpersonal relationships and the relationship between human beings and God. He claims that “[o]ur relationship with God is in a broad sense an interpersonal and hence a social relationship” (Adams, 1999, p. 249). He draws on this idea to make the case for the existence of various reasons to obey divine commands. For example, he emphasizes “reasons for compliance that arise from a social bond or relationship with God. . . . If God is our creator, if God loves us, if God gives Con: A Critique of the Moral Argument 143 us all the goods that we enjoy, those are clearly reasons to prize God’s friendship [and hence reasons to obey God’s commands]” (Adams, 1999, p. 252). The contemporary philosopher C. Stephen Evans takes a similar approach. Evans employs the familiar analogy between the parent-child relationship and the God-creature relationship this way:

It is no accident that many religious traditions have conceived of God as like a loving parent. If theism is true, creatures owe their very existence to a being whom they understand to be a just and loving Creator, and it would seem reasonable to conclude that such creatures owe to God respect and gratitude of a particular sort, as well as a duty to obey the commands the Creator might issue to them (Evans, 2004, p. 14).

But what about those who genuinely fail to recognize that they stand in such a relationship to God—nontheists, for example, who fail to recognize that God exists at all? Adams and Evans are both concerned to explain how their theories can account for the moral obligations of non-believers. Adams emphasizes that he wants to understand “the nature of moral obligation quite generally in terms of divine commands” (Adams, 1999, p. 268). An important part of Adams’s strategy for accounting for the moral obligations of nontheists is the claim that some divine commands are issued by way of “moral impulses and sensibilities common to practically all adult human beings since some (not too recent) point in the evolution of our species” (Adams, 1999, p. 270). Adams’s proposal is that God can command even nontheists by designing the human mind in a certain way—by, in effect, giving everyone a conscience. Nontheists will thus be subject to divine commands without realizing it. Similarly, Evans writes:

If a divine command theory is true, then it follows that an atheist fails to understand the nature of moral obligations at the deepest level. However, the atheist can still understand that there are actions that he or she must do or must not do, and can still know what those actions are. . . . The atheist would in that case simply fail to realize something important about those obligations. This is no more mysterious than the inability of someone ignorant of chemistry to know the true nature of water as H2O (Evans, 2004, p. 226; see also Evans, 2013, pp. 20-21).

Evans’s point that one can recognize that one has moral obligations without understanding the nature of moral obligation itself is well-taken. Nevertheless, I think that the Adams/Evans account of moral obligation has difficulty accounting for the moral obligations of nontheists. Consider reasonable non-believers, a class of people that Wes Morriston characterizes this way:

144 ERIK J. WIELENBERG

Some—Theravada Buddhists, for instance—have been brought up in nontheistic religious communities, and quite naturally operate in terms of the assumptions of their own traditions. Others, including many western philosophers, have explicitly considered what is to be said in favor of God’s existence, but have not found it sufficiently persuasive. Still others have never looked into the question in a serious way, but have seen no pressing reason to do so. I shall assume that many persons in each of these categories are reasonable non-believers, at least in the sense that their lack of belief cannot be attributed to the violation of any epistemic duty on their part (Morriston, 2009, p. 2; see also Wielenberg, 2005, pp. 60-62; and Fales, 2010, p. 157).

Adams maintains that God cannot impose moral obligations upon human beings by merely willing that such obligations exist. He argues that God must somehow share His will with us to create moral obligations for us: “Games in which one party incurs guilt for failing to guess the unexpressed wishes of the other party are not nice games. They are no nicer if God is thought of as a party to them. It is implausible to suppose that uncommunicated volitions impose obligations” (Adams, 1999, p. 261). Adams attempts to spell out this communication requirement by proposing the following account of what it takes for God to issue a command:

(1) A divine command will always involve a sign . . . that is intentionally caused by God. (2) In causing the sign God must intend to issue a command, and what is commanded is what God intends to command thereby. (3) The sign must be such that the intended audience could understand it as conveying the intended command (Adams, 1999, p. 265).

But consider the following case:

[I]magine that you have received a note saying, “Let me borrow your car. Leave it unlocked with the key in the ignition, and I will pick it up soon.” If you know that the note is from your spouse, or that it is from a friend to whom you owe a favor, you may perhaps have an obligation to obey this instruction. But if the note is unsigned, the handwriting is unfamiliar, and you have no idea who the author might be, then it’s as clear as day that you have no such obligation. In the same way, it seems that even if our reasonable non-believer gets so far as to interpret one of Adams’ “signs” as conveying the message, “Do not steal,” he will be under no obligation to comply with this instruction unless and until he discovers the divine source of the message (Morriston, 2009, pp. 5-6).

Con: A Critique of the Moral Argument 145

This line of argument suggests that only divine commands that meet a fourth condition in addition to the three Adams proposes can generate moral obligations, namely:

(4) The intended audience must recognize the command as having been issued by God.

Without this condition, Adams’s theory is implausible; but with it, the theory implies that reasonable non-believers have no moral obligations at all, since they do not recognize any command as having been issued by God. In response to this sort of worry, Evans offers a defense of the claim that “God’s commands can generate obligations even for those who do not recognize those commands as coming from God” (Evans, 2013, p. 112). The heart of that defense is the following:

Suppose I am hiking in a remote region on the border between Iraq and Iran. I become lost and I am not sure exactly what country I am in. I suddenly see a sign, which (translated) reads as follows: “You Must Not Leave This Path.” As I walk further, I see loudspeakers, and from them I hear further instructions: “Leaving the path is strictly forbidden.” In such a situation it would be reasonable for me to form a belief that I have an obligation to stay on the path, even if I do not know the source of the commands. For all I know the commands may come from the government of Iraq or the government of Iran, or perhaps from some regional arm of government, or even from a private landowner whose property I am on. In such a situation I might reasonably believe that the commands communicated to me create obligations for me, even if I do not know for sure who gave the commands. . . . In a similar manner it would seem possible for God to communicate commands that would be perceived as authoritative and binding without necessarily making it obvious to all recipients that he is source of the commands. For example, God might communicate that an act is forbidden through conscience, which could be understood as a faculty that directly perceives the wrongness of certain acts (Evans, 2013, pp. 113-114).

I agree that Evans’s example shows that commands to a given subject can sometimes impose moral obligations upon that subject even though the subject does not know the source of the commands. Thus, condition (4) above is too strong. Nevertheless, I do not think that Evans’s response adequately addresses the central worry here. Notice that the lost hiker in Evans’s example recognizes the commands he receives as commands, and his background knowledge enables him to know that the commands are being issued by some legitimate authority or other. Perhaps, then, (4) should be replaced with something like this: 146 ERIK J. WIELENBERG

(4’) The intended audience must recognize the command as having been issued by some legitimate authority or other.

It seems to me, then, that only divine commands that meet at least conditions (1), (2), (3), and (4’) can impose moral obligations upon their intended audience. However, at least some non-believers do not construe the deliverances of their consciences as commands at all; such non-believers will fail to satisfy condition (4’), and hence no moral obligation will be imposed. There is a difference between commanding someone to perform a certain act A and causing someone to believe that he is morally obligated to perform act A and, in general, the second type of process does not impose moral obligations. Suppose that in Evans’s example that instead of using signs and loudspeakers the landowner simply causes anyone walking on the path to form the belief that he is morally obligated to remain on the path. While this may be a somewhat effective method of keeping people on the path, it fails to create an actual moral obligation to remain on the path. Direct manipulation of a subject’s mental states can give the subject true beliefs (perhaps even knowledge) of pre-existing moral obligations, but such manipulation does not bring new moral obligations into being, even when such manipulation is undertaken by someone who has authority over the subject. It should be emphasized that the worry here is not an epistemological one. The worry is not that divine command theory implies that non-theists will have moral obligations but be unaware of them. Instead, the worry is that reasonable non-theists’ lack of belief prevents them from recognizing any divine signs they might receive—including their own “moral impulses and sensibilities”—as commands issued by someone who has authority over them, and that consequently such signs fail to impose moral obligations in the first place. The Adams/Evans style divine command theory is unable to account for the moral obligations of reasonable non-believers not in the sense that the theory implies that non-believers have moral obligations but are unaware of them (an epistemological worry), but rather in the sense that the theory implies that non-believers lack moral obligations altogether (a metaphysical worry). Thus, I am not convinced that the Adams/Evans approach provides an adequate theistic explanation of our moral obligations. If this is right, then defenders of the moral argument cannot draw on the Adams/Evans approach to support the moral argument. While their approach avoids the obscurity of Lewis’s proposal, it founders when it comes to explaining the moral obligations of reasonable non-believers. Those who would defend Lewis’s moral argument must look elsewhere for an adequate theistic account of human moral obligations.

Con: A Critique of the Moral Argument 147

2. Necessary Moral Truths

As Baggett points out, another important task for the defender of Lewis’s moral argument is “to argue that no alternative moral theory explains the Lewisian moral phenomena as well as theism does.” In previous writings (Wielenberg, 2005, 2008, 2009), I have defended a non-theistic account of ethics that I claim explains Lewisian moral phenomena at least as well as theism does. Baggett suggests that my view might run into trouble when it comes to accounting for the reality of necessarily true moral claims: “An affirmation of necessary moral truths . . . strikes me as less a naturalistic view than a quasi-theistic one at the least.” In this section I briefly sketch my view, explain the extent to which it is a naturalistic view, and then describe how my view might account for necessary moral truths. I should emphasize that my goal here is not to establish the truth of my non-theistic view but rather to describe its main contours and suggest one way that necessary moral truths might be accounted for without bringing God into the picture. The type of view I hold has come to be known as robust normative realism. David Enoch characterizes robust normative realism as follows: “[T]here are response-independent, non-natural, irreducibly normative truths . . . objective ones, that when successful in our normative inquiries we discover rather than create or construct” (Enoch, 2007, p. 21). On this view, normative properties—including ethical properties like moral rightness and wrongness as well as goodness and evil—are sui generis, a fundamental type of property not reducible to, or fully constituted, by some other type of property. Consequently, this view is incompatible with strong versions of naturalism according to which all properties are identical with, constituted by, or reducible to natural properties (that is, properties that can in principle be investigated empirically). On my view normative properties are also causally inert and hence do not causally affect the physical universe (see Harman, 1977, pp. 7-8 for a defense of this view). Therefore, my view is compatible with the causal closure of the physical—the principle that every physical event that has a cause at all has a complete physical cause. Commitment to that principle is among the central commitments of naturalism (see for example Chalmers, 1996, p. 128), and hence my view is naturalistic to that extent. While ethical properties do not causally impinge on the natural world, they are, in my view, causal effects of the natural world. For example, consider an act that has the following two features: (i) it is a case of causing pain just for fun and (ii) it is morally wrong (see Mackie, 1977, p. 41). I propose that the fact that the act is a case of causing pain just for fun makes the act morally wrong in that it causes (in a particularly robust fashion) the act to be wrong. A paradigmatic example of the sort of robust causation I have in mind is the causal relation that many theists take to hold between a state of affairs being divinely willed and the obtaining of that state of affairs (note that 148 ERIK J. WIELENBERG when I speak of divine willing I mean divine ordaining or, in Plantinga’s terminology [1974, p. 173], strongly actualizing, as opposed to permitting or allowing). The relation between p’s being divinely willed and p’s obtaining is a causal one of a particularly robust sort. It does not depend on the existence of a law of nature connecting acts of divine will with states of affairs (indeed, theists typically believe that whatever laws of nature hold are themselves the products of divine willing). Theists typically maintain that if God wills that p, this necessarily brings it about that p obtains. Additionally, there is no reason that divine willings cannot be simultaneous with their effects. Necessarily, if God wills at time t that p obtains at time t, then p obtains at time t. In fact, it has been argued that ordinary, non-theistic causation requires that “causes always occur simultaneously with their immediate effects” (Huemer and Kovitz, 2003, p. 556). Accordingly, I propose that being an instance of causing pain just for fun necessarily and simultaneously causes an act to be morally wrong. In thinking about this proposal, it may be helpful to consider the doctrine of divine conservation. According to that doctrine, God not only brings all contingent things into existence; He also sustains or keeps them in existence for each moment that they exist (see Kvanvig, 2008 for a useful discussion of this doctrine). On at least some versions of this doctrine, there is a robust causal relation between divine willing and every contingent thing at each moment of its existence. One way of construing my proposal, then, is as a doctrine of non-moral conservation: whatever moral properties are instantiated are conserved or sustained by various underlying non-moral properties via a robust causal relation that holds between the relevant non- moral and moral properties. God’s willing that p entails that p; it is a necessary truth that if God wills p, then p. I take it that the necessary connection between divine willings and the truth of the willed propositions is explained by the robust nature of the causal connection between the two. I suggest that the existence of necessary moral truths is similarly explained by the robust nature of the causal connection that holds between certain natural and ethical properties. For example, it is a necessary truth that inflicting pain just for fun is morally wrong because being a case of inflicting pain just for fun makes (in the sense of robustly causing) the act morally wrong. Every possible act of inflicting pain just for fun is made wrong by being an instance of inflicting pain just for fun. Accordingly, I offer this as a non-theistic alternative to the theistic explanation of necessary moral truths that Baggett sketches. (I should also note that some of Baggett's remarks suggest that what he is really interested in is an explanation of necessary truth in general. How to account for necessary truth in general in a non-theistic fashion is an interesting issue; however, it appears to have little to do with specifically moral phenomena, and hence seems not to have a direct connection with Lewis’s moral argument). Con: A Critique of the Moral Argument 149

Baggett also suggests that theism may have an advantage when it comes to explaining our knowledge of necessary moral truths. He approvingly quotes Adams’s suggestion that God has “constructed us in such a way that we would at least commonly recognize necessary truths as necessary” (Adams, 1983, p. 751). To see what my approach might have to say on this issue we should look at the specific problem with which Adams is concerned. In the discussion from which Baggett quotes, Adams is interested in explaining human knowledge of necessary truths. He is not concerned with necessary moral truths in particular. Furthermore, he is not concerned to explain how we know that necessary truths are true; rather, he is concerned to explain how we know that such truths as necessarily true. Before offering his own theistic account, he discusses the prospects for a plausible evolutionary explanation of such knowledge. He finds those prospects to be dim:

During the formative periods of human evolution, what survival value was there in recognizing necessary truths as necessary, rather than merely as true? Very little, I should think. Logical or absolute necessity as such is a philosophoumenon which would hardly have helped the primitive hunter or gatherer in finding food or shelter; nor does it seem in any way important to the building of a viable primitive society (Adams, 1983, p. 750; emphasis added).

Consider the following moral claim: it is morally wrong to inflict pain just for fun. That claim strikes me not only as true but as necessarily true. The challenge suggested by Adams’s remarks in the passage just quoted is to provide a plausible evolutionary account of our recognition of that claim as a necessary truth rather than simply as a truth. It seems to me that the key to answering Adams’s challenge (without bringing God into the picture) can be found in the following remarks by Nick Zangwill:

Suppose we judge that Queen Isabella of Spain was evil in 1492. . . . Then we do not think that she had various natural properties in 1492— such as being a torturer, a bigot, and desiring other’s pain—and by an astounding coincidence she or her actions also had the moral property of evil. Rather, we think that she or her actions were evil in virtue of those natural properties; we think that she had her moral properties because of her natural properties. In general, when we make a moral judgment we judge not just that something has a moral property, but that it has a moral property because it has some natural property. This is a fundamental principle of our moral thought (Zangwill, 2008, p. 109).

Zangwill may overstate the matter somewhat here; it is far from clear that we always or even generally consciously judge that something has a moral 150 ERIK J. WIELENBERG property because it has some other property, and many people judge that something has a moral property because it has some supernatural property. Still, I think that what is right about what Zangwill says is that ordinary moral knowledge at least sometimes encompasses not simply knowledge of the rightness or wrongness of actions; in at least some cases, it includes knowledge of which features of those actions make them right or wrong. Sticking with our familiar example, if I know of a given act of inflicting pain just for fun that it is morally wrong, I also know that what makes the act wrong is precisely that it is a case of inflicting pain just for fun and that no other feature of the action could prevent the act from being morally wrong. Putting the matter a bit more colloquially, we might say that I know that there is no excuse for inflicting pain just for fun. Knowing that, I am in a position to know that no matter how the features of the situation in which the act is performed are varied, the moral status of the act will remain the same—that is, that act is morally wrong in every possible world. In this way, we see that in the case of moral knowledge, knowledge that certain moral truths are necessarily true can be a by-product of ordinary moral knowledge and hence needs no special evolutionary explanation. (Notice that none of this implies that I know the deepest explanation for the wrongness of the act. Different normative theories will propose competing ultimate explanations for the wrongness of the act—for example that it fails to maximize utility or that it involves treating a rational being as a mere means to an end. I can know that being a case of inflicting pain just for fun makes an act wrong without knowing which of these competing ultimate explanations is the correct one.) Of course, if this account is correct, it suggests a natural follow-up question from the theistic side, namely: what is the evolutionary explanation of our recognition that inflicting pain just for fun makes an action morally wrong? Here is a very brief sketch of an answer to that question. When we form moral judgments our brains inevitably categorize the entities being judged in terms of their non-moral properties (see for example Prinz, 2008, p. 161; and Morrow, 2009). While such categorizations are sometimes non- conscious, we can often access them in one way or another. This helps to explain why, as Zangwill suggests, rather than forming “bare” judgments of moral rightness or wrongness we often judge not just that a given act is right or wrong but also what makes it right or wrong. Consider, for example, our moral judgment that it is wrong for others to inflict pain on us just for fun and that what makes such acts wrong is precisely that they are cases of others inflicting pain on us just for fun. Recognizing that it is morally wrong for others to inflict pain on us just for fun is evolutionarily advantageous in that it provides a distinctive sort of motivation to resist such treatment at the hands of others. If we are aware that our judgment of moral wrongness is responsive to the others-inflicting-pain-on-us-just-for-fun-ness of such acts then we will judge that that feature of the act is what generates its wrongness. Furthermore, it is advantageous for us to recognize that similar causes tend to have similar Con: A Critique of the Moral Argument 151 effects (because the world tends to work that way). As a result, we are disposed to recognize that inflicting pain just for fun makes an act wrong not only when it is done to us by others but also when anyone does it to anyone else. Consequently, we are led to the recognition that it is wrong for us to inflict pain on others just for fun. As Peter Singer puts it, reason is like “an escalator” that can lead us “to places that are not of any direct advantage to us, in evolutionary terms” (Singer, 2006, p. 146). In this way, evolutionary processes, unguided by the hand of God, can lead us to recognize ethical truths (see Wielenberg, 2010 and 2014 for a much fuller development of this account). In his concluding paragraph, Baggett notes that while “too much has been left unsaid here to draw bigger conclusions about the overall viability of Lewis’s moral argument . . . the argument seems to have great potential.” Like Baggett, I have left much unsaid here. However, I hope to have shown that, whatever the potential of Lewis’s moral argument, serious challenges to it remain to be overcome. I have argued that Adams’s theistic account of moral obligation—one of the most sophisticated theistic accounts of morality advanced to date—has trouble accounting for the moral obligations of reasonable non-believers. And I have sketched what I take to be promising ways of accounting for necessary moral truths and our knowledge of such truths without bringing God into the picture.

Eleven

REPLY TO ERIK WIELENBERG

David Baggett

Let me begin by saying what an unmitigated pleasure and privilege it is to have a friendly dialogue with so gracious a fellow and so clear a thinker as Erik Wielenberg. I have long appreciated his clear exposition and intriguing analysis, and this exchange has only served to enhance my appreciation. The substance of his reply makes it difficult to do it justice, so I will simply try to respond, albeit too briefly, to the two main issues Wielenberg raises: first, whether Robert Adams’s moral theory supports Lewis’s moral argument for God’s existence; and, second, what best explains moral truth, including necessary moral truth. It is important to spend some time discussing Adams’s development of the idea that God is the Good, which is, by Wielenberg’s lights, at least broadly Lewisian in spirit. Let me quickly interject that Lewis wrote quite a bit about issues of moral goodness, and he was not always concerned to connect it explicitly with God. In The Abolition of Man, for instance, he referred to the Tao—the moral law—while referring to moral resources other than theism. Although this may seem initially puzzling, perhaps at least part of his rationale was his conviction that believers and unbelievers alike can apprehend foundational moral truths. Nor did Lewis always sound like a theistic Platonist (or something in that vicinity) by equating God with goodness. He often sounded more like a Thomist and natural law theorist who connected moral goodness with what conduces to the deepest human flourishing. These pictures of goodness are not necessarily at odds, though, since God, on traditional Christian understandings, has often been thought of as the ultimate good, and union with him the ultimate end for human beings. Regarding the identity of God with the Good, though, Adams tries to articulate one way this might be explicable. Wielenberg does a fine job summarizing Adams on this issue, including the way Adams deployed insights from contemporary philosophy of language to defend an identity of God and the Good without requiring synonymy of the words “God” and “good.” Wielenberg notes that Adams’s identification of the Good with a necessarily existing God would, if successful, (1) yield Lewis’s claim that the Good is uncreated, could not have been otherwise, and features no shadow of contingency; (2) make sense of the dependence of morality on God without reducing God to a principle or an object; and (3) avoid the obscurity of Lewis’s remarks on the relation between God and the moral law. 154 DAVID BAGGETT

We can refer to these three issues, respectively, as the necessity criterion, the dependence relation, and the obscurity problem.

1. Is Lewis’s Account Too Obscure?

A word on the last of these first. On the issue of whether “God as moral law” is objectionably obscure, Wielenberg is right to say that Lewis admitted that his proposal of God as the Good is not altogether easy to understand. Lewis likened it to an attempt to grasp the nature of the Trinity, or a Flatlander’s attempt to grasp the nature of a cube. I am less inclined than Wielenberg, though, to read Lewis’s analogy as a concession to what we might call “problematic obscurity.” It seems to me rather an admission, or perhaps insistence, that the ultimate truth here is to some degree beyond our current understandings and ways of envisioning things. In light of standard categories and prevailing views, the truth of the matter—historical, scientific, and otherwise—often challenges prevailing paradigms. The deeper understandings are not a matter of going contrary to reason—though they are often a matter of going contrary to expectations—but truth can and often does stretch our understanding in new and surprising directions, some of which remain in certain respects inchoate and only partially illumined. This is a problematic epistemic deficiency, however, only if we have good reason to expect otherwise. In the current discussion, I see no good reason to believe that the large and difficult questions at issue yield answers that neatly fit into our preexisting categories. In fact, truth often tends to be more iconoclastic and less domesticated than we initially suspect. So, as for me and my house, I do not count Lewis’s concession here as problematic for his view, but rather in accordance with properly humbled epistemic expectations.

2. The Necessity and Ontological Foundations of Moral Truth

Let me turn next to the necessity criterion and the dependence relation. In terms of the former, equating the Good with God himself (or perhaps with God’s character) would make good sense of moral propositions that appear to be necessarily true. In fact, on this score theists and Platonists are kindred spirits—in contrast with many contrary prevailing views among contemporary ethicists. In his well-known piece “Religion and the Queerness of Morality,” George Mavrodes once suggested that a Platonic understanding of objective morality—central somehow both to reality and human knowledge—held much in common with classical theism, but had for its main deficiency that it provided more assertion than principled argument (Mavrodes, 1986). Most theists are not inclined to disagree with Platonists’ nonnegotiable ethical convictions. They simply argue, as Lewis does, that theism provides a more Reply to Erik Wielenberg 155 convincing ontological foundation for those convictions—namely, the thoughts and character of a holy, eternal, and perfectly loving God. In terms of the dependence relation, it is helpful at this juncture to draw a distinction between the good and the right. The good, where morality is concerned, is the axiological category pertaining to moral value; the right is a deontic issue typically involving matters of moral obligation and permissibility. Robert Adams once suggested we can see the connection between the good and the right when we note a scope ambiguity involved in negating good’s opposite, “bad.” If an action is bad not to do, it is (presumably anyway) obligatory; if an action is not bad to do, it is permissible. This is at least one way to see a connection between the axiological and the deontic. Not everything good to do is obligatory, and likewise, perhaps, not everything (in some sense) bad to do is wrong— although things irremediably bad might well be necessarily wrong. Among the moral facts in need of explanation, then, are both the good and the right. In terms of the dependence of morality on God, if God is the Good, that handily explains that dimension of dependence (though this of course requires a lot of fleshing out, some of which Adams tries to do); but what accounts for the deontic dependence? Here Adams embraces a divine command theory of moral obligation. Lewis, in contrast, is generally thought not to be a divine command theorist. In my opening salvo, the reader might recall that I suggested that even if Lewis rejected classical divine command theory, he might have believed that God’s commands invest moral obligations with a new sort of urgency or modal status. This, I suggested, might be seen as a kind of moderate divine command theory of a nonvoluntarist variety. So, supposing something like that can work, it fortifies the parallel between Lewis and Adams despite their differences. In exploring the adequacy of an Adamsian divine command theory (DCT), Wielenberg raises an interesting objection well worth pondering, although we should do so without getting too far off track from the main discussion of Lewis’s moral argument. On Adams’s account—and of course this is but the most rudimentary of sketches—God can command even nontheists by designing the human mind in a certain way by, in effect, giving everyone a conscience. Nontheists will thus be subject to divine commands without realizing it. But Wielenberg thinks this raises an insuperable objection for the account, and to make his case he directs our attention to reasonable unbelievers, who, owing to their nonculpable unbelief in God’s existence, will not, understandably enough, interpret anything as a divine command. Receiving a note from an unknown source that reads, “Let me borrow your car; leave the key in the ignition and the car unlocked,” creates no obligation, unlike the way receiving such a note from your spouse very well would result in an obligation. So Wielenberg thinks that the conditions that a divine command must satisfy should include an additional constraint (AC): “The intended audience must recognize the command as having been issued by 156 DAVID BAGGETT some legitimate authority or other.” Since at least some reasonable unbelievers will not recognize any command as having been issued by a legitimate authority—including a binding conscience—they would fail to have any moral obligations. The problem is not that they have moral obligations of which they are unaware (an epistemic problem), but that they have no moral obligations at all (an ontological problem). I hope I have done justice in short compass to this important objection to divine command theory. By way of reply, I would first note that not every version of theistic ethics embraces DCT. So, even if DCT requires AC and the objection proves intractable, other versions of theistic ethics are available, such as a divine attitude theory that circumvents the objection altogether. Matt Jordan has written an article in which he gives something very close to Wielenberg’s objection to DCT, thinking the objection insuperable, but then lays out an alternate way in which morality can depend on God that does not require any problematic voluntarist component that renders it susceptible to such an objection (Jordan, 2013). But can DCT itself be defended against this objection? I am at least tentatively inclined to think so. It is not that Lewis’s moral argument depends on this defense, but let me briefly offer a few thoughts on the matter. Let us grant that there are reasonable unbelievers in God. Wielenberg is right to suggest that no atheist is going to interpret a moral demand as a divine command, although they might construe certain commands as having come from an authoritative source. The man who knows the handwritten note asking that he leave his car keys came from his wife, for example, might be thought by an atheist to incur a duty to do so. But now consider divine commands. Suppose God were to speak to an atheist, “Go to Ninevah!” It is unlikely that any atheist would construe such a voice in his head as having come from an authoritative source. As a result, the divine command would not issue in a moral obligation. This is right, it seems to me. I would happily concede that a divine command in such a situation would not result in a duty. I am not convinced, though, that this shows that divine command theory precludes atheists from having moral obligations. Hearing inscrutable voices in one’s head and having a conscience are quite different things. Consider for a moment the distinction between revealed theology and natural theology, and between special revelation and general revelation, while bearing in mind what we are discussing. We are discussing Lewis’s moral argument for God’s existence. Such an argument is thought to be part of natural theology predicated on general revelation. In other words, for purposes of Lewis’s moral argument, even if we concede a sense in which Lewis can be called a divine command theorist, the materials required for his argument are moral facts and moral demands available to everyone, not arbitrary commands appearing as random voices in the heads of atheists. The argument attempts to explain the obvious demands morality places on us, not to defend versions of divine command theories that entail that even an atheist who hears “Go to Reply to Erik Wielenberg 157

Ninevah!” now have a moral obligation to do so. Recall that the sense in which Lewis can be argued to be a divine command theorist is only in terms of the limiting nonvoluntaristic case, which is all the more reason to confine our attention to moral truths of a more general and universal, or nearly universal variety. In other words, again, natural theology is predicated on the deliverances of general revelation, not special revelation. Instances of special revelation are interesting, but discussing them here would take us too far afield. It is true that Wielenberg’s example concerning car keys lends itself to a consideration of more particular instances of commands, but I submit that, in this discussion, there is good reason to keep the focus on more general commands—natural theology more than revealed theology, general revelation more than special revelation. It is true that Adams lays out a social requirement theory of obligation generally, and a full defense of the metaphysical contours of his divine command theory requires spelling out the way in which God’s commands best satisfy the social constraints of his analysis. The wife’s note carries authority because of her relationship with her husband; likewise, God’s commands carry authority because of the social roles he plays, his character, and the like. But this discussion is about moral apologetics, not a full-fledged argument that divine command theory is the right theory of moral obligations. Perhaps it is the right account, perhaps not. Even if I could defend divine command theory against this one objection, I would not have shown it to be the right theory. Even if Wielenberg’s objection is sound, it would not show that theistic ethics fails. It is prudent, therefore, to keep the focus more squarely on what is relevant to Lewis’s moral argument. When we do so, we recall that Lewis based his argument on such examples of morality as basic principles of fairness and justice, moral facts to which everyone (or nearly everyone) is privy. The ingredients that went into his argument, in other words, are matters of generally accessible moral truths. Believers and unbelievers alike can apprehend such truths. This was Lewis’s starting point, and this pertains to general revelation. This warrants our keeping the focus less on all the metaphysical aspects of divine command theory and more on what we might call the phenomenology of our moral experiences. We find ourselves able to apprehend clear, nonnegotiable moral facts, and they cry out for an explanation. Lewis argues that theism provides a better explanation than naturalism can. At the foundation of the moral facts in need of explanation may or may not be divine commands. The insistence that someone needs to recognize the authority of the source of morality before standing under the demands of morality gets it exactly backwards. As Lewis argues, all of us have a sense of the authority of morality, and the question he invites us to explore is where that authority came from. So in response to Wielenberg’s claim that divine commands to atheists will not produce moral obligations because the recipients are not likely to 158 DAVID BAGGETT construe the commands as having come from an authoritative source, I hesitate to agree, largely because the deliverances of general revelation on which Lewis’s argument is based seem available to everyone, believer and unbeliever alike. Consider what Lewis referred to as the Tao in his Abolition of Man: moral principles and demands that we find throughout the world and human history. Construed expansively enough, divine commands may function at their foundation, but more importantly, for Lewis, God himself resides at their foundation. This is the point of his moral argument. It is less clear that divine commands play an important role in such truths than that the truths themselves obtain. These truths can be seen to be authoritative, not just because they are practically universal, but because they are self-evident. On this view, atheists and believers alike are able to apprehend these moral truths, sense their authority, and, as a result, be obligated to obey them. A properly functioning conscience would remain binding for atheists. Wielenberg is more skeptical than I am that this story is consistent with divine command theory, but not much rides on divine command theory being the right version of theistic ethics here. The more important question is what best explains those authoritative moral facts. What I am suggesting is that instead of AC perhaps all we need is AC’:

(AC’) Setting aside instances of culpable lack of warrant, the intended audience of binding moral demands will be warranted in thinking that morality is authoritative.

This is the point of recognizing that morality hails from a legitimate authority, but more importantly, recall the apologetic context of this entire conversation: we are considering Lewis’s moral argument for God’s existence. The argument began with authoritative moral demands on us, without any assumptions about where they came from. Identification of the likely source only came later. The authority of morality was self-evident, and it is what enabled the moral argument to get off the ground in the first place. That the truths are self-evident is an epistemic point, of course; it does not settle the issue of moral foundations. Their self-evidence and authority, though, are some of the important facts that require adequate explanation. In Wielenberg’s earlier reply, he distinguishes two sorts of cases: (a) commanding someone to perform a certain act A and (b) causing someone to believe that he is morally obligated to perform act A. In general, Wielenberg claims, the second type of process does not impose moral obligations. Wielenberg could thus reiterate at this juncture that, even if God structured our noetic faculties and the like in such a way that we naturally—at least normatively—find ourselves with firm convictions about the wrongness of child torture for fun, such manipulation of our mental states is not the same as commanding. Causing someone to believe is not the same as making it the case that one is morally obligated. Reply to Erik Wielenberg 159

I might reply by suggesting that it is true that causing one to believe one is obligated is not the same as creating the obligation itself. One can wrongly believe one to be obligated to do all sorts of things. But one’s beliefs in obligations can also, presumably, correspond with reality. On an expansive construal of divine commands, as both Evans and Adams hold, God can and presumably does, either discursively or nondiscursively, form in the minds of properly functioning rational agents firm convictions about certain moral matters. If God structures things in such a way as to bring about such true beliefs, the distinction to which Wielenberg points is without a difference. Assuming that the deliverances of reason and divine commands diverge is predicated on what seems a highly unlikely view that the only way we can know the content of divine commands is by means that circumvent the functioning of reason. For classical theists the very functioning of our rationality is a gift of God and one of the ways God reveals his truth to us, both general and specific. This is exactly one of the strengths of a strongly teleological, theistic picture of moral epistemology, as opposed to a naturalistic, evolutionary story, where there is little reason to believe that the truth of moral propositions is the reason we come to believe them. What seems more likely, as Gilbert Harman, Sharon Street, Richard Joyce, Mark Linville, Angus Ritchie, and others have argued powerfully, is that, on such a naturalistic paradigm, the reason we come to hold the moral beliefs we do is not because they are true, but because, approximately anyway, holding them conduces to reproductive advantage. In sum, in response to Wielenberg’s claim that his objection entails that an Adamsian DCT does not help the cause of Lewisian moral apologetics, I am inclined to think the objection, as it stands, is in need of further argument; and that even if the additional argument can be provided, there would be other potentially fruitful ways to defend theistic ethics and moral apologetics.

3. Can Naturalists Explain Necessary Moral Truths?

Let me now turn to Wielenberg’s second point: his response to my claim that theism explains the reality of necessary moral truths better than naturalism does. I will try to cover this more briefly. In terms of Wielenberg’s own more naturalistic account of morality, he adopts a view according to which, for example, a natural property like inflicting pain for fun causes, simultaneously, a non-natural, causally inert moral fact to obtain, such as moral wrongness. Although I can see what Wielenberg is arguing here—and I readily concede that some very able thinkers may find it persuasive—I myself remain unconvinced. To me there seems something problematically incomplete about the claim that we have furnished a secular account of a necessary moral truth by pointing to the subvening natural property of inflicting pain for fun and the possibility that 160 DAVID BAGGETT such a property causes an inert, non-natural moral fact of moral wrongness to come into existence simultaneously with manifestation of the empirical property. In light of Wielenberg’s example of divine conservation—a view partially constitutive, to my thinking, of classical theism—my resistance on this score is not to disbelieve in the possibility of such causal connections in general. My resistance owes instead to a large measure of skepticism I have that such a causal relation obtains between purely natural properties and moral properties without a larger story to account for it. In the case of divine conservation, such a doctrine can arguably be seen as a function of the divine nature; God is such that all that exists apart from him depends on him for its continued existence. This makes sense both of God’s status as the ground of being and of the radical contingency of the universe. In fact, this is one of the features of classical theism that persuades me that it provides a better overall explanation than naturalism does. Moral properties, however, even if they sometimes supervene on natural properties, are reasonably thought to derive from such empirical properties only if the latter properties are causally up to the task. An eternal, perfectly actualized Being seems potentially quite adequate for the job of upholding a contingently existing universe in existence. But what is it about purely empirical properties in a naturalistic world that makes it reasonable to think them metaphysically adequate to generate authoritative, prescriptively binding, and necessarily true moral facts consisting of non-natural properties? It seems more reasonable to suppose that such mysterious non-natural properties do not exist than that they suddenly come into existence. (Notice that this is an explanatory question, not a justificatory one.) In a 2009 paper, Wielenberg wrote of ethical brute facts that they “come from nowhere, and nothing external to themselves grounds their existence; rather they are fundamental features of the universe that ground other truths” (Wielenberg, 2009, p. 26). His claim now seems to be that they do come from somewhere after all: empirical facts or states of affairs. The “comes from nowhere” hypothesis strains credulity, but so does the new suggestion. What explanatory work do necessary, objective moral facts do, after all? They are not needed to explain our (potentially false) beliefs that the action in question is morally wrong, as Harman and others have argued at length. The evolutionary moral epistemology story Wielenberg tells is less consistent with moral realism than with moral skepticism. Those moral beliefs can be explained quite handily in a naturalistic world without positing the existence of any moral truths at all, much less without investing them with adequate clout to hold authority, even if their truth can be perceived. Much of the ostensible force of Wielenberg’s strategy, to my thinking, derives from remaining moral capital in a culture whose moral convictions historically were grounded in transcendent foundations. Of course I agree with him about the existence of objective moral facts; his challenge is to account for them in a deep way with the resources of his worldview alone without using the Reply to Erik Wielenberg 161 convictions of his audience as an excuse to avoid providing such an argument. If something about a naturalistic world distinctively gave us good reasons to believe in such moral entities, that would be one thing; but merely positing the possibility of their mysterious emergence does not seem enough. At most it shows a potential consistency of naturalism and objective moral truth, although I rather doubt it does even that. As we will see in a moment, there seem to be rather intractable tensions between objective, authoritative, prescriptively binding morality, on the one hand, and naturalism on the other. But even if that stronger case does not go through, an abductive argument does not require it anyway, but merely a superior explanatory case. If theism provides a better explanation than this variant of naturalism, it wins this round, even if Wielenberg has successfully identified a genuine metaphysical possibility. As a theist, Lewis would be more inclined to suggest that, though moral properties very well may supervene on natural properties, we cannot suppose that to be enough said on the matter or else we run a serious risk of leaving out something arguably ineliminable from a full explanation. It is plausible to believe that, in order for moral properties to emerge from natural properties, either we need reason to believe that the natural properties in question are metaphysically up to the task or the nature of morality lends itself to natural explanation. Neither option seems likely on naturalism. What is it about collocations of atoms or the determined actions of conscious agents operating within a system featuring causal closure that gives us any reason to think that such empirical realities are sufficient to generate a binding set of abstract moral truths? Or what is there about non-natural, prescriptively binding moral facts that make it likely that they are altogether explicable by even the most complicated natural properties and relations within a closed causal system? Not much, in my view; and remember: even if either scenario is possible, however remotely, the challenge of winning an abductive argument is to achieve the best explanation, not a merely possible one, or even a mildly plausible one. Theism allows for a more robust metaphysical picture in which the physical order is imbued with meaning and significance that it otherwise would lack. On Lewis’s view it seems likely that this indeed is the sort of world in which we live, making it understandable why naturalists, beings created by God in his image and equipped with cognitive and affective capacities to apprehend some of the transcendence and meaning imbedded in this world, would consider it altogether natural to infer on its basis to the existence of such things as objective moral truths. But what actually explains the existence of that moral order better than naturalism is supernaturalism, Lewis would argue. God, classically construed, created this world and sustains it in existence, has imbued it with signals of transcendence and rich meaning, has populated the world with creatures made in his image in order to enter into loving relationships with him and with others, and has made human 162 DAVID BAGGETT beings with meaningful agency with which to make significant moral choices for which we can be held rightly and deeply responsible. Theism, to Lewis’s thinking at least, can thereby provide the better explanation of what needs explanation when it comes to morality; it resorts to fewer ad hoc adjustments to the paradigm, explains more of what needs explaining, and explains it better. Wielenberg himself admits that stipulating this strong supervenience relation does not address the ultimate reason why moral facts obtain. Given this, it seems to me that he should remain open to supernatural explanations of morality, but obviously much more needs to be said along such lines than can be said in so short a space. The very possibility of a theistic meta-ethic, however, shows that identifying the alleged supervenience relation potentially does little to show that a thoroughgoing secular explanation of morality has been provided. Lewis would argue that once the rest of the picture is fleshed out, theism cannot easily be left out of the overall story.

Twelve

REPLY TO DAVID BAGGETT

Erik J. Wielenberg

Predictably, this exchange between David Baggett and I will end without either of us converting the other to his own position. However, I hope (with some justification, I think) that the exchange serves both to illustrate the sorts of issues that must be settled in order to arrive at a warranted verdict about the strength of Lewis’s moral argument and as a model of civil debate between people who disagree deeply with one another. In any case, I am grateful to Baggett for the thoughtfulness and charity he has shown throughout this exchange. In these final remarks, I will (i) explain what I take to be the relevance of Adams’s theistic account of morality and what Baggett calls “the obscurity problem” to Lewis’s moral argument and (ii) say a bit more about my non-theistic account of morality and address at least some of Baggett’s criticisms of it.

1. The Relevance of Adams’s Theistic Account of Morality and the Obscurity Problem

I take it that one of the central questions about which Baggett and I disagree is the following: does the best explanation of Lewisian moral phenomena (including, among other things, the reality of objective, universal, ethical truths—some of which are necessarily true—and human knowledge of at least some of these ethical truths) entail the existence of the God of classical theism? Baggett says yes; I say no. One fact that I hope our exchange brings out is that a proper evaluation of this question requires comparing the strengths and weaknesses of theistic and non-theistic attempts to explain Lewisian moral phenomena. An important question for any theistic account of morality is: how, exactly, does God explain various moral phenomena? The question is pressing for at least two reasons. First, some ways of answering it seem plainly inadequate (and are widely regarded as such). For example, one possible answer to the question has it that the expressions “is morally obligatory” and 164 ERIK J. WIELENBERG

“is commanded by God” are synonymous, and so God creates moral obligations by issuing commands. However, it is widely recognized by contemporary philosophers, theist and non-theist alike, that such a proposal fails because it has implausible moral and linguistic implications. Second, the existence of developed non-theistic approaches to morality puts pressure on theists to flesh out their own accounts of morality. While explanation must come to an end somewhere (a point to which I return below), it seems that, all else equal, a deeper, more developed explanation is better than a less developed one. It seems to me that theistic moral philosophers have increasingly recognized the need to address the “how does it work?” question—that is, to delve into the nitty-gritty of how God might explain morality. In recent years there has been a proliferation of competing theistic accounts of morality. The longstanding divide between natural law theorists and divine command theorists (see, for example, Murphy, 2011) has been complicated by theories according to which our moral obligations are determined by (a) divine willings (Quinn, 1992; see also Murphy, 1998), (b) divine attitudes (Jordan, 2012), or (c) divine desires (Miller, 2009a, 2009b). Robert Adams’s theistic account of morality is but one of many attempts to work out the details of how God explains Lewisian moral phenomena. As Baggett points out, this means that even if Adams’s theory is inadequate, it does not follow that no plausible theistic account of morality exists. However, if Adams’s account is inadequate, then one of the most sophisticated contemporary theistic accounts of morality must be put aside when trying to determine whether the best explanation of Lewisian moral phenomena entails the existence of God. That would be a significant (but not decisive) blow to defenders of Lewis-style moral arguments. In my initial response to Baggett’s presentation of Lewis’s moral argument, I developed at some length an objection to Adams’s attempt to explain human moral obligations in terms of divine commands. I will not re- hash that objection here. Instead, I will be content to note that it is not clear to me that Baggett’s appeal to the distinction between special and general revelation helps to dispel the worry I raised about Adams’s account. On Adams’s account, all human moral obligations (including those that are revealed to and apply to all human beings) are constituted by commands issued by God to human beings. The heart of the problem I see with Adams’s proposal is that reasonable non-believers are not bound by such commands and hence, on Adams’s view, have no moral obligations—an implausible result. Again, if this is correct, it does not show that Lewis’s moral argument fails; as Baggett points out, there are lots of other theistic accounts of morality on offer. However, to defend Lewis’s moral argument adequately one must make the case that some theistic account of morality is plausible. If Adams’s account indeed fails, the defender of Lewis’s moral argument must look elsewhere for a plausible theistic account of morality. Reply to David Baggett 165

With that in mind, let us return briefly to my charge that Lewis’s account of the relation between God and the moral law is problematically obscure— what Baggett calls “the obscurity problem.” In response to this charge, Baggett suggests that the obscurity of Lewis’s proposal does not show that Lewis’s proposal is false. Baggett claims that it is sometimes the case that the truth is “to some degree beyond our current understandings and ways of envisioning things” and that “truth often tends to be more iconoclastic and less domesticated than we initially suspect.” I agree with Baggett on this point; as evidence for it, one need look no further than quantum mechanics or Einstein’s theories (special and general) of relativity. Attempts by physicists to provide explanations of these theories that are accessible to non-physicists are striking illustrations of Baggett’s contention that the truth is sometimes surprising and at odds with our current concepts. Nevertheless, I maintain that the obscurity of Lewis’s proposal is a strike against it in the following sense: everything else being equal, a theory that lacks such obscurity is better than one that possesses it. Consequently, the obscurity of Lewis’s proposal is salient when evaluating the relative strengths and weaknesses of theistic and non-theistic accounts of morality. The cases of quantum mechanics and Einstein’s relativity theories are instructive here. We accept those theories despite their obscurity because they are far superior in other relevant respects to any competing theory. So, while the obscurity of Lewis’s proposal is not a fatal weakness, it is a weakness nonetheless and one that should be taken into consideration when trying to determine whether the best explanation of morality entails God’s existence. Putting these considerations together, we reach the result that neither Lewis nor Adams has provided a particularly compelling theistic account of morality—specifically, one that adequately answers the “how does it work?” question. If that is correct, then Lewis’s moral argument is not dead; however, its defenders have more work to do to support their contention that the most plausible explanation of Lewisian moral phenomena entails that God exists. Of course, as I noted earlier, an adequate evaluation of Lewis’s argument requires a comparison of the various theistic and non-theistic accounts of morality that have been proposed. The less plausible the available non-theistic accounts are, the lower the bar of plausibility the theistic side must attain to win the prize for most plausible theory. Just as Adams’s account is but one of many on the theistic side, so my account of morality is but one of many on the non-theistic side. But since it is my favorite account on the non-theistic side and Baggett has put forward some new criticisms of it in his latest response, I will devote the remainder of my remarks to further explanation and defense of my non-theistic approach.

166 ERIK J. WIELENBERG

2. Necessary Moral Truths Again

Consider the following moral claim:

(1) The fact that an act is a case of inflicting pain just for fun makes that act morally wrong.

I suggest that (1) is not only true; it is necessarily true. I further suggest that (1) is a good candidate for a brute ethical fact—an ethical fact that is not made true by anything else (not even God). (1) is a general moral claim; it implies that all acts of inflicting pain just for fun are morally wrong. Now consider a particular instance of inflicting pain just for fun; suppose that Jake pokes Henry with a sharp stick just because Jake thinks it is funny when Henry is in pain. Even if (1) is brute and its truth has no external source or ground, the wrongness of Jake’s particular action does have a source: the wrongness of the act is caused by the fact that the act is a case of inflicting pain just for fun. Baggett’s most recent discussion of this proposal seems to suggest three main criticisms of it. The first is that the existence of brute necessary ethical facts is implausible. Baggett writes that “[t]he ‘comes from nowhere’ hypothesis strains credulity”; call this the bruteness problem. The second is that the proposal that certain natural properties (like being an instance of inflicting pain just for fun) robustly cause moral properties (like being morally wrong) is not plausible unless accompanied by “a larger story to account for it”; call this the incompleteness problem. The third is that the proposal that certain natural properties robustly cause moral properties is implausible because natural properties are simply not the sorts of properties that could generate moral properties. Baggett expresses doubt that natural properties are “metaphysically adequate” for the task; call this the metaphysical inadequacy problem. Let us consider each of these objections in turn. In evaluating the bruteness problem, it is crucial to determine whether any plausible theistic account of morality can avoid positing brute necessary ethical facts. If none can, then the fact that my view posits such facts is not a reason to favor a theistic view over mine. I noted in the preceding section that there are many theistic accounts of morality on offer; while I suspect that none of them can avoid positing brute ethical facts altogether, I lack the space to establish that here. Instead, I will argue that Adams’s theistic account of goodness is committed to such facts. This discussion will (i) show that one sophisticated attempt to support Lewis’s view that God is the moral law (an attempt to which Baggett seems sympathetic) also suffers from the bruteness problem and (ii) illustrate the difficulty of avoiding positing brute ethical facts altogether. Recall that the core of Adams’s theory consists of three claims: (i) God = the Good, (ii) finite goodness = resemblance to the necessarily existing divine nature. and (iii) moral obligation = being commanded by God. Adams’s Reply to David Baggett 167 theory commits him to the existence of at least the following brute necessary ethical facts: (a) the Good exists, (b) the Good is loving, (c) the Good is merciful, and (d) the Good is just. It might be thought that Adams’s theory provides a foundation for such ethical facts; does not the theory tell us, for instance, that the fact that the Good exists is grounded in the fact that God exists? The answer is no: since, on Adams’s view, the Good just is God, the existence of God cannot explain or ground the existence of the Good. In the context of Adams’s view, the claim that God serves as the foundation of the Good is no more sensible than the claim that H2O serves as the foundation of water. Indeed, once we see that, on Adams’s view the Good = God, we see that Adams’s theory entails that the Good has no external foundation, since God has no external foundation. It is not merely that Adams’s view fails to specify where the Good comes from; the theory implies that the Good does not come from anywhere. Therefore, if the choice is between Adams’s theistic account of morality and my non-theistic account, considerations of bruteness do not favor one account over the other. In connection with the incompleteness problem, I have three main points to make. The first is that some causal connections are such that light can be shed on them by providing a fuller story that helps us to understand how and why the relevant entities are causally connected, but some causal connections are not like this—and we recognize the existence of both sorts of causal connections. Consider, for example, the causal connection between (a) placing a pot of cold water over a gas flame and (b) the boiling of the water in the pot. There is a story about thermal and kinetic energy and molecular motion that we can tell to shed light on why there is a causal connection between (a) and (b). By contrast, there are certain causal connections between the most fundamental physical entities (quarks, strings, or whatever) for which no deeper story can be told. At some point explanation comes to an end and we say: that is just how the universe works. My first point, then, is that the fact that a given theory posits causal connections without some deeper story to explain such connections is not a fatal defect in the theory. Indeed, it appears that classical theism posits just these sorts of connections. According to classical theism, God has the power to bring entities into being ex nihilo through the sheer force of His will. So, consider the causal connection between (a) God willing that Jupiter exist and (b) Jupiter coming into existence. It seems to me that there is no story we can tell that sheds light on the nature of the causal connection between (a) and (b) that is relevantly like the story about energy and molecules that we can tell in the boiling water case. It appears that when it comes to causal connections between God willing that p and p obtaining, there is not much we can say beyond: that is just how it works. So my second point is that the incompleteness problem appears to plague classical theism and my view equally. However, I do concede that if one theory of morality can provide an account of why certain natural properties produce certain moral properties and another cannot, and all else is 168 ERIK J. WIELENBERG equal, then the first theory is preferable to the second. To that extent the incompleteness problem constitutes a weakness in my theory. But (and here is my third point) this weakness can be exploited by defenders of Lewis’s moral argument only if they can produce a plausible theistic account of morality that (a) provides such a deeper explanation and (b) is, taking all other relevant factors into consideration, at least as plausible as my theory. While I have my doubts that such a theistic account can be produced, I must be content here simply to note that whether such a theory can be produced should be included on the list of issues to be settled in order to arrive at a warranted verdict about the success of Lewis’s moral argument. Finally, let us consider the metaphysical inadequacy problem. I suspect that Baggett and I simply have differing intuitions here: whereas it seems to him that natural properties cannot robustly cause moral properties, it seems to me that they can. I note that, as far as I can see, Baggett offers no argument in support of his view on this issue; at the same time, I must confess that I similarly lack an argument in support of my view on this issue. How, then, to proceed? One option is simply to invite readers to consult their own intuitions here. However, I think there is some reason for us to view our intuitions about what can cause what with some suspicion. The history of science teaches us that the causal powers of the natural world are often surprising and (at least initially) counterintuitive. For example, Newton’s law of universal gravitation was originally criticized on the grounds that so-called “action at a distance” is impossible; it seemed to many of Newton’s critics that one physical object could causally affect another only if the two objects came into contact with one another. Leibniz, for example, claimed that Newton’s views imply that gravity is some sort of mysterious, incomprehensible “occult quality” of physical objects (see Janiak, 2014 for a helpful discussion of this episode in the history of science). So, when it comes to what can cause what, the truth, as Baggett says, “often tends to be more iconoclastic and less domesticated than we initially expect”; accordingly, I suggest that those who share Baggett’s intuition that natural properties cannot robustly cause moral properties should view those intuitions with “properly humbled epistemic expectations.” I conclude that the metaphysical inadequacy problem is not a serious problem after all.

3. Conclusion

I said at the outset that Baggett and I disagree over the answer to the following question: does the best explanation of Lewisian moral phenomena (including, among other things, the reality of objective, universal ethical truths—some of which are necessarily true—and human knowledge of at least some of these ethical truths) entail the existence of the God of classical theism? Those who would defend Lewis’s moral argument must make the case that the correct Reply to David Baggett 169 answer to this question is yes. Throughout this exchange I have argued for a negative answer to this question by (a) identifying what seem to me to be difficulties in some of the theistic accounts of morality that have been proposed and (b) providing an admittedly incomplete sketch and defense of a non-theistic account of morality (see Wielenberg, 2009 and 2014 for more). While I can hardly claim to have settled the matter, I hope to have pointed toward some significant challenges faced by defenders of Lewis’s moral argument. To determine whether Lewis’s argument succeeds, one must evaluate the various theistic and non-theistic theories of morality that have been proposed and determine which of them (or at least, which type of them— theistic or non-theistic) is more plausible. If that is right, then arriving at a warranted verdict about the adequacy of Lewis’s moral argument is a large and difficult task. Unfortunately, the truth is sometimes not only surprising but also hard to find. That, as they say, is life.

Thirteen

PRO: A DEFENSE OF C. S. LEWIS’S “TRILEMMA”

Donald S. Williams

No apologetical argument that C. S. Lewis ever made is better known—or more controversial—than his famous “Trilemma” (not his word), or “Lord/Liar/Lunatic” (not his phrase) argument for the deity of Christ. N. T. Wright observes accurately that “[t]his argument has worn well in some circles and extremely badly in others” (Wright, 2007, p. 32). Some of the sharpest critiques, in fact, have come from within the believing community. It is curious that an argument that has become a staple of popular Christian apologetics should be rejected as fallacious by many who presumably accept its conclusion. With not only the validity of a much used argument but also the competence of the greatest apologist of the twentieth century at stake, it is time to take a fresh look at Lewis’s argument and its critics. Can we still use the Trilemma? If so, how should we approach it? At the end of the day, how does Lewis come off as an apologist and an example to other apologists? First, let us remind ourselves of the argument itself as it is presented in Mere Christianity. (See Brazier, 2002, pp. 91-102 for a survey of other works in which Lewis gives a version of the argument.) Lewis is addressing a person who says, “I’m ready to accept Jesus as a great moral teacher, but I do not accept his claim to be God.” We note first of all that the Trilemma is presented not so much as an argument for the deity of Christ per se, but as a refutation, a heading off at the pass, of one popular way of evading the claims of Christ. This, Lewis argues, is the one thing we cannot say:

A man who was merely a man and said the sort of things Jesus said would not be a great moral teacher. He would either be a lunatic—on the level with the man who says he is a poached egg—or else he would be the Devil of Hell. You must make your choice. Either this man was, and is, the Son of God: or else a madman or something worse. You can shut Him up for a fool, you can spit at Him and kill Him as a demon; or you can fall at His feet and call Him Lord and God. But let us not come with any patronizing nonsense about His being a great human teacher. He has not left that open to us. He did not intend to (Lewis, 1943, p. 56).

172 DONALD S. WILLIAMS

Many critics treat Lewis’s Trilemma as original. But it is actually a refinement of a much older argument, the aut Deus aut malus homo (“either God or a bad man”) which goes back at least to the Patristic period. (See Brazier, 2012, pp. 103-126 for a survey of its use before and after Lewis.) Lewis makes the dilemma a trilemma by subdividing the malus homo option into two types of badness—mendacity and insanity—which are potentially relevant to the case of the claims of Christ to be God. Later thinkers have expanded it again to a Quadrilemma: Lord, Liar, Lunatic, Legend, or alternatively, Lord, Liar, Lunatic, Innocently Mistaken. In this chapter I will use the familiar term Trilemma to refer to the aut Deus aut malus homo (or “Mad, Bad, or God”) argument in whatever iteration we find it, because it was Lewis’s tripartite form that gave it classic expression for most of us. Lewis’s version of the argument involves the following steps:

(1) Jesus claimed to be God. (This is assumed in Mere Christianity.) (2) There are three logical possibilities in the case of such a claim: (a) He was telling the truth. (b) He was lying. (c) He was mistaken (and hence insane, given the nature of the claim). (3) A liar or a megalomaniac (the relevant form of insanity) could not be a great moral teacher. (4) Therefore, we must either accept Jesus’ claim or reject him as immoral or insane. The merely mortal great moral teacher option is logically eliminated.

Note that one could go on to argue that (5) Jesus was not a liar, (6) Jesus was not insane, therefore (7) Jesus was God. One could; many have; I might; in the next chapter Lewis does—but in the original passage from Mere Christianity Lewis leaves it at (4). He is explicit about his purpose: “I am trying here to prevent anyone saying the really foolish thing that people often say” (Lewis, 1943, p. 55). Lewis does not claim to have proved the deity of Christ beyond a shadow of doubt, but only to have clarified our choices. Jesus was (a) telling the truth, and is the Son of God; he was (b) lying; or he was (c) mistaken—and one cannot be mistaken about the particular claim being made (deity) and be fully sane. The only choice Lewis claims to have eliminated absolutely is that Jesus was simply a great, but merely human, moral teacher—for a person who is a liar or a megalomaniac hardly qualifies as a great moral teacher. Now, the argument is surely presented as support for the deity of Christ in that Lewis thinks that the other two choices will be hard choices for most people to make, as well as choices that give inferior explanations for the full data of the phenomenon of Christ. But people could still make them. “You can shut him up for a fool. . . .” The easy choice—that Jesus was a great moral Pro: A Defense of C. S. Lewis’s “Trilemma” 173 teacher but not God—is the only one Lewis actually purports to have eliminated completely. How well does he succeed? The basic problem Lewis’s critics have had with this argument, even in this limited understanding of it, is their contention that it commits the fallacy of False Dilemma, the premature closure of options. Marvin D. Hinten uses it as an example of one of Lewis’s alleged weaknesses: he “overlimits choices” (Hinten, 2008, p. 8). If it can be shown that there are other legitimate possibilities for how to understand the claims of Christ, it is urged, the argument fails. The other possibilities suggested fall into basically two categories: first, the possibility that Jesus did not actually make the claims attributed to him, or that if he did, he did not mean them as the bald claims to deity for which conservative Christians have taken them; and, second, the possibility that someone could indeed be sincerely mistaken about his identity without being truly insane in a way that would necessarily compromise his views of ethics or his status and authority as a moral teacher. We will examine each of these possibilities in turn, and then look at an additional objection: that, even if the propositions of the Trilemma are probably true individually, their combined probabilities fall below the threshold of persuasiveness (the “diminishing probabilities argument,” or DPA).

1. First Objection: Biblical Criticism

First, it is argued, modern biblical criticism does not allow us to make the naïve assumption either that Jesus said everything that the New Testament attributes to him or that what he did say has the meaning conservative Christians have always attached to it. Few believers are ready to sign up for the Jesus Seminar and question wholesale whether the words of Jesus as reported in the canonical Gospels are authentic. But believers do need to concern themselves with the fact that many secular people today will not begin with a presumption of their authenticity. Thus, Wright thinks that Lewis’s argument “backfires dangerously when historical critics question his reading of the Gospels” (Wright, 2007, p. 33). It is equally common to question whether Jesus’ statements really add up to a clear and unequivocal claim to deity. All that is needed to deprive Lewis’s argument of its logical force is the probability that Jesus’ words should be taken in some other sense. For some, Lewis’s failure to consider such a possibility robs him of all credibility. “Lewis’s view that Jesus’ claims were so clear as to admit of one and only one interpretation reveals that he is a textually careless and theologically unreliable guide” (Beversluis, 1985, p. 54). What are these other possible readings? Here things get a bit murky. It is apparently easier to suggest that a greater knowledge of, say, the first-century 174 DONALD S. WILLIAMS

Jewish background would make such readings possible than it is to come up with specific examples. Thus, Beversluis: “Lewis’s discussion suggests that all individuals of all times and places who say the kinds of things Jesus said must be dismissed as lunatics. But this overlooks the theological and historical background that alone makes the idea of a messianic claim intelligible in the first place” (Beversluis, 1985, p. 56). How exactly knowledge of that background would alter the nature of Jesus’ claims is not made clear. The best Beversluis can manage is, “When they did dispose of him, it was not on the ground that he was a lunatic but on the ground that he was an imposter” (Beversluis, 1985, p. 56). N. T. Wright takes a different tack, appealing to the “strong incarnational principle,” which was the Jewish Temple, the sign of God’s presence among his people (Wright, 2007, p. 32). Lewis does not so much get Jesus’ deity wrong as “drastically short circuits” the original Jewish way of getting there: “When Jesus says, ‘Your sins are forgiven,’ he is not claiming straightforwardly to be God, but to give the people, out on the street, what they would normally get by going to the Temple” (Wright, 2007, p. 33; emphasis in the original). By not taking us deeply enough into first-century Jewish culture (at least as understood by Wright), Lewis fails to give us “sufficient grounding in who Jesus really was” (Wright, 2007, p. 33). Readers willing to brave the technicalities of biblical criticism can easily get the impression that there is a solid scholarly consensus that we cannot assume that Jesus really said everything that the Gospels present him as saying. Representative of this view is Frances Young’s contribution to John Hick’s symposium The Myth of God Incarnate, “A Cloud of Witnesses.” Young takes it for granted that the New Testament writings were produced by people trying to come to grips with the meaning of Christ and doing it in terms of their own developing situations in their churches. Few would question that picture of things; I do not. But Young draws from it the conclusion that the picture we get of Jesus is “the result of believers searching for categories in which to express their response to Jesus, rather than Jesus claiming to be those particular figures” (Young, 1977, p. 15). Thus, she asserts, “[t]he titles were attributed to Jesus by the early Christians and were not claimed by Jesus himself” (Young, 1977, p. 17). Only in John’s Gospel are claims of divinity actually put into Jesus’ own mouth, and John according to Young is not a historical account at all but a later meditation on the meaning of Jesus’ life. If this conclusion is true—or is even as solidly supported by a real scholarly consensus as is implied—then the Trilemma would have great difficulty getting off the ground, with its initial premise (that Jesus claimed deity) being questionable at best.

Pro: A Defense of C. S. Lewis’s “Trilemma” 175

2. Biblical Criticism: A Response

Lewis’s argument as presented in Mere Christianity simply presupposes that Jesus said and meant the things he is traditionally taken to have said and meant. The argument is conditional in form, “Given that Jesus said and meant these things, this is what follows.” To note that the antecedent of this conditional is controversial in some circles is not a refutation; a refutation would require establishing that the supposition is false, or at least probably not true. And this, as I will argue, has simply not been done. Why does Lewis, though, make an initial assumption that does not appear to be one that we can safely afford to make? It was not because he was unaware of biblical criticism. It seems to me that most critics of Lewis have simply ignored the original audience for the Broadcast Talks that eventually became Mere Christianity: not college educated people but simple British laypersons during World War II. To bring up the technical issues of biblical criticism with that audience would have been a foolish introduction of questions they were not asking, unnecessary complications they did not need to deal with. With a more sophisticated audience, one would of course have to be prepared to make a case for the authenticity of the Gospel accounts and deal with alternative interpretations, because the truth of the initial assumption is indeed essential to the argument. That Lewis knew of this challenge and was prepared to meet it when appropriate is proved by essays such as “Modern Theology and Biblical Criticism.” Peter Kreeft and Ronald Tacelli also recognize the necessity of having a response to the critical argument; they expand the Trilemma into a Quadrilemma: Lunatic, Liar, Lord, or Legend (Kreeft and Tacelli, 1994, pp. 161-174). Their divinity-claiming Jesus is not a legend because the New Testament documents are too early to have allowed for a long period of gradual magnification of Jesus’ reputation by later followers. Beversluis in 1985 rejected this context-based defense: “When Lewis . . . justifies the popular approach on the ground that ‘if you are allowed to talk for only ten minutes, pretty well everything else has to be sacrificed to brevity,’ he presents not a justification but an excuse. . . . Why not write a longer book in which ‘everything else’ can be fully and fairly discussed?” (Beversluis, 1985, p. 57). But here Beversluis falls prey to that regrettable tendency of reviewers to criticize the book they would have preferred the author to have written rather than the book he actually wrote. Would Beversluis have an audience of simple laypersons remain unaddressed? Does he really think it makes sense to confuse them with technicalities that do not concern them? As for the “longer book,” one could say that it exists in Miracles or can be reconstructed from various essays that do address different, more sophisticated audiences. In C. S. Lewis’s Case for the Christian Faith, Richard Purtill has a fine discussion of that larger argument gleaned from a more 176 DONALD S. WILLIAMS generous sampling of the Lewis corpus (Purtill, 1981, pp. 45-71). Most of Lewis’s critics simply ignore the original context. In the much-expanded second edition of C. S. Lewis and the Search for Rational Religion, Beversluis responds to the arguments of Lewis and others that support a traditional reading of the Gospels as giving an accurate and reliable report of Jesus’ claims. He says that all such arguments “uncritically assume that the synoptic Gospels are historically reliable sources” (Beversluis, 2007, p. 116). Instead of scholarship, apologists like Kreeft and Tacelli offer “a flurry of unscholarly pseudo-questions” (Beversluis, 2007, p. 118), such as why the apostles would be willing to die for what they knew was a lie. “Real” New Testament scholars do not ask such questions because they “know” that none of the original apostles had anything to do with the Gospels. “All mainstream New Testament scholars agree that the synoptic Gospels are fragmentary, episodic, internally inconsistent, and written by people who were not eyewitnesses” (Beversluis, 2007, p. 123). For someone who claims to find fallacious motes in the eyes of others, Beversluis has a curious blindness to the beams in his own eyes. His whole argument here depends on the fallacies of ad verecundiam (inappropriate appeal to authority) and dicto simpliciter (sweeping generalization). Even if all serious biblical scholars did agree with Beversluis, that fact in itself would not make them right. But they can only be said to agree by the sleight of hand of simply (and arbitrarily) defining a “mainstream” scholar as a skeptical one. Beversluis’s unqualified generalization—all?—has never in fact been true, and is less true now than it has been at any time in the modern age. Richard Bauckham’s magisterial Jesus and the Eyewitnesses is just one recent counterexample. A basic source like Stephen Neill’s classic The Interpretation of the New Testament could have provided Beversluis with many more. Beversluis in his revised edition also responds specifically to Lewis’s own arguments in “Modern Theology and Biblical Criticism.” He simply dismisses Lewis’s point that people who claim to find myths and legends in the Gospels need to know something about myths and legends, as well as Lewis’s observation that source criticism when applied to modern authors where it can be checked is almost always wrong. Beversluis patronizes these concerns as “the Argument from Personal Incredulity” (Beversluis, 2007, p. 123). Nevertheless, Lewis’s incredulity is not just a rhetorical ploy but has very good and specific grounds in his claim that the whole enterprise of skeptical criticism is methodologically flawed—an issue that Beversluis just fails to address. But that claim is central to the case against this alleged “consensus.” We will have more to say about this below. So far, we have to conclude that the authenticity of the sources simply has not been overturned by this argument. The alternative interpretations of Jesus’ claims are not impressive either. Recall Beversluis’ claim that “[w]hen they did dispose of him, it was not on Pro: A Defense of C. S. Lewis’s “Trilemma” 177 the ground that he was a lunatic but on the ground that he was an imposter” (Beversluis, 1985, p. 56). How exactly is that a problem? “Liar” is one of the horns of the Trilemma. Is not an imposter just one form of liar? Is not Liar at least as incompatible with Great Moral Teacher as Lunatic? And N. T. Wright seems to expect of his readers a sophistication in modern interpretations of Jewish culture that even the Pharisees of Jesus’ day did not manifest. After Jesus’ declaration that the sins of the paralytic were forgiven prior to his healing, they were not saying, “Who is this who speaks blasphemies? Where can sins be forgiven but in the Temple alone?” They were saying, “Who is this who speaks blasphemies? Who can forgive sins but God alone?” (Luke 5:21; emphasis added). In other words, Lewis’s argument deals with the reactions Jesus’ contemporaries actually made to him, not the one Wright thinks they should have made! Wright thus tempts one to apply to him Lewis’s verdict from “Modern Theology and Biblical Criticism”: These critics are so adept in reading between the lines that they have forgotten how to read the lines themselves. Beversluis fares no better when he claims that all that is needed is to suppose that Jesus had been “authorized to forgive sins by God” (Beversluis, 2007, p. 124, emphasis added). This again simply ignores the actual reaction of Jesus’ contemporaries. They took Jesus’ words as a claim to deity, and he did nothing to allay their concerns. To understand their reaction, as well as the significance of Jesus’ allowing it to take place, modern readers might be helped by imagining the reaction of a devout Muslim to a mere human being who claimed to be Allah. It is ironic that Lewis is accused of ignoring the cultural context of the Gospels’ claims for Jesus by people who have obviously failed to make the effort to imagine the fierce monotheism of first- century Judaism—a basic and essential prerequisite to any audience analysis of the words of Jesus! Far from Lewis’s views of the Gospels revealing him as “a textually careless and theologically unreliable guide” to them, it would seem that the accusation would better fit Lewis’s critics. G. K. Chesterton asked a pertinent question in his version of the argument: “Mahomedans did not misunderstand Mahomet and suppose he was Allah. Jews did not misunderstand Moses and identify him with Jehovah. Why was this claim alone exaggerated unless this alone was made?” (Chesterton, 1925, p. 246). Frances Young commits the same kind of fallacious bandwagon appeal to scholarly consensus as Beversluis, and adds to it a brazen non sequitur. Surely the New Testament writers were indeed struggling to understand Jesus in terms of their own problems. This is simply to say that they were human beings. It does not follow that they put their own ideas into Jesus’ mouth, or into the mouths of his close associates (like Peter in his famous confession), or that they manufactured incidents like Jesus forgiving sins, along with the reactions of those present. These are conclusions that would have to be reached independently, needing more grounds than the assumption that things 178 DONALD S. WILLIAMS just must have happened that way because that is how “real scholars” understand the evolution of the New Testament. That Young is imposing a concept of evolution on the New Testament documents rather than reading it out of them is suggested by the strange statement that their “dates of origin span approximately three quarters of a century” (Young, 1977, p. 14). First, that is very unlikely. The earliest documents are the first epistles of Paul, which are probably from the fifties. But practically the whole New Testament, including all four canonical Gospels, was already being quoted as Scripture by the Apostolic Fathers by the end of the first century, meaning it had to be in circulation some time before that (Bruce, 1960, pp. 18-19). The actual period of composition then may be as little as half what Young suggests, and her suggestion is hardly indisputable—but it is needed to give time for the evolution of the early Christians’ understanding of Jesus that is assumed to have happened. And that is precisely the point. Included in the collection accepted by the end of the first century are all four canonical Gospels and the undisputed Pauline epistles—all the major documents on which the traditional account of the claims of and for Christ are based. Even if later dates for a few of the disputed epistles be granted, the earlier dates we must accept for the rest make it harder to posit the kind of evolution critics like Young assume. Young is very honest about the source of the presuppositions that drive such an understanding: “The Christians of the early church lived in a world in which supernatural causation was accepted without question.” But such a world view is “unthinkable now.” “There is no room for God as a causal factor” in the modern mind, and Christian scholars according to Young must simply bow to that situation (Young, 1977, p. 31). But if we want honestly to examine the question whether Jesus could have claimed to be—and was—the Son of God, that is precisely the point on which we have to keep an open mind! Young’s closed mind, and that of her cohorts in the mainstream critical “consensus,” renders what looks like textual scholarship an exercise in philosophy determining in advance what texts are to be allowed to say. In this she is typical of the whole enterprise of negative biblical criticism. That is precisely why that critical consensus is unimpressive to conservative believers. It is philosophically prejudiced and methodologically flawed, not to mention actually balanced by a significant body of criticism that, without the predisposing naturalistic bias, reaches very different conclusions. Recall Lewis’s observation that the kind of reconstructive techniques practiced by skeptical scholars have an accuracy record near zero when applied to contemporary documents where the results can be checked (Lewis, 1967, pp. 159-161). I would argue, much as Lewis did, that Jesus’ contemporaries, who were or had access to eyewitnesses, are in a better position to know what he said than modern experts trying to reconstruct the documents according to Pro: A Defense of C. S. Lewis’s “Trilemma” 179 their own preconceived modernist philosophies. For anyone who looks at the critical issues in that light, the initial premise of the Trilemma remains strong. In summary, Lewis’s Trilemma did not, in fact, “backfire” with the audience for whom it was intended, even if it does not work with negative historical critics, a “failure” that Lewis himself would have expected. Even a more sophisticated audience that objectively examined the data would have to admit that the complications raised by modern biblical criticism do not overturn the initial premise of the Trilemma. According to the documents (as opposed to tendentious theoretical interpretations and reconstructions of them), Jesus in fact claimed to be God: he made the statements and performed the actions, and he meant what he said. This is confirmed by the reactions his contemporaries actually had to those words and deeds. Anyone using the Trilemma argument today should be prepared to make the case that Jesus actually made the claims that the Gospels attribute to him. The wise apologist will not simply repeat Lewis’s paragraph from Mere Christianity, but instead adapt it to his own audience. This will involve notations such as “Here be prepared to insert ‘Modern Theology and Biblical Criticism,’ along with further updated arguments.” Unlike his critics, we should look to Lewis’s other books and essays as evidence for how he himself would have used the argument from Mere Christianity in different contexts, and then follow suit ourselves.

3. Second Objection: Mistaken Identities?

The second major attempt to show that Lewis failed to cover his bases involves, amazingly, the denial that only an insane person could sincerely but mistakenly believe himself to be God, or that such a mistake would automatically disqualify him as a great moral teacher. McGrath thinks that “[t]he option that Jesus was someone who was not mad or bad, but was nevertheless wrong about his identity, needs to be considered as a serious alternative” (McGrath, 2013, p. 227). Along that line Beversluis originally asserted that “[w]e could simply suppose that although [Jesus] sincerely believed he was God, he was mistaken” (Beversluis, 1985, p. 55): not lying or insane, just mistaken. He elaborates: “If we deny that Jesus was God, we are not logically compelled to say that he was a lunatic; all we have to say is that his claim to be God was false. The term lunatic simply clouds the issue with emotional rhetoric” (Beversluis, 1985, p. 55). In his second edition, he adds documentation from psychological studies of insanity to the effect that “delusional people are deluded about something . . . but they are rarely, if ever, deluded about everything” (Beversluis, 2007, p. 126). Just because a person is deluded about who he is does not necessarily mean that he is deluded about the content or value of his moral teachings. 180 DONALD S. WILLIAMS

Well, this assertion is generally correct; but surely its application to the specific case of Jesus would take some supporting. No doubt people may be sincerely mistaken about a lot of things, even having to do with their own identity, without being necessarily insane; and they can be insane without being wrong about morals. But make no mistake: We are being asked here to believe that a person could be mistaken about the claim that “Before Abraham was, I Am” (John 8:48), a person who was in a position to be familiar with the standard translation of the Tetragrammaton, the Old Testament name of God, and still be considered a sound thinker about morals (or anything else). Is this really credible? Marvin D. Hinten shows how such support might look. When he teaches Mere Christianity, he asks his class

if they believe angels really did appear to Joan of Arc to say she was God’s chosen instrument to save France. Half the class shake their heads no; the other (quicker-thinking) half simply sit and think it over, because they already see where it is going. None of them see Joan as insane or demonic, so if they apply Lewis’s line of reasoning they will have to admit God really did send angels to Joan, which they have no intention of admitting. I then bring Mohammed into the mix, a man who genuinely seems to have felt Gabriel appeared to him with teaching from God. We discuss ways in which a goodhearted person could be genuinely mistaken about their [sic] role in life: an idée fixe, a hallucination, etc. (Hinten, 2008, p. 8).

Daniel Howard-Snyder has the most sustained and rigorous argument for the idea that Jesus could have been merely mistaken about being God. He admits that believing one is divine when one is not is believing something “importantly false,” but then claims that “[m]erely being wrong about something important, even something as important as whether one is divine, neither implies nor makes it likely that one is a lunatic, insane, deranged, or otherwise fit to be institutionalized” (Howard-Snyder, 2004, p. 463). To support this audacious claim he tries to imagine scenarios in which Jesus could have had credible but false grounds for believing he was God, reasons that could have beeen accepted by someone who was not insane. Perhaps Satan could have given him the ability to perform miracles and duplicated in his mind the subjective experience of being divinity incarnate. Perhaps Jesus, convinced that he was the Messiah, found exegetical grounds in the Old Testament for believing that the Messiah was in some sense divine. (This would be plausible because in fact the early Christians did find such textual arguments for Christ’s divinity after the fact.) For Howard-Snyder, these are “good but fallible grounds” that a person might have for believing in his own divinity (Howard-Snyder, 2004, p. 474). Jesus might have made such deductions in error, or applied them to himself in error, without being insane. Howard-Snyder does not claim that either scenario actually obtains, or is Pro: A Defense of C. S. Lewis’s “Trilemma” 181 likely, but simply that their possibility makes it impossible to dismiss the “sincerely mistaken but still sane” option; therefore the Mad, Bad, or God argument fails. So, Howard-Snyder claims, you can be mistaken about your identity without being insane. Likewise, you can be mistaken about your identity without undermining your views of ethics. According to Beversluis, Lewis “apparently thought that if certain factual claims Jesus made about himself were false, a disastrous conclusion would follow about the truth, sanity, and reliability of his moral teachings. But why say that?” (Beversluis, 1985, p. 55). Beversluis goes on to ask, “Did Lewis think that if Jesus were not God, there would no longer be any reason for believing that love is preferable to hate, humility to arrogance, charity to vindictiveness, meekness to oppressiveness, fidelity to adultery, or truthfulness to deception?” (Beversluis, 1985, p. 55). For Howard-Snyder, we are not in a position to say that the diabolic deception or exegetical misapplication scenarios “are significantly less likely or plausible than the God option” (Howard-Snyder, 2004, p. 478). So the Trilemma fails at every point by this view. You can in theory be mistaken about your identity without being insane and without having false views of ethics; therefore, Lewis has failed to eliminate the “Great Moral Teacher but not God” view of Jesus and hung his apologetic on a fallacious hook. “Contrary to what Lewis claims, we can deny that Jesus was God and say that he was a great moral teacher” (Beversluis, 2007, p. 135).

4. Mistaken Identities: A Response

Let us begin by remembering the conclusion of Lewis’s Trilemma: that Jesus could not have been a great moral teacher but not God. The response of the critics is, well, why could he not have been just sincerely mistaken about God without being insane, or have been mentally unbalanced in some sense but still be a great moral teacher? To rebut this response, we need to be clear about what it takes to be a great moral teacher. I would suggest the following criteria: First, you have moral teachings that resonate with humankind’s most basic instincts about right and wrong, and state them in ways that are both profound and challenging. Second, you are a person of great integrity and act consistently with your own moral teachings. Third, you must be sufficiently in touch with reality that your teachings have general credibility. Clearly, if Jesus had been lying about his claims, he would be disqualified by the second test; but few accuse him of that. More importantly for this discussion, a person who failed the third test would also have problems with the trustworthiness needed to fully inhabit the role of a great moral teacher, even if he consistently lived up to his own ethical ideals. This is where the rubber meets the road in evaluating the claim that Jesus could have been simply mistaken about his deity. 182 DONALD S. WILLIAMS

Most of Lewis’s critics succeed in undermining his argument only by use of a clever sleight of hand known as the straw-man fallacy. The argument most of them attack is simply not the one that Lewis made. Most of the criticisms deal with the general concept of mistaken identity, whereas Lewis is dealing with a very specific case of it, the false claim to be God. As David Horner rightly puts it, Beversluis’s representation of the case (if “certain factual claims Jesus made about himself were false”) is hardly adequate. “The factual claims in question are of cosmic, as well as supremely personal and existential, consequence” (Horner, 2008, p. 77). Treating such vastly different cases of mistaken identity as equivalent is illogical at best and dishonest at worst. But Lewis’s critics have to do it to make their criticisms sound plausible. (Howard-Snyder does deal more directly with the specific claim to divinity, but does not take it with sufficient seriousness, as I will try to show.) This weakness becomes very clear when we examine the examples Hinten uses to support the claim that mistaken identity does not necessarily imply insanity. Joan of Arc and Mohammed thought they had seen angels and had a special role in history as a result. One can imagine that they could have been victims of some kind of hallucination or had some kind of experience that they misinterpreted, and that this could have happened without compromising their general soundness of mind, or their views of ethics. But the problem is that such examples are simply not relevant to Lewis’s argument. Joan and Mohammed did not claim to be God. That is, they did not claim to have existed from eternity in a special relationship with God the Father that made them Lord and gave them the authority to command the elements and to forgive sins. They did not claim that they had a prior existence in which they were omniscient, omnipotent, and omnipresent—all of which is implied in and entailed by the specific nature of Jesus’ claims. They did not claim that he who had seen them had seen the Father. They did not claim to be the Jahweh of the Patriarchs and Moses incarnate in human flesh! How is it possible to miss the profound difference between all other mistakes about one’s own identity and this one? One who wrongly believes that he is Napoleon has only confused himself with another finite human being. (Even this would present problems for the claim to be a great moral teacher. As Horner correctly observes, having correct views on ethics is a necessary but hardly a sufficient condition for being a great moral teacher. (Horner, 2008, p. 77).) As Kreeft notes, “A measure of your insanity is the size of the gap between what you think you are and what you really are” (Kreeft, 1988, p. 60). Indeed. Chesterton makes a similar point: “Normally speaking, the greater a man is, the less likely he is to make the very greatest claim. Outside the unique case we are considering, the only kind of man who ever does make that kind of claim is a very small man: a secretive or self- centered monomaniac” (Chesterton, 1925, p. 247). Pro: A Defense of C. S. Lewis’s “Trilemma” 183

Kreeft and Chesterton are right: To believe that one is Jahweh differs from all other such mistaken claims by an order of magnitude that is . . . well, infinite. It compounds a mistake of fact (“I am this finite created being, not that one”) with an error in metaphysics (“I am not a finite being at all, but the Ground of all Being”). This is not, as Lewis’s critics want to believe, merely a matter of degree. The gap between any creature and the Creator is a difference of kind. One might object that while the difference between the Creator and the creature is a difference of kind, the claim itself does not so differ from other claims, since all delusions are ontologically false to the same degree, that is, completely. But even if we accept this analysis and agree that all false claims are equally incorrect, it does not follow that all such errors are equally serious, much less morally equivalent. Falsely claiming to be Napoleon, for example, does not make one guilty of blasphemy. Mistaking one creature for another is an error, conceivably innocent; mistaking a creature for the Creator is idolatry. The error attributed to Jesus would be of the latter variety, and surely not irrelevant to his status as a great moral teacher—especially among first- century Jews! Anyone sincerely mistaken about being God would fail our third criterion for being a great moral teacher, being clearly out of touch with reality. Any first-century Jew so mistaken (and not so insane as to be wholly lacking in culpability) would run afoul of the second as well, being guilty of two of the most serious sins recognized by that society: blasphemy and idolatry. To put it bluntly, therefore, Lewis’s critics’ ability to rebut his argument depends on their ability to substitute a different and inferior argument while no one is looking and get away with it. When, like Lewis, we remember the radical nature of what Jesus actually claimed, and compare it with the ridiculously inadequate examples urged against the Trilemma, the attempts to evade its force become laughably absurd. An equal lack of attention to what Lewis actually said appears in the attempt to evade his claims about the implications of the relationship between Christ’s person and his teaching. Beversluis asks, “Did Lewis think that if Jesus were not God, there would no longer be any reason for believing that love is preferable to hate, humility to arrogance, charity to vindictiveness, meekness to oppressiveness, fidelity to adultery, or truthfulness to deception” (Beversluis, 1985, p. 55)? But Lewis was not evaluating the moral truth of Jesus’ teaching; he was examining the claims of the Teacher. His whole argument presupposes the truth of the teachings (cf. Lewis, 1943, p. 137), which is part of the evidence to be considered in evaluating the sanity of the Teacher. What is under scrutiny is the claims of the Teacher. Lewis is not saying that, if he were insane enough to wrongly think he was the omnipotent God, Jesus’ moral teaching would be refuted. He is saying that the truth of those teachings and their widely acknowledged superiority to all other attempts to state the same ideals is inconsistent with the notion that their 184 DONALD S. WILLIAMS source was either a blatant liar or a megalomaniac. Nothing that his critics have said makes those propositions any more consistent than they ever were before. Beversluis’s question is simply beside the point. Howard-Snyder is an exception to my dismissal of the attempts above to show that mistaken identity does not entail insanity, because he does try to deal with the specific case of mistakenly believing that one is God. Yet in reading his argument I cannot escape the impression that, having used the word “God” in one sentence, he immediately forgets in the next sentence what that word means. How else could anyone write with a straight face a sentence like this: “Merely being wrong about something important, even something as important as whether one is divine, neither implies nor makes it likely that one is a lunatic, insane, deranged, or otherwise fit to be institutionalized”? (Howard-Snyder, 2004, p. 463). It is not so much the “importance” as the nature of the claim to divinity that calls into question the sanity of any mere mortal who makes it, and guarantees the insanity of anyone who makes it falsely. Indeed, some of Jesus’ opponents, and for a while even members of his own family (see, for example, Mark 3:20), questioned his sanity—not surprisingly. They had not evacuated the word God of its meaning, or the concept of God of its transcendence. Howard-Snyder rhetorically softens the nature of the claim even with his diction, speaking of believing oneself to be “divine” rather than what is at issue, the concrete and personal claim to be God. I repeat: it is the claim to have existed from eternity in a special relationship with God the Father that made a person Lord and gave him the authority to command the elements and to forgive sins. It is the claim that he had a prior existence that was omniscient, omnipotent, and omnipresent. It is the claim that he who had seen this one had seen the Father. It was, particularly for Jesus, the claim to be the Jahweh of the Patriarchs and Moses, incarnate in human flesh. Howard-Snyder also confuses the issue by introducing the word institutionalized. The Trilemma does not require that a Jesus falsely claiming divinity would qualify for any specific modern diagnosis of a pathology justifying institutionalization; it only requires that he be unbalanced enough to be out of touch with reality and thus disqualified as a great moral teacher. Surely megalomania would suffice as such a disqualification? And surely the false claim to be God, made sincerely, would count as megalomania? If not, perhaps our requirements for “great moral teacher” have receded as far as our concept of what it takes to be God! If we remember what it means to be God, then, we must agree with Stephen T. Davis that we are “not prepared to allow that anybody other than God ever has sufficient reason to consider himself divine” (Davis, 2004, p. 491). Howard-Snyder’s attempts to imagine scenarios in which a sane person could be falsely persuaded that he is God fail at two points. First, he forgets the full meaning of what it would have meant for a devout first-century Jew to think he was God. Howard-Snyder realizes correctly that it would not be Pro: A Defense of C. S. Lewis’s “Trilemma” 185 enough for Satan to grant the power to do miracles, because prophets were believed to have performed miracles. So he has to have Satan reconstruct for Jesus the subjective experience of being God incarnate. The problem with this is that no one who has not been God incarnate could possibly know what that experience is. Hence, we have to ask, how would falsely assuming that one is having it not be megalomania? Having Jesus conclude his divinity through faulty exegesis of the Hebrew Bible runs up against the same problem. Surely a sane, nondivine person who understands the Judeo-Christian concept of God would conclude of any text that persuaded him that he was, contrary to all his experience, immortal, omniscient, and omnipotent, that there was a problem either with the text or with his reading of it. The second problem with Howard-Snyder’s scenarios is that, to establish the reasonableness of the sincerely-mistaken option, they would have to establish it, not for just any imaginable abstract figure, but for Jesus. Howard- Snyder lays down two ground rules at the outset: we must not treat the historical accounts as inspired Scripture, and we must not import into the discussion any independent evidence for Jesus’ divinity, such as his miracles, teaching, or resurrection (Howard-Snyder, 2004, p. 458). Many apologists are prepared to accept the first condition for the sake of argument; few are prepared to accept the second. There is a good reason for this refusal. The purpose of the Trilemma is not just to establish some abstract truth but to facilitate an encounter with Christ by clarifying the options of how we can understand him. And so the question is not whether Satan could have persuaded some typical first-century Jew that he was divine, but whether it makes sense to say that he could so have persuaded Jesus. Does Jesus strike us as a person who had been deluded by Satan? Interestingly, Jesus had his own answer to that scenario: if he did his great works by the power of Satan, then that would mean that Satan was fighting against his own kingdom, since Jesus’ works were clearly works of mercy and goodness (Matt. 12:25-28). In sum, the attempts to show that the Trilemma omits valid but unconsidered options all fail. To reject Lewis’s argument, you have to be prepared to affirm that a person in his right mind can sincerely but mistakenly believe, not simply that he has been visited by an angel, but that he is Almighty God, the Creator of the Universe, and still retain any credibility on anything else he might say. Since very few people in their right minds are prepared to accept that conclusion, most of Lewis’s critics are forced to try to undermine his argument by sneakily substituting a straw man for it. Refuting that weak substitution, they then pretend to have refuted the Trilemma. But no reader who is actually paying attention should fall for this shell game—for that is what it essentially is. Howard-Snyder’s attempt to support the sincerely mistaken option must be taken more seriously, for it does attempt to deal with the claim to be God rather than merely with the concept of mistaken identity in general. But it also fails by omitting to keep the full concept of deity in the forefront of our minds throughout the discussion. 186 DONALD S. WILLIAMS

5. Diminishing Probabilities?

Another attempt to find problems with the Trilemma does not attack its individual propositions but accepts for the sake of argument that they are each probably true. The problem is that when there are many such propositions, even if each is probably true, when the probabilities are multiplied together, the probability of the whole is significantly weakened. For example, if you have four propositions that are each probably true with a probability of .85, the probability of all four being true together is only .522—even odds, but hardly a compelling case. In the case of the Trilemma as Howard-Snyder analyzes it, you have to affirm that Jesus claimed to be God, that he was not lying, that he was not insane, and that he was not merely mistaken without being insane. If all four of these propositions are true, and these are the only possible alternatives, then it follows with deductive validity that he was telling the truth and was God. But all four are historical propositions, and therefore only probably true, because historical investigation cannot yield mathematical certainty. And all four, especially the first, are contested. Howard-Snyder gives what he considers charitable and generous ranges of probability to each proposition, ranging from .7-.9 for the claim to divinity to .85-.95 for the others, and ends up with a range of only .43-.77 for the whole. Therefore, he concludes, we should “profess ignorance and suspend judgment about the matter,” instead of claiming that the Trilemma shows it to be rational to believe in Jesus’ divinity (Howard-Snyder, 2004, p. 462). There are a number of ways we could respond to this argument. We could argue for higher values for the probabilities; but skeptics would have their own arguments for why they should be lower, and we would really be arguing the case for the truth of each proposition, which we are going to have to do anyway. Nevertheless, we have already argued that it is not necessary to follow Howard-Snyder’s rule about excluding evidence for the deity of Christ from outside the Trilemma itself. A person who looked at these four propositions in the light of the evidence for the resurrection set forth in a book like Frank Morison’s Who Moved the Stone? and in the light of the fulfillment of prophecy and other such apologetical evidence, might well come up with high enough values that the final result would still be quite believable. Howard-Snyder’s “range” (.43-.77) is simply a recognition that people come to different conclusions. One who thought with good reason that the actual probability was .77 (or higher) would hardly be required to suspend judgment simply because they knew that some people believe that the probability is much lower. Still, whatever values we assign must be less than absolute certainty. So far, therefore, the diminishing probabilities argument at worst can only qualify our confidence in the conclusion of the Trilemma; it does not overturn it. Pro: A Defense of C. S. Lewis’s “Trilemma” 187

I think the analysis I just gave is correct; but I also think that there is a deeper problem with the diminishing probabilities argument. It is easy to forget that in the Trilemma we are not simply debating various abstract propositions but ultimately dealing with our response to a person. The purpose of the argument is to enable us more intelligently to answer the basic question Jesus puts to us: “Who do you say that I, the Son of Man, am?” (Matt. 16:13- 18; cf. Brazier, 2012, pp. 103-106). Even Howard-Snyder admits that the Trilemma is deductively valid; his problem is the extent to which we can have confidence in the individual propositions. But the bottom-line question is whether I trust this Person that the historical accounts and the preaching of the Gospel present to me—even when He makes the most audacious claims. And one does not decide to trust another person simply by juggling a probabilistic calculus, but by responding to the Gestalt of his total personality. Of course, one is justified in doing so only as long as the propositions of the formal argument are believable both individually and together. If they were not, the Gestalt would not matter; if they were not, it would be a sign that the Gestalt was leading you astray. But one does not decide to trust a person on the basis of propositions and their logical relationships alone. In making this judgment in Jesus’ case, we gain clarity by using the Trilemma: by asking, “Is he lying? Is he crazy? Could he be just simply mistaken about this claim?” The Christian hopes that the response will be, “In his case—no, I don’t think so,” and that the Trilemma will then help to guide the seeker toward the logical response of faith: “He is telling the truth.” It will not be the Trilemma alone which generates this response, but rather the totality of Christ’s person as revealed in the Gospels (aided and brought into focus by the Trilemma and its validity) and brought home to the seeker by the Holy Spirit. Nothing less has ever produced that response or ever will. I think Lewis understood this truth, for at the end of his presentation in Mere Christianity he hopes that his elimination of the great moral teacher cop-out will push us back to Christ himself: “He has not left that open to us. He did not intend to” (Lewis, 1943, p. 56, emphasis added). Thus, the diminishing probabilities argument is not as impressive as it first seems, and is ultimately irrelevant to the way the Trilemma actually works. A similar attempt to weaken the apparent force of the Trilemma is the Extraordinary Claims Argument. On this view, an extraordinary claim (like the deity of the man Jesus) requires exceptionally strong evidence to be rationally believed. Historical arguments, by their nature are never more than probabilistic, and are thus incapable of providing such rational support. Therefore such claims cannot be supported by apologetic argument and must be believed if at all by sheer blind faith. The problem with the argument from extraordinary claims is that it cuts both ways. Is the notion that this vast, intricate, mathematically rational and fine-tuned universe just randomly popped into existence out of nothing and 188 DONALD S. WILLIAMS then proceeded to organize itself by pure chance not an extraordinary claim? Is the notion that the Disciples were all transformed from clueless cowards to men who turned the world upside down by a contention they knew to be false not an extraordinary claim? Is the notion that a merely human person can believe himself to be the omnipotent, eternal Creator of the universe and not be insane not an extraordinary claim? Surely they are. So if, when you think it through, you can avoid one extraordinary claim only by affirming another set of them, equally extraordinary, we must realize that the argument from extraordinary claims takes us nowhere and should therefore be abandoned. We simply have to make the best judgment we can on the evidence we have, however “extraordinary” the conclusion may seem to some to be.

6. Conclusion

How then do we evaluate the Trilemma as an apologetic argument? Brazier asks whether it is a failure and concludes, “No, because it generated speculation, got people talking” (Brazier, 2012, p. 186). It has certainly done at least that! And it has done much more as well. Lewis’s Trilemma is still a strong argument and can be used with confidence if we allow it to be nuanced and strengthened by its context in Lewis’s body of writings as a whole and if we understand its proper role in clarifying the options. It is unfair to take a paragraph aimed at a lay audience and complain that it is inadequate to deal with people who have a more sophisticated set of issues. Of course the classic passage from Mere Christianity needs to be supplemented when used with more sophisticated audiences, by Lewis’s other writings and by information and arguments that have come to light since he wrote. But the basic argument is sound. It is one thing to claim that it commits the fallacy of False Dilemma; it is quite another to show that other credible options actually exist. Lewis’s critics have simply failed to do that. The argument as presented by Lewis does not purport to prove the deity of Christ by itself, but it supports it by analyzing the logical options available and pointing out the difficulty of seeing Jesus as a liar or a lunatic. Attempts to see him as a liar or a lunatic are tendentious and ignore the actual facts of his life, and attempts to find other options, such as a sane person sincerely mistaken about his deity, fail in the same way, and fail doubly when we understand the real magnitude of the claim being made. Second, Lewis’s position as the dean of Christian apologists remains unchallenged. He was not infallible, but neither was he guilty of writing something in the Trilemma that was “not top-flight thinking” (Hinten, 2008, p. 8). His unique combination of wide learning, no-nonsense clarity, elegant language, and apt analogy remains as the standard to which Christian apologists should all aspire and the example they should seek to emulate. Pro: A Defense of C. S. Lewis’s “Trilemma” 189

When examined carefully, the Trilemma supports that conclusion; it is not an exception to it. Liar, Lunatic, or Lord? Lacking, Ludicrous, or Logical? Plunk for Liar or Lunatic if you must. But let us not come with any patronizing nonsense about how Lewis gave us a fallacious argument. He has not left that open to us. He did not intend to.

Fourteen

CON: LEWIS’S TRILEMMA: CASE NOT PROVEN

Adam Barkman

Donald Williams, whom I personally know and respect, is, like me, an unashamed admirer of C. S. Lewis. Yet for all our shared admiration, I suspect that Williams and I do not always admire all the same things about Lewis. Williams, for instance, begins his chapter noting, “It is curious that an argument that has become a staple of popular Christian apologetics should be rejected as fallacious by many who presumably accept its conclusion.” I am not sure what Williams finds “curious” about this, but certainly a Christian who is interested in the truth of the matter could find the Trilemma Argument problematic. Indeed, the thing I myself admire most in Lewis is his passion for truth and his willingness to pay the price in pursuit of it (cf. Lewis, 1999, p. 331). Williams, however, seems to suggest that Christian thinkers should simply accept an argument—never mind its validity—if its conclusion is something that we agree with. This is not the spirit of Lewis even if Lewis is the very person who articulated the Trilemma. Nevertheless, Williams has done us all a service by helpfully pointed out that not enough philosophers who have examined Lewis’s Trilemma Argument have properly appreciated what was in the culture at the time Lewis articulated this argument and to whom Lewis was speaking. If not brilliant, Lewis’s Trilemma Argument is fairly compelling if (1) we grant Lewis’s assumption that Jesus claimed to be God, and (2) we take into account Lewis’s intended audience (namely, ordinary BBC listeners during World War II). Williams is right that when Lewis is understood to be concerned with the same things that a modern analytic philosopher like Daniel Howard- Snyder (Howard-Snyder, 2004) wrestles with, Lewis is misrepresented and made to look a fool (when he was no such thing). But, admitting this, our discussion neither should assume (as Lewis did) that Jesus claimed to be God nor forget that what we are primarily interested in is not the Trilemma as a piece of popular apologetics, but its soundness as an argument directed at informed believers and nonbelievers today. In this extended philosophical discussion, we want to go further than Lewis could, and this means unpacking some of his assumptions. My claim is that taken as a rigorous piece of philosophy or theology, the Trilemma ends in a False Dilemma, fallaciously asserting that Jesus is either “liar, lunatic, or LORD (as in God).” Both prominent expanded versions of 192 ADAM BARKMAN the Trilemma Argument—(1) either liar, lunatic, legend, or Lord, or (2) either liar, lunatic, innocently mistaken, or Lord—add real possibilities to the argument, and thus make the argument better and stronger, even if the conclusion—Jesus is Lord—is no longer obvious. Yet my aim is not to rehash an old, or create a new, “Quadrilemma” (or something even more fanciful). If one simply wants a wide-ranging rehash, go with the much-discussed Howard-Snyder/Davis debate (Howard-Snyder, 2004; Davis, 2004). My take on the argument is a bit different; I think the Trilemma primarily fails because the argument wrongly assumes—that is, it assumes without argumentation where argumentation is needed—that Jesus claimed to be God (or “LORD”). To undercut the Trilemma Argument, all I really need to do is to show that reasonable doubt can be cast upon the claim that Jesus thought himself God. This can be done, but my larger intention in all this remains one with Lewis’s—first, to seek the truth, and second, to use the truth to show why people should become followers of Jesus.

1. A Failure of Imagination?

Unlike a few critics of the Trilemma, I side with biblical scholars like Craig Evans, Mark Strauss, Ben Witherington III, and N. T. Wright in asserting that there are good reasons to think the Gospel reports about Jesus are generally reliable, and so I have no particular interest in challenging the reliability of the reports themselves. My interest, rather, is in what is and is not reported. In particular, I want to begin by noting how often people, including Christians, demonstrate a notable failure to imagine new possibilities when new possibilities would be justified. Many Bible readers, whether they be ancient Hebrews or contemporary Christians, want to put difficult-to-classify entities into easily identifiable boxes—a tendency that is often responsible for asserting False Dilemmas. To make my point, consider two examples. A number of Christians have assumed that the “living creatures” or “cherubs” in Genesis, Exodus, Isaiah, Ezekiel and Revelation are simply really powerful angels; that they essentially belong to the same group (“angel” or “archangel”) as, say, Gabriel and Michael. While this is certainly a possibility, it is not much stronger than this. Yet the degree to which people embrace this connection between living creatures and angels shows a certain failure of imagination, in my opinion. Why not entertain the possibility of a new category of created being? Bats are not the same thing as birds. Or again, many Christians who read the Bible without a proper imagination as to real possibilities simply assert that there will be no marriages in Heaven or the New Earth, largely because Jesus said, “there will be no giving or taking of marriage, you will be like the angels” (Matt. 22:30). These Christians narrowly assert that there is only one way to enjoy marriage—here on this Earth. But note that although Jesus does indeed deny Con: Lewis’s Trilemma: Case Not Proven 193 the possibility of new marriages in Heaven or on the New Earth, he (1) does not deny that some marriages could continue in the New Earth, and (2) does not say that new marriages between humans are not possible ever again—he speaks, it seems, only of Heaven (which will pass away and give way to the New Heavens) and the New Earth. Ockham’s Razor—the principle that, all other things being equal, simpler explanations are to be preferred—is appropriately used when there are no good reasons to postulate further entities, but this is not what we see with, say, the living creatures or the possibility of some marriages continuing in Heaven. And something similar seems to be true of Jesus. Few contemporary New Testament scholars would argue that Jesus’ identity is very clear, yet what is remarkable among nearly all of these scholars is the felt need to slot Jesus into some pre-existing category—to simplify if you will. Rejecting the liar, lunatic, legend and simply mistaken categories, a large number of New Testament scholars see Jesus either as merely a wise man and so not God (“liberal theologians”), or the Son of Man, Son of God and the Messiah, and therefore God (“conservative theologians”). Lewis himself makes this substantial leap, moving from Jesus’ claim to be the Son of God to him being God himself:

A man who was merely a man and said the sort of things Jesus said would not be a great moral teacher. He would either be a lunatic—on the level with the man who says he is a poached egg—or else he would be the Devil of Hell. You must make your choice. Either this man was, and is, the Son of God: or else a madman or something worse. You can shut Him up for a fool, you can spit at Him and kill Him as a demon; or you can fall at His feet and call Him Lord and God. But let us not come with any patronizing nonsense about His being a great human teacher. He has not left that open to us. He did not intend to (Lewis, 1999, p. 56; emphasis added).

Now generally my own sympathies lie with the conservatives on these matters, yet what I find odd is that none of them seem willing, despite the evidence they themselves often present, to posit that while Jesus likely saw himself, and indeed was, the Son of Man, Son of God and Messiah, the evidence for him thinking himself, and being, part of the Godhead is much less compelling than the evidence for his thinking himself and being the Son of Man, Son of God and Messiah. And this is really at the heart of the argument, for even if the evidence for Jesus being the Son of Man, Son of God and Messiah is compelling (as I think it is), that does not warrant the quick assertion that Jesus is more likely God or part of God than simply a mysterious being—a being far more mysterious than a mere cherub—whose titles include Son of Man, Son of God and Messiah. If the Bible spoke clearly on the issue of Jesus’ identity, the tremendous Christological debates of the first three Christian centuries would not have occurred 194 ADAM BARKMAN

2. Son of Man, Son of God, and Messiah . . . But God?

Jesus’ favorite self-designation is “Son of Man” (Luke 22:29; Mark 10:35), which likely alludes to the “son of man” figure in Daniel 7, a figure given all authority over creation by God, the Ancient of Days: “He was given authority, glory and sovereign power; all peoples, nations and men of every language worshipped him. His dominion is an everlasting dominion that will not pass away, and his kingdom is one that will never be destroyed” (Dan. 7:13-14). Jesus as the Son of Man would certainly be a figure above the angels, cherubs (if they are different), human patriarchs, kings and prophets (Matt. 12:3-6, 41- 42), yet there is nothing in Daniel 7, Luke 22 or Mark 10 that suggests that the Son of Man is God or part of the Godhead. Indeed, without begging the question, a fairly straightforward reading of this is that the Son of Man is distinct from both God and man. If the Son of Man was given authority, there must have been a period when he lacked such authority. I myself can make little sense of the orthodox notion of the Son being eternally given authority. Consequently, if we do not downplay these passages (“liberals”) or read into them more than is there (“conservatives”), the Son of Man, whom Jesus does appear to have claimed to have been, appears to be God’s vicegerent in creation. And let us say, for the sake of argument, he was in fact this. It would be in this sense, rather than in the sense of Jesus having an authority distinct from God’s charge, that we should read Jesus’ claim that “the Son of Man is Lord even of the Sabbath” (Mark 2:28). Or again, when Jesus asserts that “one greater than the temple is here” (Matt. 12:6), he should be understood to be saying this as the Son of Man, who has been given authority to clarify and add to the law that God Himself gave, rather than saying this as God himself. This probably also applies to Jesus forgiving sins, for he does not say that he does this inasmuch as he is God, but rather “that you may know that the Son of Man has authority on earth to forgive sins” (Mark 2:10). To be fair to the evidence, it is worth noting that when the Jewish leaders say things like, “Who can forgive but God alone?” (Luke 5:21), Jesus does not feel the need to clarify for us the precise relation between the Son of Man, who has been given authority to forgive sins, and God Himself, who has the eternal right to forgive sins. With the notable exception of his trial (Mark 14:62), Jesus generally responded to false charges and yet was silent when people spoke truthfully of him. For instance, when Jesus eats with sinners (Matt. 26:57-68), heals the paralytic man (Mark 2:13-17), is accused of colluding with Beelzebub (Luke 11:14-28), is tested for signs (Matt. 16:1-4), is anointed by the sinful woman (Luke 7:36-50), and so on, Jesus speaks up. But when Jesus is called “the Holy One of God” (John 6:60-71), is identified as the Christ (John 11:25-27), triumphantly enters Jerusalem (Matt. 21:1-11), walks on water (Matt. 14:22-35), and so on, he is reported to have been silent, as if to acknowledge the claim being made of him. Is Jesus’ silence in the face Con: Lewis’s Trilemma: Case Not Proven 195 of the Jewish leaders’ charge that he is a blasphemous sinner clear evidence— as Lewis and others have taken it—that Jesus claims to be God or part of the Godhead? Certainly this could help that position somewhat, but it is nothing like clear evidence. After all, Jesus had already explicitly told us that he is the Son of Man, a being who has all the authority of God on Earth, which would ipso facto include the authority to forgive sins on God’s behalf. Without reading into this more than we ought to, the best explanation for Jesus’ silence is that he accepts the identification of God not as a literal identity claim, but rather as the Son of Man, who has God’s authority to heal, body and soul. Moreover, it seems that Jesus’ self-understanding as the Danielian Son of Man would also sufficiently account for his accepting certain types of “worship” or respectful gestures, for “He was given authority, glory and sovereign power; all peoples, nations and men of every language worshipped him” (Dan. 7:13). Jesus, we are told, received “worship” (proskuneō) from his disciples, and may have even encouraged it (John 14:1). But worship here— and in fact, the kind of worship that might be appropriate to the Son of Man— might simply have to do with prostrating oneself before him as toward a great being or king—“a customary gesture in the ancient world for someone to use when beseeching an authority figure” (Blomberg, 1997, p. 403). Of course—and again in fairness to the evidence—when such gestures are offered to some of God’s lesser authorities such as the disciples or angels (Acts 14:14-15; Rev. 22:8-9), such worship is refused as inappropriate— probably because it is distracting from the real object of worship, and possibly because such worship would violate the command, “Have no other gods before me” (Ex. 20:3). Here it could be argued that Jesus accepted worship because he accepted a strong identity claim as God. But again, if read in light of Daniel 7 (and not putting more into this than is warranted), Jesus is probably best understood as accepting a kind of worship appropriate to his being the Son of Man, and not the kind of worship due to God Himself as laid out in the Ten Commandments. Indeed, this interpretation would better fit the way Jesus himself taught us to pray—namely, to the Father (thus, “no other gods before Me”)—yet still in Jesus’ name (because Jesus was given authority over all creation by God, and so is the “way” to God) (John 16:23-24, 14:13- 14). A person’s faith in Jesus is credited to him as faith in God probably because God gave Jesus authority on Earth and not necessarily because Jesus is God Himself. In addition to claiming to be the Son of Man, Jesus also appears to identify himself not as a messiah (like King David or even Cyrus the Great (Is. 45:1), but as the Messiah—the anointed king who, on behalf of God, will establish God’s Forever Kingdom, which is to say a Kingdom where, among other things, the ethical will of God is accomplished everywhere. Caiaphas the high priest asks Jesus, “Are you the Messiah, the son of the Blessed [God]?” Jesus replies, “I am; and you will see the Son of Man seated at the right hand of Power [God], and coming with the clouds of heaven” (Mark 14:61-62). 196 ADAM BARKMAN

Here we can see that due to the Messiah’s scope of authority, he is clearly identified with the Son of Man, and, in fact, the two terms—Son of Man and Messiah—now seem to function as one. And this also appears true of Jesus being the unique Son of God. In the Bible, the term “son of God” is often used in a fairly weak sense and applies variously to certain celestial beings (including Satan) (Gen. 6:2; Job 1:6) and to mere humans (Luke 3:38). Yet in a few places in the Gospels it seems to be more a proper title or name than a term referring to “created rational beings,” if you will. For example, when Peter declares that Jesus is “the Son of the living God” (Matt. 16:16), this seems to refer to someone who is greater than “a person of angelic stature” or “a godly prophet.” Indeed, in the passage quoted in the previous paragraph, when Jesus admits that he is the Christ or Messiah, this term is connected not only with his being the Son of Man, but also his being “the Son of the Blessed One” (Mark 14:61-62). Ditto for “All things have been committed to me by my Father. No one knows the Son except the Father, and no one knows the Father except the Son and those to whom the Son chooses to reveal him” (Matt. 11:27; emphasis added). Although above the angels and other “sons of God,” Jesus—the Son of God in a unique sense—still identifies himself as a person occupying an epistemological (and possibly an ontological) position below the Father’s: “But about that day or hour no one knows, not even the angels in heaven, nor the Son, but only the Father” (Matt. 24:36); “The Father is greater than I” (John 14:28); “Why do you ask me about what is good? One there is who is good” (Matt. 19:17). Now again in fairness to the evidence, some have pointed both to Jesus’ numerous “I am” statements in the Gospel of John and his use of metaphors about himself which typically were applied to God Himself as evidence for Jesus claiming to be more than the Son of God, but also a person that is part of the Godhead. Eight times in the Gospel of John Jesus uses the phrase “I am” (for example, “I am the good shepherd,” “I am the resurrection and the life,” “Before Abraham was, I am,” and so on) to refer to himself, and this has been seen as important since “I am” was a name God used when He revealed Himself to Moses. Also, Jesus stated that “I and the Father are one” (John 10:30). Moreover, Jesus used metaphors about himself (the Lord of the harvest, the shepherd, the rock, and so forth) that were also used to talk about God Himself. Should the “I am” statements in John and in these metaphors be seen as compelling evidence of Jesus claiming to be God or the Lord? They certainly can help the traditional position somewhat, but even “conservative” scholars like Craig Blomberg warn that we “must be careful not to make either too little or too much of these claims” (Blomberg, 1997, p. 163). John Robinson argues that these passages need not be taken as unambiguous claims of Jesus’ divine self-consciousness, since neither the grammar of “I am” nor the metaphors need be exclusively used of God (Robinson, 1985, pp. 343- Con: Lewis’s Trilemma: Case Not Proven 197

397). To be clear, these passages do support the traditional Trinitarian case, but not as much as many—including, perhaps, Lewis—have thought. Limiting myself to things that Jesus probably said about himself, there seems to be good evidence that Jesus thought of himself as the Son of Man, Son of God and Messiah—a being above all cherubim, angels, prophets and kings, and one who can truly be called the Lord or adonai over all creation; nevertheless, it seems his Lordship is still one derived from a being distinct from his Lord—the LORD, if you will, Yahweh (cf. Ps. 110:1). This is to say that if we look for Trinitarian ideas in what Jesus said of himself, the evidence is not as strong as Lewis assumes, which is to say without much argumentation. And if this is so—if reasonable doubt can be cast on the question of whether Jesus thought of himself as God—then the Trilemma does not even get off the ground.

3. Would Jesus Agree With John?

Nevertheless, Lewis might be able to modify his Trilemma, arguing that even if it is not likely that Jesus himself claimed to be God, Jesus would have agreed with what his disciples said about him—notably John, who is often understood to be claiming that Jesus is God. Of course to go this route weakens an already weak argument since it seems just as likely as not that John misunderstood Jesus as it is to think that John was right to go a step beyond what Jesus claimed for himself. Lewis, though, does appeal to John in his Trilemma. He quotes John 20:28—where Doubting Thomas says to Jesus, “My Lord and my God!” upon seeing the scars of the resurrected Jesus. And while this passage might appear to help the Trinitarian interpretation insofar as Jesus is silent in the face of this exclamation (and silence in the case of Jesus typically means assent, as I have said), it is not obvious that Thomas was referring to one being rather than two. The question is, is Thomas referring to his Lord and master—Jesus—and God, or is he saying that Jesus is God? Lewis appears to have thought it referred to one being—that Thomas is calling Jesus God—but this is far from clear. Does the context and prologue of the Gospel of John help clarify this? Could this be why Lewis understands this passage in this way? The prologue of John is a theological reflection on Jesus which, even if we admit that it is inspired (and, of course, our sceptic here would not admit that), can be read in various ways. The “Word” that is “with God,” and indeed “was God” (John 1:1), could be either “God” or “god.” That is, the prologue could be read to agree with the Trinitarian idea that Jesus is of the same substance as the Father, but it could also be read, as the Arians did, as saying that Jesus is a unique divine being who existed before all else save God the Father (here Jesus as lowercase “g” god might be seen as a perfect reflection, as St. Paul 198 ADAM BARKMAN would have it, of uppercase “G” God). In other words, John could “merely” be saying that Jesus is a unique divine being—God’s unique Son—who helped create, and later redeem, all that was made. This use of “god” could then be seen as heightened usage of “god” in a passage like Psalm 45:6-7, where the Judean king is given the loose title of “god” (elohim, which would be translated as theos in Greek). Indeed, in John 10:34-36, Jesus himself, quoting Psalm 82:6, applies the term “god” to himself, and he seems to use it with a connotation of himself being below God but also more than a mere “god” (a created rational soul if you like):

Is it not written in your Law, ‘I have said you are gods?’ If he called them ‘gods,’ to whom the word of God came—and Scripture cannot be broken—what about the one whom the Father set apart as his very own and sent into the world? Why then do you accuse me of blasphemy because I said, ‘I am God’s Son?’ (John 10:34-36)

One could go on—pointing to John 14:28, where Jesus (at least as likely talking about ontology as function) says, “the Father is greater than me”—but I think I have sufficiently shown that Lewis and others have not been so obviously warranted in thinking Jesus (and even his disciples) understood the Son of God, Son of Man and Messiah to also be God (capital “G”) and LORD. And though space does not allow it, I think I could expand this to say that even in the letters of Paul, Peter and so on, such a claim is not as strong as Christians typically have thought.

4. “Another Credible Option”

In my opinion, the Trilemma is inadequate when push comes to shove because it requires a high probability that Jesus actually claimed (or even would have agreed with claims) that he is LORD or God. Many—though especially those with little taste for theological and philosophical history—have thought the issue obvious. On one side of the debate, Bart Ehrman thinks that Jesus is just a human being, and on the other Michael Bird (who mentions the importance of Lewis’s Trilemma) thinks that Jesus is the same substance as God Himself (Bird, 2014, p. 11). And yet neither—perhaps lacking imagination—ask if the best evidence about Jesus could be that he is the unique Son of God, the Son of Man, and Messiah—but not a mere human being, nor an angel, nor a cherub, nor a being consubstantial with the Father. Here I do not say more of what this might look like (since we are on the fringe of mystery), nor do I say this Arian view is my own personal view (such is beside the point); but I do insist that the possibility that Jesus is such a unique being is “another credible option,” as Donald Williams would have it, even if few bother (non-Christian) or dare (Christian) to entertain it. Con: Lewis’s Trilemma: Case Not Proven 199

Lewis also did not really explore this option, and one could more nearly imagine him working out something like Richard Swinburne’s a priori set of reasons for the Trinity (Swinburne 2010) before challenging the doctrine of Jesus’ consubstantiality with the Father. Perhaps it is his honest reading of Scripture that eliminated this “mere” option, or perhaps his faith in the apparent divinely-guided authority of the Church, or even his latent Idealism (which stayed with him in some form or another to his death). Who is to say? And so though I could imagine Lewis agreeing with a Gregory Boyd who says, “I would never say I am ‘inclined’ toward believing in the Trinity . . . or any other essential Christian teaching” (Boyd, 2013, p. 255 emphasis added), Lewis did wisely argue that believing X to be the case is different than knowing the internal workings or precise details of X. For example, Lewis was skeptical of the dominant Anselmian “satisfaction” theory of the Atonement, but that did not stop him from believing that Jesus’ life, death and resurrection makes possible the way back to the Father (Lewis, 1999, chap. 4). And so even if one would not mind if Lewis gave Arius as much due as Athanasius, or listened a bit more closely to John Milton’s claim that there is no good reason why a serious Christian should feel the need to uncritically accept the doctrine of Trinity (Milton, 1957, pp. 932-964), Lewis was certainly a lover of truth and a great constructor of arguments. Unfortunately, the Trilemma was not one of his great arguments, at least when one gets down to the details.

Fifteen

REPLY TO ADAM BARKMAN

Donald S. Williams

It is a privilege to share in this conversation on the Trilemma with Adam Barkman, whose response to my defense of the argument insightfully and helpfully focuses our attention right on the crux of the matter: the question of whether Jesus did or did not claim to be God. Let me then begin by highlighting our significant areas of agreement before proceeding to make some comments that I hope will help our readers in the task of deciding between us on the disputed point that remains.

1. Areas of Agreement

First, we agree that the “truth of the matter” is more important than our ability to score apologetic points. What I found curious in my initial statement was that so many profess to find the argument fallacious when (unlike Barkman) they are clearly analyzing an inferior argument to the one Lewis actually made—and when one would think they would have no motive for doing so. Second, we agree that the argument “is fairly compelling if we grant Lewis’s assumption that Jesus claimed to be God.” It is clear that the Trilemma becomes pointless in the absence of that claim. Third, we agree that what matters is “its soundness as an argument directed at informed believers and nonbelievers today.” (I would however distinguish soundness from effectiveness. The former is not dependent on the audience, while the latter is.) And, perhaps surprisingly, I even agree with Barkman that Lewis’s argument as presented in Mere Christianity “assumes without argument” something for which argument is now needed: “that Jesus claims to be God.” As I said in my original defense:

Anyone using the Trilemma today should be prepared to make the case that Jesus actually made the claims whenever it is needed. The wise apologist will not simply repeat Lewis’s paragraph from Mere Christianity, but rather adapt it to his own audience. This will involve notations such as “Here be prepared to insert ‘Modern Theology and Biblical Criticism,’ along with further updated arguments.”

202 DONALD S. WILLIAMS

2. At Issue: The Claims of Christ

What then remains? Our main point of disagreement appears to be over how strong the case for Jesus’ claims to deity actually is—and perhaps over how strong it needs to be. Barkman thinks that all that is needed to undermine the Trilemma is to “show that reasonable doubt can be cast upon the claim that Jesus thought himself God.” One could ask how much doubt is reasonable doubt? Do we need a slam-dunk case for the traditional view of Jesus’ claims, or is it sufficient simply to make a strong case that the traditional reading is the most reasonable interpretation of the texts? I would say that even if (which I am not admitting) the latter were all we could manage, that would be enough to preserve the relevance and usefulness of Lewis’s argument. It would be a “proof” only to the extent that we were certain about the claims. But even if we only thought them fairly likely, the Trilemma would still be a pointed exposition of the implications of those claims—and a compelling one, for all the reasons I gave in my original defense. But it is time to turn to the claims themselves. In his effort to show that Jesus’ claims to deity can reasonably be doubted, Barkman expends a great deal of labor establishing a point that no informed defender of the traditional reading of the Gospels contests: that titles such as Messiah, Son of Man, or even Son of God do not in and of themselves necessarily constitute claims to deity in a full Trinitarian sense. He is right. Each of them in isolation can be read either as claiming nothing out of the ordinary or as compatible with the claim to be, say, some kind of Arian god or angelic being or Platonic demiurge as opposed to being God Himself. Even the claim to forgive sin could be read as a claim only that God had authorized Jesus to pronounce forgiveness on His behalf. All this being correct, it might seem at first glance that Barkman has succeeded, not in overturning the traditional understanding of Jesus’ claims (which he does not claim to have done), but in showing that it can reasonably be doubted. Before we draw that conclusion, though, perhaps we should see what can be said on the other side.

3. The Case for the Traditional Reading

The case for the traditional reading is that, while some of the claims can be read otherwise in isolation, the claims of Christ taken as a totality, and read in the light of his contemporaries’ response to them, do indeed amount to a claim to be God. That totality is not limited to the possibly ambiguous titles mentioned above, but includes language such as “Before Abraham was, I AM” (John 8:58). Not, note, “Before Abraham was, I was,” a claim that would have been audacious enough. Jesus here, in no uncertain terms, identifies his own ego, his “I,” with that of Jahweh. In a first-century Jewish Reply to Adam Barkman 203 context, a more brash and bold claim to deity can hardly be imagined. And his contemporaries showed that they understood this claim quite well: They attempted to stone him for blasphemy (John 8:59). The other titles should be read, not in isolation, but in the context of the more unambiguous statements such as this one. Some will no doubt object that John’s language may be a later theological interpretation, not the ipsissima verba (“actual words”) of Jesus himself. I agree with Lewis (1967, p. 215) that “the assured results of modern criticism” on such matters is highly suspect on both methodological and presuppositional grounds. (See Gilson, 2014, for an interesting recent response to the “legend” hypothesis.) But even if we accept the theory that John’s reportage is actually an interpretation—something that has by no means been proved—it would still remain true that the totality of Jesus’ claims, as reported and understood by those nearest to them, includes a context that makes them read very differently from the impression given by modern attempts to scour them for potential loopholes. Starting with Jesus’ own contemporaries and coming down to the modern period, the strength of those claims demanded a response, whether that response was the attempt to evade them in favor of the brand of common- sense rationalism that was Arianism, or whether it was the careful and final definitions of the Council of Chalcedon. Barkman asks, if the New Testament were clear, why did the Christological controversies of the first three centuries take place? The answer is that they took place precisely because the claims were clear: They asked people to believe things that some of them had a hard time understanding or accepting; hence the various Christological heresies, which were attempts to alleviate those difficulties. But the fact that the totality of the claims was unavoidably there as a hard and inescapable datum meant that when every inadequate explanation had been weighed in the balance and rejected, the high and orthodox Chalcedonian Christology was what remained. You could not believe all that the New Testament says about Christ, including all that it reports him as saying about himself, without believing that he was “fully God and fully man.” It is obviously unfair to expect fully developed Nicene Trinitarianism from the earliest disciples of Jesus who first recorded his words and deeds as they remembered them. That is a bar that is anachronistic and unreasonable, so it is not surprising that the first generation of Christians does not clear it. What we could expect is that their recorded memories of Jesus’ life and claims when seen as a whole were such as to put them on a trajectory that would naturally and logically lead to Nicene and Chalcedonian orthodoxy once time to reflect on those claims had passed. And that is exactly what we find. By their own admission, at the time Jesus’ disciples did not know what to make of him. “What manner of man is this?” was their initial response— just what one would expect of people trying to get their heads around the experience (not yet the idea) of God incarnate. What they reported about him 204 DONALD S. WILLIAMS eventually led the church to conclude that he was “fully God and fully man.” How do we know they did not get it wrong? We know that because God raised him from the dead, which they took (as early as Peter’s sermon on the day of Pentecost) as the vindication of the claims. For me, the reactions of the original audience nail down this already strong case for the traditional reading. For example, we can speculate all we want about ways that Jesus’ act of forgiving the sins of the paralytic could have been taken. But such speculation is pointless, because we know how in fact it was taken: “Who is this man who speaks blasphemies? Who can forgive sins but God alone?” (Luke 5:21). (As I said in my original defense, Lewis’s argument deals with the reactions Jesus’ contemporaries actually had to him, not with the ones that modern scholars think they ought to have had.) Jesus, who was never slow to challenge preconceived notions he thought wrong, said nothing to correct this conclusion. His silence on this point speaks loudly when we read it in the light of John 8:58 (“Before Abraham was, I am”).

4. Conclusion

Now, Barkman agrees that these considerations weigh in favor of the traditional reading. He does not think they weigh as strongly as I do. What matters in the end is what our readers think. If you agree with me that the case against the traditional reading is not at all compelling, then the Trilemma tells you what you should conclude about Jesus. If you agree with my learned colleague that significant doubt remains about Jesus’ claims, then for you the Trilemma ceases to function as a proof. (Remember that Lewis did not present the Trilemma as a proof of the deity of Christ as such; he only claimed that it had eliminated the easy way out of seeing Jesus as merely a great moral teacher. Nevertheless, people since his time have used it as a proof, understandably so, as its implications do lead in the direction of accepting Christ as God incarnate. Consequently, evaluating the argument as a proof is a legitimate part of our enquiry.) But even if we do not use the Trilemma as a proof of Christ’s deity, it still remains useful and insightful as a guide to the implications of his claims, should the traditional reading of them be correct. Perhaps a kind of Pascal’s Wager comes into play at this point. Given the seriousness of the implications of Christ’s claims as traditionally understood, given the fact that those claims have been called into question but hardly disproved, given the weight of the testimony to those claims of the whole Christian tradition, I would want a pretty strong case against them, not just some theoretical possibility of doubt, before I would dismiss them. I do not think that such a case exists. And so, like Lewis, I conclude as Jesus’ first followers did: Jesus was not a liar or insane; he was not merely a good human moral teacher; therefore, Jesus Christ is Lord. Sixteen

REPLY TO DONALD WILLIAMS

Adam Barkman

Despite a few loaded terms and the rare uncharitable implication, both Williams’s initial paper and his subsequent reply to my chapter have been instructive in content and courteous in tone. I personally have been bettered by this exchange with him, and I trust the reader has been as well. Nevertheless, while progress has been made, some disagreement between us remains. Williams’s reply emphasizes three key issues that he hopes will establish Jesus’ likely claim to Deity, which will then open up the possibility of an effective Trilemma Argument. These three issues are: (a) Jesus’ alleged implicit claims to Deity, (b) the reliability and proper interpretation of the Gospel of John, and (c) the apostolic origins and orthodox status of the doctrine of the Trinity in Christian tradition.

1. Jesus’ Apparent Implicit Claims to Deity

Like Williams, philosopher Richard Swinburne argues that in his first-century Jewish context Jesus could not “publically and explicitly say that he was God,” yet he implied as much in both his actions and his words (Swinburne, 2010, p. 98). Now in general, I agree that God does seem to be slowly, progressively preparing the world for what is coming next. But is this what is happening with Jesus’ alleged implicit claims to Deity? Did Jesus tacitly claim to be God (by forgiving sins and uttering a number of “I am” statements), and then guide further generations to explicitly make this claim? Possibly, but I have some doubts. Swinburne, Williams, and Lewis himself all take Jesus’ professed authority to forgive sins as an implicit claim to Deity that Jesus later made explicit. But two objections arise, one of which I highlighted in my first paper. First, is the fact that Jewish leaders regarded Jesus’ claim to forgive sins as “blasphemous” strong evidence that they thought Jesus was actually claiming to be God? The Gospels have some of the scribes say, “Who can forgive sins but God alone?” (Mark 2:7), but they could just be saying that only God or God’s Appointed could forgive, and that Jesus was neither. Second, the Jewish leaders likely knew—but perhaps did not understand—the Book of Daniel, in which “one like a son of man” (Dan 7:13) is given “dominion and glory and kingdom” by God, including (presumably) the authority to forgive 206 ADAM BARKMAN sins. These scribes might have known this passage in Daniel but denied that Jesus was this divinely appointed figure, or they might have interpreted the passage differently, or they might not have read Daniel very carefully. On any scenario, it is questionable whether Jesus’ apparent identification of himself with the Danielic son of man is tantamount to a claim to be consubstantial with God Himself. As I see it, the burden of proof here lies with the Trinitarians. Granting the admittedly debatable claim that the Gospel of John is a reliable record of what Jesus said, how should we understand Jesus’ eight “I am” statements in that gospel? Williams is certainly right that these statements could recall Exodus 3:14 (Yahweh’s “I Am Who I Am”), but is this the best interpretation? If Jesus’ words were understood to allude to the Exodus passage, then we would have expected them to provoke extreme outrage each time they were uttered. But that is not what generally occurred. When Jesus says, “I am the bread of life” (John 6:35), the Jewish leaders “complain,” but not about those two words themselves, but about Jesus’ claiming to be from Heaven. The second time—“I am the light of the world” (John 8:12)—Jesus is again “challenged,” but only over his claim to possess teaching authority. The fourth and fifth times—“I am the gate for the sheep” (John 10:7) and “I am the good shepherd” (John 10:11)—Jesus is called “crazy” by some, but only because he claims to be the means back to God. The sixth time is when Jesus states, “I am the resurrection and the life” prior to his raising Lazarus from the dead (John 11:25). Here, too, the fact that Jesus uses the “I am” phrase does not outrage anyone, though the Jewish leaders do feel threatened by Jesus’ apparently miraculous powers. Finally, the seventh and eighth times are when Jesus says, “I am the way, and the truth, and the life” (John 14:6) and “I am the true vine” (John 15:1). These utterances are directed at the disciples, and so do not even elicit the reaction of complaint or challenge, but rather confusion. So, there is little evidence that Jesus’ use of “I am” statements in John were either intended or understood to claim that he was God. The one apparent exception is John 8:58, when the Jews in the courtyard of the Temple pick up stones to throw at Jesus after he declares, “Before Abraham was, I am.” Here it is important to understand the context, however. Jesus’ statement is the culmination of a long-running argument with the Jews in which he declares that they are not true children of Abraham, that their father is the devil, and that they will “die in [their] sins.” Jesus’ declaration that he is a pre-existent Messianic being, older and greater than Abraham, and that “God is the source of my being” (John 8:42, NEB), is simply the final straw in this provocative exchange. The only other time Jesus uses an “I am” statement that causes a real uproar is in Mark 14:62, when Jesus is asked by the chief priests whether he is the “Messiah, the Son of the Blessed,” and Jesus replies that “I am.” Here, though, the uproar is likely about Jesus’ claiming to be the Messiah, for if Jesus is the Messiah, and the priests are against him, then the priests are in Reply to Donald Williams 207 serious trouble with God. But, of course, the priests are convinced that Jesus was not the Messiah because, as they saw it, he did not keep the law of Moses, as they supposed the Messiah surely would. They probably thought that Jesus blasphemed God by contradicting Him, favoring his own interpretations of Scripture over God’s clear commands. But again, the “I am” is not the central issue here. In short, the Trinitarian explanation needs to do much more work if the “I am” statements are to be understood as claims to identity with God.

2. John’s Prologue

Williams thinks—and he is far from alone in thinking—that the Prologue of John’s Gospel is more or less a slam dunk for Trinitarians. But here we need to note two things. First, the Trilemma Argument—which is what our exchange is ultimately about—is about whether Jesus claimed to be God, not about John’s understanding of Jesus. Second, even if we concede that Jesus would have fully agreed with John, one can still reasonably ask whether the Prologue supports belief in the Trinity, especially given the way “God” and “gods” are used throughout that Gospel. I agree with Williams that the Prologue, taken in isolation, is best read as saying that the “Word is the God” (capital “G”), based on the typical usage of article transferences in ancient Greek. But the Greek phrase could also be read as saying “a god” or “a divine being,” which could speak to Jesus’ role as the Son of God, Son of Man, Messiah, and as a “divine being.” Even if Jesus was not God, he could be, as John says, a being who existed before our universe, and the one by whom all angels and humans and other, lesser “divine beings” were created. Indeed, when the Prologue is read in the context of John’s Gospel as a whole, the Trinitarian interpretation becomes less convincing. Consider the following. When Jews in the Temple want to stone Jesus for saying “The Father and I are one” (John 10:30), thus committing blasphemy for “mak[ing] yourself God,” Jesus responds:

Is it not written in your Law, “I said, you are gods”? If he called them gods to whom the word of God came—and Scripture cannot be broken— do you say of him whom the Father consecrated and sent into the world, “You are blaspheming,” because I said, “I am the Son of God”? If I am not doing the works of my Father, then do not believe me; but if I do them, even though you do not believe me, believe the works, that you may know and understand that the Father is in me and I am in the Father.

Here Jesus distances himself from the strong claim to be literally God, but merely asserts himself to be the “consecrated” Son of God who lives in especially close communion with the Father. To this end, Jesus quotes Psalm 82:6 (“You are gods”) to establish that talk about humans and angels as 208 ADAM BARKMAN

“gods” or “divine beings” is legitimate and scripturally-sanctioned. Yet while not claiming to be literally of the same substance as God Himself, Jesus does insist that he is not merely a regular “divine being” in the angelic or human sense, but something more. Regardless of whether the Gospel of John gives or does not give us a slightly more theological interpretation of Jesus, the burden of proof still rests with Trinitarians to show that John—and John’s Jesus— really thought Jesus to be God Himself. The claim to Deity is such a radical claim—and one so weakly, if at all, implied by the Synoptic Gospels’ portrayal of Jesus—that we need better arguments for Jesus’ claim to be God than Williams gives us. Now I know that my arguing as I do here seems to Williams that I am looking for “loopholes” rather than the truth, but the truth is that Scripture as a whole does not give the strong support to Trinitarianism that Williams would like. Of course, as a fairly conservative, largely orthodox Christian, part of me would like Scripture to give more support to the doctrine; yet I am with Lewis the philosopher here: better the hard truth than a sweet lie. Indeed, if the doctrine of the Trinity is to be believed, we likely will need some non- Scriptural—yet potentially God-given—support to turn the tide. And this, Williams thinks, he can give us.

3. Tradition, Probability and Wager

Williams speaks about the dangers of a “common-sense rationalism” that would challenge the doctrine of the Trinity, but what exactly is his preferred alternative? Possibly Williams means that non-common-sensical Trinitarians will be more epistemically humble and make more room for mystery than common-sensical non-Trinitarians do. But both of those implications seem off. I take it that a mystery in the broadest sense of the word is an unsettled logical possibility or probability—something that humans should leave open or not believe with cocksure assurance—and a theological mystery is a logical theological possibility or probability that humans do not yet understand well, yet still know enough to know that it is neither a contradiction nor a certainty. Whether God has made a cotton candy planet somewhere is a mystery; whether Confucius is in Heaven is a mystery; and whether Jesus is, or is not, of the same substance as God is also a mystery. The point is that both the Trinitarian and non-Trinitarian views of Jesus are in some sense mysteries. Of course, some things are greater mysteries than others. Is the doctrine of the Trinity more mysterious than, say, a conception of Jesus as a pre-existent being who became incarnate in human flesh? Perhaps. But it hardly follows that both are equally good explanations of the relationship between Jesus and the Father. Very few things are genuinely equal, and the same is true here. But all this I understand as Lewis did, which is to say by way of Williams’s “common-sense rationalism,” that is, by thinking hard about the evidence Reply to Donald Williams 209 before me. My point is that “common-sense rationalism” is humble in recognizing the limits of what we can know, yet “rational” in insisting that beliefs be properly supported by evidence. The question between Trinitarians and non-Trinitarians is not whether either view is a contradiction or a certainty, but rather which view, after careful reflection, is more probable. Williams notes that the Christian tradition favors the doctrine of the Trinity. I agree, though I would argue that the early Church’s understanding of Jesus was not as obviously Trinitarian as Williams claims (cf. Hanson, 2005, pp. xviii-xviv; Wiles, 2004, p. 3). In support of Williams, I would further add that Trinitarianism has in its fold nearly all the best Christian minds (certainly all my heroes, save for the non-Trinitarian, Sir Isaac Newton). Yet it hardly follows—at least not for Protestants, who feel a certain liberty to challenge tradition here and there—that tradition and the best minds are always right. Serious Christian philosophers nowadays are showing some very real problems with some traditional doctrines, such as the Lake of Fire as perpetual suffering (as opposed to Annihilationism, the view that the holding- cell “hell” is utterly destroyed in the metaphorical “Lake of Fire”), and omniscience as knowing all future propositions (as opposed to knowing all possible future propositions). Of course, being a fairly traditional Christian myself, I agree with Williams that all things being equal, the general consensus of the Church is evidence in favor of a traditional belief, and so, in this case, evidence of Trinity being true. Yet Williams puts more stock in this witness than I do. To immerse one’s self in the early years of the Arian- Trinitarian debate, for example, has the effect of making “tradition” appear to be a much less clear entity. In this debate—where both sides can claim the guidance of God—there is a lot of rhetoric, but often not a lot of careful, objective analysis. Moreover, even if one were hypothetically to grant that Trinitarianism had the better arguments when it triumphed over its various opponents in the early Christian centuries, it hardly follows that it is still the best view today. We have to go, as Lewis says, where the evidence leads us. Now against this general claim, Williams’s proposes a kind of Pascalian Wager: if the evidence for and against Trinity is roughly the same, go with Trinity because of tradition. But, for one thing, neither Williams nor I think the evidence is exactly or even nearly 50-50. For another thing, Williams fails to note the potential dark-side of his Wager. For consider: If the non- Trinitarian view is wrong, nothing of substance will be lost for non-Trinitarian Christians (salvation presumably being secured by Jesus’ sacrifice and by faith in him). Yet if Trinitarians are wrong, then they will have committed a grave blasphemy, namely, elevating Jesus to the same level as God. I trust that Jesus would forgive us poor Trinitarians even of this, but it is still a very serious matter that Williams completely ignores. So ironically, Williams’s Pascalian Wager, if it succeeds at all, seems to support non-Trinitarianism. Of course, my aim here is not to problematize the doctrine of Trinity for its own sake, but rather to show that it can reasonably be challenged, and, 210 ADAM BARKMAN more to the point, that Jesus’ claim to Deity or “Lord” is not as well- established as Lewis supposed. Thus, while Williams and I agree that “Lord” is a far more plausible alternative than either “liar” or “lunatic,” I believe that there is another reasonable possibility: that Jesus (without being either insane or unethical) claimed to be “merely” the unique Son of God, Son of Man, and Messiah. I do not say, of course, that this alternative is my own personal view, but I do admit that it is a reasonable view, and so, ipso facto, enough to make Lewis’s Trilemma Argument a failure.

Seventeen

PRO: THE PROBLEM OF PAIN DEFENDED

Philip Tallon

In 1939, Ashley Sampson of Centenary Press invited C. S. Lewis to write on the problem of evil for the Christian Challenge series. Lewis asked if he could write anonymously, because he felt he did not live up to the hard-edged principles he would need to espouse. This would not have been first time (or the last) that Lewis published anonymously. His early poems appeared under the name “Clive Hamilton” and the memoir of his wife’s death, A Grief Observed, was attributed to “N. W. Clerk” during Lewis’s lifetime. The publisher, however, refused his request for anonymity. And so with the publication of The Problem of Pain in 1940, Lewis kick-started his role as a public philosopher and theologian. Later works in this vein would include the radio talks that were published as Mere Christianity (1952), as well as books like Miracles (1947) and The Four Loves (1960), along with numerous essays on theological and philosophical topics. But The Problem of Pain was the first, and constituted a strong entry into the public sphere. Despite some self-deprecatory remarks by Lewis at the start of the book (“If any real theologian reads these pages he will very easily see that they are the work of a layman and an amateur”), Lewis’s work is thoughtful, scriptural, steeped in the Western tradition, and ambitious in scope (Lewis, 2001c, p. vii). Within 150 pages or so, Lewis handles the origins of religion, articulates the problem of evil in various forms, deals with free will and soul making, offers an imaginative reconstruction of the origins of humanity and its fall, discusses the nature of pain in some depth, grapples with the problem of animal suffering, and theologizes about heaven and hell. Despite the wide range of subjects, Lewis gives each topic serious thought. Though his treatment of any particular topic is rarely comprehensive, the quality of Lewis’s thinking on the problem of evil is impressive, and he anticipates many major developments in the philosophical literature over the next 60 years. It is worth emphasizing at the outset that Lewis is responding to the intellectual problem of evil in The Problem of Pain (as well as in sections of Mere Christianity). Though pastoral or emotional concerns are not far from his heart, they are not his focus:

[F]or the far higher task of teaching fortitude and patience I was never fool enough to suppose myself qualified, nor have I anything to offer my readers except my conviction that when pain is to be borne, a little 212 PHILIP TALLON

courage helps more than much knowledge, a little human sympathy more than much courage, and the least tincture of the love of God more than all (Lewis, 2001c, p. vii).

If one is looking for an existential treatment of suffering, it is hard to find a better example than Lewis’s own book, A Grief Observed, which many sufferers have praised for its honesty and insight. Meghan O’Rourke, author of The Long Goodbye, a memoir about the loss of her mother, said that “A Grief Observed got me through the raw, hallucinatory first months. It made me feel less alone, because he really captures the experience of obsessiveness, loneliness, and embarrassment that many mourners feel” (O’Rourke, 2011). In the long run, of course, the intellectual problem and the existential problem are inseparable. The source of ultimate hope and comfort rests on the existence of a good God with the power to set everything right in the end. Though critical thinking about the reasons for pain may not always be comforting in the grip of suffering, the trustworthiness of the Christian gospel, which Christian theodicy partly defends, is necessary for any truly hopeful response to the problem of pain.

1. The Main Points of Lewis’s Theodicy

The argument of the The Problem of Pain gets rolling at the beginning of the second chapter, where Lewis offers a variation on the traditional form of the logical problem of evil:

If God were good, He would wish to make His creatures perfectly happy, and if God were almighty He would be able to do what He wished. But the creatures are not happy. Therefore God lacks either goodness, or power, or both (Lewis, 2001c, p. 16).

This argument is formally valid; its logical pattern is plain old modus tollens (“If A then B; not B; so not A”). Lewis rightly sees then that challenging the argument requires challenging the premises, which he does by arguing that the meanings of “good,” “almighty,” and possibly “happy” must be shown to be “equivocal” (16). The word “equivocal” here can be slightly confusing, since its usage in common theological jargon sometimes implies the same word having two completely different meanings (for example, the ‘bark” of a tree and the “bark” of a dog). But Lewis uses “equivocal” to mean ambiguous. “[I]f the popular meanings attached to these words are the best, or the only possible, meanings,” Lewis writes, “then the argument is unanswerable” (Lewis, 2001c, p. 16). The second and third chapters of The Problem of Pain focus on defining God’s power and goodness in such a way that they do not imply an idyllic Pro: The Problem of Pain Defended 213 world that is free of pain and suffering. Lewis accomplishes his goals by challenging both of the initial premises (the ones about God’s goodness and power). If his challenge to either premise succeeds then the argument fails to refute traditional Christian theism. I think that both his responses do succeed, as we will see by the end of this essay. It is also important to note that in the process of responding to these two premises, Lewis unfolds his discussion in such a way that these topics shape and guide the book’s larger concerns. Thus, a helpful structure for understanding the shape of Lewis’s theodicy (and a helpful way from preventing this essay from being a simple reiteration of the book’s main points) is to look at Lewis’s treatment of omnipotence and free will, followed by his treatment of divine goodness and suffering (in both its human and animal varieties).

2. Omnipotence and Free Will

Lewis understands God’s omnipotence to mean the “power to do all that is intrinsically possible, not to do the intrinsically impossible” (Lewis, 2001c, p. 18). “Intrinsic” possibility is synonymous here with broadly “logical” possibility. Logical impossibilities include explicit contradictions in terms, such as the notions of a “shapeless cube” or a “married bachelor.” Lewis quickly elaborates on the need for this qualification to divine omnipotence as it pertains to free will. We might ask whether God, in His almighty power, could give us free will and yet also withhold free will from us (Lewis, 2001c, p. 18). Lewis argues that He could not, for this would involve an “intrinsic” contradiction akin to a shapeless cube. It might make a certain sort of linguistic sense. We can combine this adjective and noun in common usage. But it fails to make logical sense. God’s power is not limited by His inability to do utterly nonsensical things. In a memorable turn of phrase, Lewis quips, “meaningless combinations of words do not suddenly acquire meaning simply because we prefix to them the two other words ‘God can’” (Lewis, 2001c, p. 18). God wills that humans obey His commands and seek the good, but He also wills that we do so freely. It is reasonable to assume that libertarian free will (where humans supply the sufficient conditions for a free decision) is assumed in all of Lewis’s theology. There are some passages in Lewis that might seem to imply a kind of theological determinism. In Surprised by Joy, Lewis describes his search for God as “the mouse’s search for the cat,” and he uses strong language about the feeling of compulsion that accompanied his decision of faith (Lewis, 1955, pp. 227-229). But these passages reflect a traditional understanding of God’s grace drawing sinners to salvation, and so imply only that grace enables our free response to God. Nowhere does Lewis endorse anything like compatibilism (the view that our choices can be both fully determined and significantly free). 214 PHILIP TALLON

It is hard to overestimate the importance of libertarian free will in Lewis’s theology. The theme is repeated many times in The Problem of Pain, and is also taken up in Mere Christianity: “[F]ree will, though it makes evil possible, is also the only thing that makes possible any love or goodness or joy worth having. A world of automata—of creatures that worked like machines—would hardly be worth creating” (Lewis, 2001a, p. 48). Here Lewis passes along key elements of Augustinian theodicy, and also anticipates later work done by Alvin Plantinga. Plantinga’s words echo Lewis’s in Plantinga’s book, God, Freedom, and Evil. “A world containing creatures who are significantly free,” Plantinga writes, “is more valuable, all else being equal, than a world containing no free creatures at all” (Plantinga, 1974, p. 30). Free will offers a greater-goods defense against the problem of evil. If some significant good can only be achieved only at the cost of permitting some lesser evil, then God might be justified in allowing that evil. A runner, for instance, may know that training for a marathon will be the cause of many aches and pains, and will require great personal cost. But she weighs those costs against the greater good of high levels of fitness and great accomplishment. Evils, in many cases, can only be avoided by withholding some greater good. As Plantinga says: “The fact that free creatures sometimes go wrong . . . counts neither against God's omnipotence nor against His goodness; for He could have forestalled the occurrence of moral evil only by removing the possibility of moral good” (Plantinga, 1974, p. 30). God wills to create free creatures, and in doing so, allows the possibility of willful rebellion. Even in this, God’s ultimate will is not thwarted. In Mere Christianity, Lewis provides the analogy of the mother who wishes for her children to keep the room clean, but also for them to do it of their own choosing. The mother returns at the end of the night and finds “the Teddy bear and the ink and the French Grammar all lying in the grate” (Lewis, 2001a, p. 47). This is against the mother’s will for a tidy room, but consistent with the possible states of affairs which she herself has allowed. But here an objection might be raised. Does free will necessarily produce suffering? Could not God have given us libertarian free will but also rigged the system so that no sin ever hurts anybody? This would be a bit like the mother in the example above giving her children the choice to clean up after themselves, but also employing an army of robots to immediately put away any toys they did decide to drop on the floor. This would preserve free choice in one sense, but also avoid any bad effects that might result from it. Against this objection, Lewis argues that a necessary condition for “a society of free souls” is “a relatively independent and ‘inexorable’ Nature” (Lewis, 2001c, p. 19). To have relationships, we must have an awareness of the existence of others, and this requires an environment within which to relate and connect. Could embodied beings such as us even interact with God, others, or themselves without a physical world in which to do so? Without such a world, how could we distinguish between ourselves and others, or make choices that Pro: The Problem of Pain Defended 215 have effects outside our own minds? “A creature with no environment would have no choices to make,” Lewis notes (Lewis, 2001c, p. 20). Any environment for personal interaction and effective decision-making must have a consistent order, and so any stable environment carries the possibility of running afoul of our immediate desires or transmitting the repercussions of our bad choices. Careless walking may result in stubbed toes. Unreciprocated affections may cause disappointment. The same solidity that makes a piece of wood useful for building houses also allows it to be used as a fearsome weapon. Lewis acknowledges that at times God does intervene supernaturally in the world by preventing some bad effect or initiating something good, but overall stability “demands that these occasions should be extremely rare” (Lewis, 2001c, p. 25). A world in which all ill choices (moral or otherwise) caused no ill effects would undermine the integrity of a stable environment, which is required for us to be meaningfully responsible and free. Richard Swinburne offers a similar argument in Providence and the Problem of Evil, developing it into what he calls “the principle of honesty” (Swinburne, 1998, p. 138). “God could of course arrange things so that our bad choices never had any effects. When we chose kind words, they came out of our mouth; when we chose to insult, the air did not convey the message” (Swinburne, 1998, p. 138). Without the possibility of evil choices humans would be held captive to an illusion about the meaning of their own agency. Such a consistently unstable environment would constitute a systematic deception on God’s part and be inconsistent with His goodness. For Lewis, then, God’s choice to create genuinely free creatures independent from Himself and others implies both the real possibility of rebellion and a stable natural order within which to exist. The conditions of freedom must carry consequences. In his discussion of the Christian doctrine of the fall, Lewis applies a number of his points about the necessity of a stable environment. God’s intention to be in relationship with meaningfully free creatures entails a firm commitment on God’s part to conditions in which their freedom can be expressed. Because of our rebellion, humans are now bent toward sin and able to suffer and die. God could have miraculously removed the results of human sin, for a first, a second, and a third time, but eventually He would have had to allow humans to pay the price. If not, the integrity of the system would be undermined. You can waive the rules a few times in a game, but not every time, or else the game loses its purpose. We may wonder at this point how it was that humans went from a state of happiness and blessedness to their fallen condition. If everything was perfect, indeed, if humans were “completely good and completely happy,” how did we ever make so ruinous a decision as to throw all that away? (Lewis, 2001c, p. 66). How can sin arise from perfect goodness? Lewis responds to this objection by arguing that the necessary conditions for sin are extremely easy to come by. In fact, they are unavoidable. All that is needed is an awareness of 216 PHILIP TALLON

God and of the self as something other than God, and the choice to be selfish is already at hand (Lewis, 2001c, p. 70). Even in the paradise-like Garden of Eden, the possibility of wrongdoing was present. Lewis’s treatment of Eden is interesting. In a long and creative aside, he imagines what life before the fall might have been like. The earliest humans were perhaps elevated out of the primate herd and blessed with a spiritual element. They also had (Lewis imagines) control over their animal nature and a greater measure of mastery over the environment. Such creatures would be raised above many of the sufferings of embodied life, and thus be protected from the travails of a world outside the Garden that developed by natural means (including Darwinian processes) (Lewis, 2001c, p. 77). Self-will, however, intervened and the creatures decided to claim as their own what was always borrowed from God, and thus fell into ruin (Lewis, 2001c, pp. 74-76). Disorder, disease, and death followed from disobedience. Humans are now a “spoiled species,” “a horror to God” in their fallen state (Lewis, 2001c, pp. 81, 63). As a sign of his thorough commitment to the notion of the integrity of the environment God created, Lewis stresses that humanity’s current state is not simply a matter of retributive punishment, but rather the natural and necessary consequence of original sin (Lewis, 2001c, pp. 64-65). Because early man rebelled against God, “the total organism which had been taken up into his spiritual life was allowed to fall back into the merely natural condition from which, at his making, it had been raised” (Lewis, 2001c, p. 78). A just and loving God could not pretend that a ruptured relationship was still intact, or that serious sin should be ignored. Because of the “rules of the game,” human evil inevitably leads to calamity. In a passage that evokes the aesthetic motifs invoked by earlier theologians such as Athanasius and Irenaeus, Lewis uses the analogy of artistic creation, whereby God, the master artist, instead of stopping the play when it has gone wrong, finds a way to enter into the disrupted performance to set things right from the inside: “The world is a dance in which good, descending from God, is disturbed by evil arising from the creatures, and the resulting conflict is resolved by God’s own assumption of the suffering nature which evil produces” (Lewis, 2001c, p. 80). The “rules” which govern creation should not be understood as mechanistic physical laws, but more like the standards that govern a great work of art with its own integrity and sense of proportion. By the miracle of divine creativity humans participate in the co-creation of the work, and even when they disrupt the drama, God’s ingenuity finds a way to bring about a beautiful resolution through the incarnation of Jesus Christ and the redemption of the world. Lewis is keen to point out that the ultimate undoing of the effects of human wickedness does not excuse the wickedness itself. God weaves simple evil into “complex good” (Lewis, 2001c, p. 111). But the doers of simple evil Pro: The Problem of Pain Defended 217 may still suffer the consequences, even into eternity. In his discussion of hell, Lewis brings to bear the twin emphases on free will and the natural consequences from rebellion in a helpful way that also has been built upon by later philosophers. Hell, for Lewis, is simply the ultimate consequence of the seriousness with which God takes the gift of freedom. In The Great Divorce, a fictionalized version of the nineteenth-century Scottish writer George MacDonald explains:

There are only two kinds of people in the end: those who say to God, "Thy will be done," and those to whom God says, in the end, "Thy will be done." All that are in Hell, choose it. Without that self-choice there could be no Hell. No soul that seriously and constantly desires joy will ever miss it. Those who seek find. To those who knock it is opened (Lewis, 2001b, p. 75).

The Problem of Pain reiterates these core ideas. Those who are in hell are successful “rebels to the end” (Lewis, 2001c, p. 130). God has offered every chance of escape from destruction. Forgiveness offers a chance to undo the relational consequences of disobedience. Without repentance, however, ultimate alienation is inevitable. Misery follows in course. Connection to God is the only possible source for ultimate fulfillment. Lewis writes, “To be God—to be like God and to share His goodness in creaturely response—to be miserable—these are the only three alternatives” (Lewis, 2001c, p. 47). Lewis’s defense of hell as a natural consequence of ultimate rebellion accords with many contemporary defenses. The traditional Christian defense of hell focused on retributive justice: since any sin against God is infinitely serious, the punishment for sin must also be infinite (Walls, 1999, p. 494). Contemporary defenses such as those offered by Lewis, Jerry Walls, and Richard Swinburne do not deny that sin merits punishment, but they do not defend the idea of eternal hell along those lines. Instead, they support key aspects of the traditional doctrine of hell by an appeal to natural consequences. Damnation is self-imposed, and may even be reversible. Though Lewis never makes this explicit in his work, it is implicit in The Problem of Pain, as well as in The Great Divorce and The Last Battle, that separation from God is not decisively decided at death, as we see various imaginative instances whereby God offers additional chances for humans to repent after death. Jerry Walls takes up Lewis’s insistence that the gates of hell are “locked on the inside” (Lewis, 2001c, p. 130; emphasis omitted) by arguing that God’s character implies that God offers postmortem “optimal grace” to everyone (Walls, 1999, p. 496). Walls writes, “if sinners are forever lost, either because of annihilation or their persistent refusal of grace, their fate depends on their continuing choices rather than only upon choices they made in this life” (Walls, 1999, p. 499). It is free rebellion, even to the utter end, that separates the damned from eternal happiness. 218 PHILIP TALLON

3. Divine Goodness and Suffering

Another major line of Lewis’s argument focuses on the relationship between God’s goodness and human suffering. As Lewis indicated before, our use of the word “good” in formulating the logical problem of evil must be shown to be “equivocal,” that is, ambiguous (Lewis, 2001c, p. 16). If God’s goodness is in some sense completely beyond our normal use of the word good, how are we to hold on to any content in the term whatsoever? The dilemma the Christian theist faces is to maintain both that (a) “if God is wiser than we His judgement must differ from ours on many things, and not least on good and evil,” and also (b) “if God’s moral judgement differs from ours so that our ‘black’ may be His ‘white,’ we can mean nothing by calling Him good” (Lewis, 2001c, p. 28). The difficulty, then, is to speak meaningfully about perfect goodness while acknowledging our own finite and flawed understanding of God’s moral nature. Total discontinuity, as has sometimes been affirmed in the Christian tradition, would resolve the logical problem, but at great cost. It would cut the Gordian knot in a very unhelpful way. If, as Lewis writes, “our idea of good is worth simply nothing—[it] may . . . turn Christianity into a form of devil-worship” (Lewis, 2001c, p. 29). How then are we to acknowledge that God’s reasons may surpass our understanding, without sacrificing the content of affirmations about God’s goodness? Here Lewis again anticipates an important issue in the problem-of-evil literature. A few decades ago, William Rowe proposed a modified form of the problem of evil known as the evidential problem. His strategy is to reformulate the challenge around so-called “gratuitous evils” that do not seem to lead to any greater goods. Rowe’s examples of a little girl’s rape and murder and the slow death by fire of a small deer in the woods offer instances where there appear to be no good reasons for allowing those evils to occur. Stephen Wykstra, in responding to the evidential problem of evil, replies by questioning whether the lack of appearance of any compelling reasons means that there are no such reasons. If God’s goodness far surpasses our own, we should not be surprised that there often appear not be compelling justifying reasons. Wykstra proposes what he calls the “condition of reasonable epistemic access” (CORNEA) as a way of showing that if we have reasons to think we do not have access to God’s good reasons, we should not be surprised not to perceive them (Wykstra, 2001, pp. 128-130). This introduced a response known as “skeptical theism”—the claim that, given human limitations, we should not expect to be able to understand all the reasons why a good God might allow evil. Wykstra’s defense is a powerful response to Rowe’s objections, but it risks the kinds of problems Lewis is concerned about in discussing God’s goodness. If God’s goodness is so radically different from human ideas of goodness that we cannot see why He might allow terrible evils, what grounds Pro: The Problem of Pain Defended 219 do we have for thinking that He does have good reasons? Richard Swinburne has suggested that Wykstra’s skeptical maneuver is too simplistic, and may fall prey to Lewis’s worry about introducing radical skepticism. Responding to the crux of Wykstra’s argument—the assumption that God’s ways are so much higher than our ways that we should not expect to understand why he allows all forms of evil—Swinburne writes:

But the trouble with this version of the argument is that while our moral beliefs (and factual beliefs, we may add) may indeed be in error in relevant respects, we need some further argument to show that they are more likely to be biased in the direction of failing to understand that some apparent bad states really serve greater goods, rather than in the direction of failing to understand why some apparent good states really serve greater bad states (Swinburne, 1998, p. 25).

This, too, is a version of Lewis’s “devil worship” objection. Lewis says little else that bears specifically on the evidential problem of evil. It should be noted, however, that his discussion is compatible with later work done by William Alston and others, who offer responses to the problem that acknowledge the limited nature of our ability to know all reasons why a good God might allow terrible evils. Alston acknowledges human cognitive limitations and suggests that the fact that we can think of many broadly compelling reasons why God might allow ordinary evils gives us good reason to think that God has sufficient reasons for allowing extreme and seemingly “pointless” evils as well (Alston, 2001). In like manner, a child who has known her parent to be trustworthy on things she can understand might properly trust her parents on matters she cannot understand. The nearest Lewis comes to threading the needle on this issue of how far our ideas of God’s goodness might be off-base is when he makes it clear he would oppose any strong form of skeptical theism while also holding on to an appropriate agnosticism about all God’s reasons. He writes:

The Divine “goodness” differs from ours, but it is not sheerly different: it differs from ours not as white from black but as a perfect circle from a child’s first attempt to draw a wheel. But when the child has learned to draw, it will know that the circle it then makes is what it was trying to make from the very beginning (Lewis, 2001c, p. 30).

This affirmation of some natural sense of God’s goodness is borne out throughout Lewis’s theodicy, as he offers various human analogies for God’s character and the reasons God might allow evil. To be sure, our human understanding of God is always in need of correction. Because of human wickedness and inherent limits of human understanding, we will not easily understand the reasons God has for allowing evil, nor indeed may we ever 220 PHILIP TALLON clearly understand all of them. We are, after all, a “spoiled species” (Lewis, 2001c, p. 81), and we can only see “as in a mirror dimly” (I Cor. 13:12) what God may have in store.

4. The Fall

In his discussion of original sin in The Problem of Pain, Lewis offers a bare- bones doctrinal formulation of the fall, claiming that “man is now a horror to God and to himself and a creature ill-adapted to the universe not because God made him so but because he has made himself so by the abuse of his free will” (Lewis, 2001c, p. 63). Beyond this conventional theological account, Lewis says little, and instead devotes a fair amount of energy to resisting additional doctrinal accretions. A number of times Lewis opposes a version of total depravity which would hold that “our idea of good is worth simply nothing” and that every part of the human moral condition is completely wicked (Lewis, 2001c, pp. 29, 61). As far as I can tell, almost no reputable theologians have explicitly held this extreme view of the fall (the sixteenth- century Lutheran theologian, Matthias Flacius, is one exception). There is also no indication anywhere that Lewis believes that humans inherit guilt. His view, rather, is that humans are naturally self-centered creatures that are born into a society of sinners in which they naturally form untrue moral beliefs and sinful attitudes. “The situation is not nearly so hard to understand as some people make out,” Lewis writes. “It arises among human beings whenever a very badly brought up boy is introduced into a decent family” (Lewis, 2001c, p. 81). Encountering God’s goodness, we are like the badly brought up boy who must radically revise his old ways of being and understanding, but the reversal is never “total” (Lewis, 2001c, pp. 29-30). In the memorable conclusion to the chapter, Lewis offers a possible “Socratic myth,” which imagines the heights from which we have fallen by showing how the Original Man might have been almost godlike in his control and mastery of his own nature (Lewis, 2001c, p. 71). Because of the fall, humans now live in a corrupt society, and are much more at the mercy of their inordinate desires. It might be easy to question whether Lewis, at this point, is a bit softer on depravity than biblical theology might allow, but his discussion of the universal experience of sin makes it clear that he sees sin as a deep and pervasive human condition.

5. Suffering and Soul Making

This brings us back to the problem Lewis sees with our culturally shaped conception of God’s goodness, namely that it consists in nothing more than “kindness” or a kind of “senile benevolence” (Lewis, 2001c, p. 31). Against Pro: The Problem of Pain Defended 221 the widely shared notion that God’s goodness equals kindness (which would seem to rule out the allowance of terrible and seemingly gratuitous evils), Lewis attempts to maintain the tension between God’s higher goodness and our partial understanding of it by exploring the concept of God’s goodness through various images of divine love. Love, in the short run, often looks nothing like kindness (at least in the common usage of the word). God’s love for us aims at our good, that is, our perfection and flourishing. Given our native selfishness and love of pleasure and ease, this requires growth. This process of growth is often painful as God makes fierce demands on us to become the sort of person He intends us to be. Lewis writes, “[T]he Love that made the worlds [is] . . . persistent as the artist’s love for his work and despotic as a man’s love for a dog, provident and venerable as a father’s love for a child, jealous, inexorable, exacting as love between the sexes” (Lewis, 2001c, p. 39). As Lewis sees it, humans are like a great work of art which the master artist will not leave be until it finally achieves perfection. God will take “endless trouble” over us, his terrestrial masterpiece, but He will also give us “endless trouble” until we are just right (Lewis, 2001c, p. 34). We are like a smelly, incontinent dog God takes to obedience school until we are fully housebroken (Lewis, 2001c, p. 36). We are like a son who must become the person a father wants him to be (Lewis, 2001c, p. 37). We are like an adulterous wife whose affections must be recaptured by her rightly jealous husband (Lewis, 2001c, p. 38). All of these analogies suggest God’s love for us, but also point to the difficult process of transformation required for God’s loving purposes to be fulfilled. Because of sin, of course, the process of correcting and perfecting our characters is understandably painful. This brings us to the final great subject that gives The Problem of Pain its title. As we saw, some pains should be expected in our lives simply because of the necessity of an orderly and stable physical world. Minor pains (what Lewis calls A-level pains) are simply physical sensations that result from living a vulnerable, embodied life in an environment governed by consistent and predictable laws of nature (Lewis, 2001c, p. 87). These pains do not raise any serious problems for God’s goodness, and may not even be disliked. More severe forms of conscious physical or mental suffering, what Lewis calls B- level pains, require a deeper analysis. Lewis’s response to the problem of B-level pains begins by his clear recognition of the highest good for any human: “to surrender itself to its Creator—to enact intellectually, volitionally, and emotionally, that relationship which is given in the mere fact of its being a creature” (Lewis, 2001c, p. 88). Perfected relationship—not a life of uninterrupted pleasure—is the highest good, and this is the ultimate reason why God allows pain, including even extreme B-level physical and mental suffering. In his final chapter of The Problem of Pain, Lewis gives emotional weight to this heavenly glory. The import of the final chapter is largely to give credence to 222 PHILIP TALLON his affirmation of communion with God as the highest value. A key focus for Lewis is affirming that heaven is not a one-size-fits-all experience, but one in which each person’s unique character will be uniquely satisfied. “Your soul has a curious shape,” Lewis writes, “because it is a hollow made to fit a particular swelling in the infinite contours of the Divine substance, or a key to unlock one of the doors in the house with many mansions” (Lewis, 2001c, p. 152). To be in the presence of God does not diminish, but completes, our one- of-a-kind selfhood. But the beatific blessing of heaven is not something that can be granted without personal cost. Unless we learn to truly love God and others, heaven can offer nothing for us. Our union with God is “almost by definition, a continual self-abandonment—an opening, an unveiling, a surrender, of itself” (Lewis, 2001c, p. 156). Thus, the self’s greatest satisfaction depends on the continued surrender of itself. Given the egocentric creatures that we are, this implies a great deal of painful moral transformation. Lewis here anticipates, explicitly this time, later development of the so- called “soul making” theodicy (Lewis, 2001c, p. 109). John Hick, in his classic Evil and the God of Love, uses the same phrase (borrowed from John Keats) to plumb the deeper mysteries of why God permits evil. Those who offer the standard version of the argument from evil, Hick writes, “are confusing what heaven ought to be, as an environment for perfected finite beings, with what this world ought to be, as an environment for beings who are in the process of becoming perfected” (Hick, 1978, pp. 257-258). God’s plan for us is modeled after the pattern of Christ, a man of much suffering. The love of the Father for the Son is echoed in the analogy of human parenting, inasmuch as responsible parents must impart to their children that worldly pleasure is not the “sole or supreme value” (Hick, 1978, p. 258). The long-run goal of a parent is to help their children grow into happy, mature, and ethical persons, even if some suffering is required in the process. The evil of suffering, beyond being a consequence for bad choices, may be instrumental in reuniting us with God’s highest good for us, Himself. Lewis discusses three ways that suffering can serve in this way. First, because of the fallen state of humans, pain “shatters the illusion that all is well” (Lewis, 2001c, p. 94). Without some suffering, a sinner may never recognize and own the “illusion” of his rebellious course, and therefore never repent (Lewis, 2001c, p. 93). In a famous passage Lewis describes pain as God’s “megaphone to rouse a deaf world” (Lewis, 2001c, p. 91). Despite the somewhat callous way that Lewis’s pronouncement is sometimes depicted, he seems fully aware of the danger in the image:

No doubt Pain as God’s megaphone is a terrible instrument; it may lead to final and unrepented rebellion. But it gives the only opportunity the bad man can have for amendment. It removes the veil; it plants the flag of truth within the fortress of a rebel soul (Lewis, 2001c, pp. 93-94).

Pro: The Problem of Pain Defended 223

Second, in addition to being a sharp call to repentance, pain reminds us that we are utterly dependent on God. Lewis quotes a friend who said, “We regard God as an airman regards his parachute; it’s there for emergencies but he hopes he’ll never have to use it” (Lewis, 2001c, p. 94). We proceed in life, very often, without a deep sense of our reliance on God. In moments of suffering, however, we are reminded “in advance of an insufficiency that one day… [we] will have to discover” (Lewis, 2001c, p. 95). Finally, pain allows us to distinguish between obedience for self-centered motives and obedience for God-centered motives. Lewis agrees with Aristotle that virtue unites pleasures with duties, and so the more Godlike we become, the more we will enjoy godly activities. But for fallen creatures, still in the process of perfection, he sees the Kantian wisdom in a certain necessity for pain in making correct moral choices. His focus is primarily epistemological: pain helps us to know that we are acting rightly. “We cannot therefore know that we are acting at all, or primarily, for God’s sake,” Lewis writes, “unless the material of the action is contrary to our inclinations, or (in other words) painful” (Lewis, 2001c, p. 97). Pain thus gives teeth to our obedience to God by clarifying our true motives:

[T]his one right act includes all other righteousness . . . the supreme cancelling of Adam’s fall, the movement “full speed astern” by which we retrace our long journey from Paradise, the untying of the old, hard knot, must be when the creature, with no desire to aid it, stripped naked to the bare willing of obedience, embraces what is contrary to its nature, and does that for which only one motive is possible (Lewis 2001c, p. 100).

Here the issues of suffering and free will intersect, as Lewis ties a bow on the twin themes that govern his solution to the problem of evil. It is through active self-giving that we become “collaborators” in God’s redemption of the world (Lewis 2001c, p. 102). Our self-giving helps break the “spell” that Adam cast over the human race (Lewis 2001c, p. 102). Christ’s sacrifice is the ultimate example of this, as He freely chooses to obey the Father’s will in the garden and undertakes the suffering of martyrdom to redeem the broken world. In short, by showing the reasons God may have for causing or allowing pain, Lewis shows why the assumption that a good God would always make humans happy is far too simple. Experience suggests that suffering can often help guide us toward higher goods. As Lewis writes,

I have seen the last illness produce treasures of fortitude and meekness from the most unpromising subjects. I see in loved and revered historical figures, such as Johnson and Cowper, traits which might scarcely have been tolerable if the men had been happier. If the world is indeed a “vale 224 PHILIP TALLON

of soul-making,” it seems on the whole to be doing its work (Lewis, 2001c, p. 109).

6. Animal Suffering

One final issue, animal pain, remains to be discussed in Lewis’s work. Though it is disconnected in some ways from Lewis’s other concerns, Lewis is obviously concerned about the reality of animal suffering on both an emotional and intellectual level. The sheer amount of animal suffering is both “appalling” and also difficult to comprehend because the “Christian explanation” for the problem offers little help in understanding it (Lewis, 2001c, pp. 132-133). One of the reasons animal pain is such a mystery is that the reasons God permits it are largely “outside the range of our knowledge” (Lewis, 2001c, p. 133). Not being given much revelation or insight on the matter, we must resort to “guesswork” (Lewis, 2001c, p. 133). However, even here Lewis’s response is thoughtful and accords well with later reflections on the issue, especially work done by Michael Murray in his important book, Nature Red in Tooth and Claw: Theism and the Problem of Animal Suffering (2009). Lewis offers three points that, taken together, make good mileage toward an adequate response to the problem of animal pain. First, Lewis recognizes that many animals, though they clearly respond to negative stimuli, may not be truly suffering (that is, experiencing pain in the B sense) because they lack integrated consciousness, or a “self” (Lewis, 2001c, p. 136). Some animals may have a level of sentience similar to what Michael Murray describes as “blind pain” (analogous to the experience of “blind sight,” when a person can physically “see,” but cannot mentally process the visual stimuli) (Murray, 2009, p. 54). Some animals that lack higher-order consciousness may experience all the neural components that go into pain, but never process the experience mentally, and thus do not meaningfully suffer. However, Lewis is quick to acknowledge that this defense may not cover many of the higher mammals, such as dogs or chimpanzees. “How far up the scale such unconscious sentience may extend,” Lewis writes, “I will not even guess. It is certainly difficult to suppose that the apes, the elephant, and the higher domestic animals, have not, in some degree, a self or soul” (Lewis, 2001c, p. 136). Second, Lewis considers whether moral evil may be partly to blame for animal pain. Recognizing that evolutionary theory shows that predation predates the emergence of humans, the fall of humans cannot be a total explanation for animal suffering (Lewis, 2001c, p. 37). Lewis speculates that God may have originally intended for humans to redeem animals from suffering, but that the fall altered this plan. This, however, is highly speculative and is offered as mere conjecture (Lewis, 2001c, p. 139). A Pro: The Problem of Pain Defended 225 slightly more promising theodicy for animal suffering is the idea that Satan is partly responsible for animal pain. Just as Scripture suggests that Satan, as “the god of this world” (2 Cor. 4:4), is the cause of some diseases (Luke 13:16), so might he be responsible for some aspects of animal suffering. (For a literary presentation of this theme that may have influenced Lewis, see J. R. R. Tolkien’s Silmarillion (Tolkien, 1977, p. 36)). These suggestions should be understood merely as possible explanations. We may not have strong reasons for affirming them, but they are entirely compatible with Christian theology. Finally, and most movingly, Lewis considers the plight of animals as a question of justice. Even if we can see possible reasons why innocent animals might suffer, what can God do to make it up them? Lewis finds some wisdom in those who have longed for animal immortality (including John Wesley): “If . . . the strong conviction which we have of a real, though doubtless rudimentary, selfhood in the higher animals, and specially in those we tame, is not an illusion, their destiny demands a somewhat deeper consideration” (Lewis, 2001c, p. 142). Lewis seems to think that animal resurrection, if it does occur, will likely depend on an animal’s connection to a human master, because he believes that true animal selfhood might emerge only as a result of a nurturing human- animal relationship (Lewis, 2001c, p. 145). This final qualifier does introduce a question about whether higher mammals (such as primates) that possess rich emotional lives but die without masters suffer and die without the possibility of ultimate justice. If innocent animals do deserve recompense, then perhaps their fate should not be dependent on relationship to a human master (Bassham, 2005, pp. 278-285). Might God Himself, as He indicates in the Book of Job (38:39-41; 39:1-30), be the master and shepherd of all? That Lewis leaves this—and a number of other important issues— incompletely explored is hardly a major fault of his short and accessible work. He effectively rebuts the traditional logical problem of evil; God’s goodness and power do not logically contradict the unhappiness we see all around us. Lewis achieves his primary aim and lays the groundwork for later thinkers to add greater richness and depth to Christian theodicy. On the whole, Lewis deals with the troubling questions raised by suffering with such probity and rigor that The Problem of Pain should be regarded as an astounding accomplishment. I challenge anyone to produce another work in the 20th century that deals with the problem of evil with greater scope, sensitivity, and seriousness.

Eighteen

CON: C. S. LEWIS ON EVIL: AT BEST A LIKELY STORY

David L. O’Hara

I have to say that I like Phil Tallon’s chapter, and, more importantly, that I think he is generally right. When I say “I like it,” I mean that I think it is well- written and persuasive, and when I say “I think he is right,” I mean that Tallon has given an accurate account of Lewis’s book. Moreover, I think Lewis is an important apologist for a certain kind of Christian position, and that he represents that position well. Unfortunately, all this makes my task here more difficult, since I intend to argue against that position. I am not off to a good start. Trying to refute as sharp a thinker as C. S. Lewis is a daunting task from which any wise person would recoil. Lewis is a master, and his prose is full of gems of thought, evidence of a powerful mind allied with a profound faith. With each of his books I find real insights, often in casual observations about medieval philosophers and poets. “We think we are kind when we are only happy,” Lewis writes, and I find my own heart laid bare (Lewis, 1996, p. 49). Like a twentieth-century Justin Martyr, Lewis learned wherever he read. He draws insight from Bernardus Silvestris in several places in his writings, and I have to confess that were it not for Lewis, I would never have read Bernardus. Similarly, Lewis’s casual observations about Aristotle, or about medieval angelology, may seem like offhand remarks but turn out to be life-changing ideas. Let me make my job even harder: Lewis’s position has strong support in the broad Christian tradition. He argues very much as St. Augustine argues in his Confessions, for instance. Augustine repeats the scriptural idea that God “fashions pain to be a lesson” (Augustine, 1998, p. 25), an idea Lewis famously echoes in The Problem of Pain, where he calls pain “God’s megaphone to rouse a deaf world (Lewis, 1996, p. 91). Nevertheless, in what follows, I hope to show that while Lewis’s argument may be persuasive and even orthodox, it is nevertheless inadequate and of very limited value. Our response to the problem of pain should not be primarily one of trying to construct a system of theology in which God must make sense. Instead it should be a response of awe towards God and the life we have been given, and of love and compassion toward our neighbors. It is probably the rare exception when those who suffer really need, more than anything else, to know what God is up to. It is not metaphysics or theodicy 228 DAVID L. O’HARA that gives us reason to believe that our suffering is meaningful, that there is room for awe and wonder in the midst of our pain, and that it is possible to find anything more worth our attention and worship than grief and agony. It seems to me that this kind of argument, while well-meaning, can have the effect of making us less concerned with the problem of evil by making it into an intellectual puzzle instead of a call to prayer and action. Fortunately for me, Lewis himself offers much that is necessary for demonstrating the weaknesses of his own argument. I will try to lay those out as simply as I can, and then to sketch another possible solution to the problem posed by The Problem of Pain. Those of us who think that charity is as important in philosophy as logic will want to note that Tallon’s representation of Lewis’s position is based on The Problem of Pain alone. This matters for two reasons: First, my argument against that view, then, will not be an attack on Tallon or Lewis, but on the view that Tallon so deftly extracts from Lewis’s book. Second, Lewis has more to say about evil and suffering than he lays out in that book. Arguably, Lewis’s best writing is in his fiction. As we will see, Lewis does not draw a sharp line between narrative and philosophical theology. His philosophical and theological writings often turn to narratives for their support, and his fiction is shot through with philosophical references and ideas. For example, consider the words of the Professor in The Lion, the Witch, and the Wardrobe: “Plato, it’s all in Plato!” (Lewis, 2004, p. 170). In his postscript to Out of the Silent Planet Lewis pretends to “remove the mask” of fiction and to allow the reader to peek behind the curtain to see the novelist at work. His reason for doing so, he says, is that fiction can serve a philosophical purpose. “What we need for the moment is not so much a body of belief as a body of people familiarized with certain ideas,” he writes (Lewis, 1965, p. 152). Sometimes belief is not as important as imagination, after all. So instead of trying to refute Lewis head-on, let me offer some ways in which Lewis himself offers a critique of his own position.

1. “I Am Not a Theologian”

First, as Lewis wisely volunteers several times in The Problem of Pain, he is not a theologian. There can be no doubt that his writings make real contributions to the popular understanding of theology and of philosophy of religion. Nevertheless, he often pointed out that he had no special training in those fields. In his Preface he writes, “If any real theologian reads these pages he will very easily see that they are the work of a layman and an amateur,” and that it “is not a work of erudition” (Lewis, 1996, p. xii). This may seem unimportant, especially in an age when most of us are content to imagine that theologians do not know much that is worth knowing, and that we are all competent theologians on our own. Lewis was undoubtedly Con: C. S. Lewis on Evil: At Best a Likely Story 229 well-read in theology, but he knew his own limitations. His caveats about his scholarship are not just a matter of Christian humility, but of scholarly integrity. He was reluctant to write The Problem of Pain, and he takes care several times throughout the book to point out the limits of his studies and of the usefulness of his argument. From the outset he insists that in most of the book he is merely “re-stating ancient and orthodox doctrines” (Lewis, 1996, p. xii). It would be both inconsistent and unwise to take Lewis as an authority on philosophical theology but not as an authority on the limitations of his own studies. Of course, one need not be a professional or trained theologian to speak theological truth. St. Francis of Assisi was, by all accounts, not a learned man. His writings are brief, and not at all erudite. He appears to have struggled to write well, and he employed the services of at least four secretaries throughout his life to help him get his ideas written down. Yet nobody disputes the profundity and value of his spiritual writings, nor denies that his Christian beliefs are both ancient and orthodox. At the same time, we take him to be one theological voice among many, and we recognize that no matter how powerful his mystical insight into poverty, chastity, and obedience, his is not the last word. If anything, Francis’s words point to a higher authority than Francis. In a similar way, Lewis is attempting to help us solve a problem, but it is a very particular, local, and parochial problem. By reminding us that he is not a theologian, Lewis effectively urges us take his voice as also being just one among many. His is not the orthodox view. Orthodoxy, after all, offers us some latitude of belief. Lewis’s position is only one possible Christian view, and, as I will argue, it is not necessarily the best.

2. The Limited Value of The Problem of Pain

This brings me to my (or rather, Lewis’s) second point: Lewis himself points out that The Problem of Pain will be of limited usefulness, and that in fact he is addressing a relatively unimportant question. The Problem of Pain is not so much a solution to the problem of evil, but an exhortation to practice the virtue of courage, emboldened and empowered by the promises of the Gospel. Arguably, Lewis viewed the book as a prefatory work in character ethics. In the Preface, Lewis writes “that the only purpose of the book is to solve the intellectual problem raised by suffering; for the far higher task of teaching fortitude and patience I was never fool enough to suppose myself qualified” (Lewis, 1996, p. xii). He adds, “when pain is to be borne, a little courage helps more than much knowledge, a little human sympathy more than much courage, and the least tincture of the love of God more than all” (Lewis, 1996, p. xii). If he could not give relief to sufferers, he could at least clear the intellectual ground so that they might find it easier to begin to face suffering bravely. He is working at the level of “knowledge,” knowing that courage, 230 DAVID L. O’HARA sympathy, and divine love are all more excellent. He is not trying to supply knowledge but to overcome the false belief that an understanding of pain is what we really need. At first glance, this claim may seem surprising. The topic of pain and suffering is, after all, important to everyone at some point. Everyone suffers; no one escapes pain. For the theist, this can be no small matter. Besides, surely pain is one of the strongest arguments against the existence of God: if God is all-loving He should wish the end of pain; if He is all-knowing He cannot be ignorant of our suffering; and if He is all-powerful, then He stands in a position to bring about His own wishes. Yet apparently He does not. Does this not argue against the existence of God? But notice what we have just done. Plainly, the problem of pain is not one problem, but several. First, it is the universal problem that all of us must confront pain and suffering; that we generally wish to be free of it; and that we need to figure out how to conduct ourselves when we suffer. Second, pain presents us with a problem in apologetics. This second problem is itself not one but several problems as well, which we may compare to St. Anselm’s “ontological argument” in his Proslogion. Anselm attempts to demonstrate that the existence of God is self-evident, and that once the definition of God is known we may be sure of the existence of the God that is defined. One of Anselm’s aims in his writing is to help the insipiens (Latin for “the one who does not know”) to know that God exists. But who, exactly, is this insipiens? Some of Anselm’s readers took the insipiens to be the non-believer, and Anselm’s reply to his critic Gaunilo suggests that Anselm would agree. But Anselm clearly has another person in mind in writing his apology: himself. He begins his argument with a prayer, after all. Plainly he begins already believing. Even so, his apologetic plays an important role for himself, in showing that there is at least one way in which he can demonstrate that faith is not irrational. In simpler terms, Anselm is writing from the position of a person who has faith but who seeks understanding of that faith. So his Proslogion has two aims: to help the unbeliever to overcome an intellectual hurdle to faith, and to help his own intellect to be reconciled to what his heart knows. This, I believe, is Lewis’s position as well. While there may be some benefit to persuading non-believers to believe, and in defending the faith from skeptical attacks, there is also considerable value in going through an intellectual exercise that allows us to see that our faith is not unreasonable. The Problem of Pain may be helpful for the individual who wishes to see that it is not impious to inquire into possible reasons why God allows suffering. But there is a pitfall to avoid here. The temptation will be to allow the argument to seem so strong that it comes to supplant the faith, forgetting that the faith came first and is of more value. The reasoning may seem sturdier, but in the end it is only as strong as its premises, and its premises may be doubted. In other words, the argument itself rests on faith. The only real value Con: C. S. Lewis on Evil: At Best a Likely Story 231 of the argument is to remove an obstacle to faith, so that faith may get on with its real work. The danger here may not be apparent, so let me turn to the pragmatic philosophy of Charles Peirce to show why this matters. Peirce points out an uncomfortable truth about our species: we like to say lofty things about our love of the truth, or about being seekers of truth, but in fact we rarely press our inquiries very far. Usually, we stop asking questions when we no longer feel the irritation of our doubts. If we discover that our previous knowledge of a subject is inadequate, we seek new answers. But, Peirce points out, we do not tend to seek true answers; we just seek satisfying answers. We seek adequate solutions to our problems, and then we stop asking questions because we are comfortable with what we know. In itself this is not a problem; there is nothing wrong with seeking comfort. However, being in possession of comfortable answers tends to make us stop examining our beliefs (Peirce, 1955, pp. 9-14). For the theist, believing that we have solved the problem of pain can become an obstacle to both work and worship. If we believe that we are in possession of sufficient understanding of why God ordains suffering, this can impede our will to work to alleviate that suffering. And if we believe we have a good enough understanding of why God permits suffering, this can be an impediment to prayers of supplication and adoration. Here we return to Lewis’s admonition not to take him as a theologian; he does not intend his argument to supplant either sound theology or faith lived in love. In other words, one may live a faithful life without understanding why God allows suffering; but if one understands Lewis’s argument about pain, that alone does not make for a good life, and it is a poor substitute for charity.

3. The Limits of What We Can Know

But surely we should not reject Lewis’s argument just because we fear that accepting it might lead us to complacency. The threat of complacency should be a warning to us, not a reason to deny the validity of a claim. We should only reject the argument if we see that it is not justified by its premises or by its chain of reasoning. As it turns out, Lewis himself points out that this is the case with his own argument. The weakness of the argument, ultimately, is this: it depends on speculation. Lewis is writing about what no one really knows. The danger is that the reader of The Problem of Pain will come away from it with the mistaken view that now she does in fact know what she does not know. I want to note once again that I am not saying Lewis is wrong. What I am saying is that Lewis is aware that what he says about the intellectual problem of pain must necessarily run up against the very limitations of our intellect. This is an important point, one that Lewis notes several times, and yet (at least when he was writing The Problem of Pain) he does not seem to take his own point seriously enough. In addressing one serious question, he writes, “I 232 DAVID L. O’HARA doubt whether the question has any meaning: and even if it has, I am sure that the answer cannot be attained by the sort of value-judgments which men cannot significantly make” (Lewis, 1996, p. 64). Why does he allow for limitations of our knowledge here, but not in other places? Here we learn more from Lewis’s example as a thinker than from his conclusions. If we are generous readers, we may say that he is acting Socratically, attempting to follow the argument wherever it leads. Then, when he finds himself in aporia, unable to see the way forward, he acknowledges that. It should surprise no one to hear it said that there are things about God that we simply do not know. Even so, some people may be concerned that in saying so we are giving up without a struggle. But why should we not allow God’s reasons to exceed the length of our intellectual reach? We know God to be good, but we also know He is not a tame lion. Here again, Lewis’s example is better than his conclusions. In chapter five of The Problem of Pain, Lewis approaches the limits of his knowledge in a way that I argue is more ancient and more orthodox (and probably more Biblical, too) than the attempt to justify God in the face of suffering: he tells a story. Recall Lewis’s words in Mere Christianity, in the first of his two chapters entitled “Faith.” Faith, he says, “is the art of holding on to things your reason has once accepted” (Lewis, 2002b, p. 116). He distinguishes between faith in this sense, and mere belief on the other hand, which is a kind of intellectual acceptance of a fact. Faith is not simply intellectual acquiescence to propositions, and it may thrive even in the absence of certainty (Wettstein, 2012, p. 147n. 30). Here in this central chapter of The Problem of Pain, the hinge on which his book swings is a myth that he tells about the fall of humanity. “What exactly happened when Man fell, we do not know; but if it is legitimate to guess, I offer the following picture—a ‘myth’ in the Socratic sense, a not unlikely tale” (Lewis, 1996, p. 71). In a footnote he adds that by “myth” he means “an account of what may have been the historical fact” (Lewis, 1996, p. 71n. 3). Let us grant him every literary license in writing this myth, just as we afford that license to Plato. But in granting it, let us also make sure to remember that it is just that—a possible story, but not necessarily historical fact. It is a speculation, a useful fiction. And this is exactly what we expect from Lewis. Just as he said in the title of one of his famous essays, “sometimes fairy stories may say best what’s to be said” (Lewis, 1982, p. 45). Lewis’s gift as a philosopher is like Plato’s: both of them stir us to reconsider our casual words; both of them show us our common errors; and both of them, at their best, leave us with stories more than with doctrines. Those who are inclined to read either one only for his conclusions, beware!

Con: C. S. Lewis on Evil: At Best a Likely Story 233

4. The Problem of Pain and Pastoral Theology

This leads me to my final and most important criticism of The Problem of Pain: the argument easily lends itself to bad pastoral theology. That may seem unrelated to Christian apologetics, but if our apologetics oppose our pastoral theology, then I submit that something has gone wrong with our apologetics. Lewis’s Problem of Pain is a powerful work, but Lewis himself appears not to have been satisfied with it. He published The Problem of Pain, with its bold and memorable proclamations (“God whispers to us in our pleasures, speaks in our conscience, but shouts in our pains: it is His megaphone to rouse a deaf world”) in 1940. Twenty-one years later, he wrote far more profoundly about pain—his own pain, after the death of his beloved wife—in A Grief Observed. There can be no doubt that the profundity of the later book has much to do with the fact that it is the story of his own suffering. It is, as the title suggests, a phenomenology of grief. This means he does not need to force the grief to mean one thing or another; he sets himself free to observe it, and to acknowledge it. Near the end of the book he tells us he will stop writing when he runs out of notebooks to write in, not when he has resolved the question. That small detail of his life, that resolution to attend to his own existence in all its contingency, speaks a louder word than all of The Problem of Pain and its megaphones. Perhaps the most important point to make here is that grief is always personal. This brings me back to my previous point: if our apologetics insists on treating pain impersonally, it will wind up being offensive and repulsive to persons. If we take The Problem of Pain as a work intended to persuade suffering people that their pain is no obstacle to belief, we may only add to their pain. At this point I have to make an observation—one which I hope will not be unkind—about both Lewis and Tallon. Both of them suggest that God cannot eliminate all of our suffering because this would somehow violate the “rules of the game.” It is precisely this kind of talk that concerns me. How can we take the suffering of others seriously if we regard their lives as ludic frivolity? But this seems to be important for Lewis’s account of pain, as though God were bound to “play” by some rules. Are we pieces on a divine gaming board? To insist that we are stands in stark contrast to words like, “I have called you servants, but now I call you friends” (John 15:15). Perhaps someone will object that philosophy is supposed to be clear and firm, not swayed by love or matters of the heart. Philosophy, they will say, should cast a cold, steely gaze on life, and accept whatever consequences good reasoning leads to. It should approximate the methods of the sciences, analyzing with precision and without bias. Fair enough. But I would like to remind such people of three things: First, science itself is always—always—a matter of the heart. Its methods may not be, but its aims are. We select what we will analyze not by scientific means of selection but by following our inclinations and by having in mind the kind of knowledge we long to acquire. 234 DAVID L. O’HARA

Second, there is, as Berengar of Tours claimed when discussing the Eucharist, a science of the heart, a scientia cordis. The heart may aid us in our search for rational understanding with a reason of its own, given by God. Third, as the very name indicates, philo-sophy is and ought to be a matter of love, especially when discussing something as important as religion. Philosopher of religion Charles Taliaferro reminds us of this when he writes about The Cloud of Unknowing: “The writer [of the Cloud] recommends that whenever you feel lost in the search for God, you should repeat the words “love” or “God” a lot; both words amount to the same thing” (Taliaferro, 2006, p. 175). We are not robots or logical machines. We are people in relationship with others, people enmeshed in communities, and it is love above all else that preserves those communities. Reason helps us to accomplish the ends of the heart, but as Pascal says, the heart has reasons that reason itself cannot comprehend. Charles Peirce reminds us that “the heart is a perceptive organ,” to which he adds what Augustine and Anselm and Pascal and Lewis all knew: it is the heart that perceives God, and the mind only races behind the heart, trying to catch up (Peirce, 1955, pp. 377-378).

5. A Better Story

Like Socrates at his own trial, I feel the need to ask the jury to stick with me, aware that I am probably saying things that seem unappealing to those who love Lewis. So let me repeat: I do not think Lewis is wrong. I do think, however, that he is also not right, if we regard his Problem of Pain as his final and best (or worse, as our final and best) word about God and suffering. He offers a plausible story, but it is just that, a story. Lewis reminds us in chapter 6 of The Problem of Pain that “we must be careful to attend to what we know and not to what we imagine” (Lewis, 1996, p. 107). This is good advice, and I will add to it that Lewis’s “response” to the problem of evil is what he can imagine might be the case, and not at all what he knows for certain. This is part of the work of theology, and of philosophical theology, to weigh competing stories and to ask how they may help us navigate the situations that face us concretely, here and now. Richard Rorty, not usually thought of as a philosopher of religion, nevertheless makes a very helpful observation about this. Too often we have allowed apologetics to be simply the response of one institution to another in a competition for recognition and legitimacy. This, Rorty notes, distracts us from the real work of holiness. And to this he adds that holiness looks forward to a world in which “love is the only law” (Rorty, 2005, pp. 39-40). Why should not this be what guides our philosophical theology? Metaphysical speculation has its place as a kind of training-ground for the intellect, but surely Christian philosophers, at least, will want to devote their attention to the practical problems of living lives of love. Some theories, even “ancient and orthodox” ones, are distractions from this. Lewis says as Con: C. S. Lewis on Evil: At Best a Likely Story 235 much about several venerable theological theories: “These theories may have done some good in their day but they do no good to me” (Lewis, 1996, p. 82). In fact, maybe what we need most is not understanding, but courage, sympathy, and love, as hard as those things may be to understand. This is especially true when our “understanding” is not perfect knowledge but a guess based on amateur theology and speculative myth-making, as it is in The Problem of Pain. “The best is perhaps what we understand least,” Lewis writes in A Grief Observed (Lewis, 2002a, p. 688). Let Plato be right when he says that “the love of wisdom begins in wonder”; if so, then that is all the more reason not to try to dispel wonder and mystery from our lives. When faced with the problem of evil, I find the best I can do is turn to the stories, and to the Story. Here is where Lewis is at his best, when he is not trying to put the great question of pain to rest but instead showing how we might approach pain with the kind of courage that wrests meaning from the suffering. I often think of Puddleglum the Marsh-Wiggle in this regard. In The Silver Chair, Puddleglum and two children, Jill Pole and Eustace Scrubb, are imprisoned underground by a witch. They become enchanted by her magic fire, so that they are unable to remember that there is a world aboveground to which they belong. In a moment of courage, Puddleglum stamps out the fire with his bare foot, causing himself some pain. The pain—or perhaps the courageous act, or both—clears his head and break the spell. And then Puddleglum makes a speech that includes one of Lewis’s best cases for God. He describes the beauty of God’s world under the sun, acknowledging that there, underground, they cannot see it. Perhaps, he admits, that world does not exist at all. Even so, a life spent believing in that world and seeking it is better than a life that gives up hope of finding it. “That’s why . . . I’m on Aslan’s side even if there isn’t any Aslan to lead it,” Puddleglum concludes (Lewis, 1980, p. 159). This is a beautiful argument, and one that rests on the beauty of hope. And it is a pragmatic argument, one that is shot through with uncertainty and human fallibility. It calls to mind the words of Socrates in Plato’s Meno: “I do not insist that my argument is right in all other respects, but I would contend at all costs in both word and deed as far as I could that we will be better men, braver and less idle, if we believe that one must search for the things one does not know, rather than if we believe that it is not possible to find out what we do not know and that we must not look for it” (Plato, 2002, 86b-c). Sometimes we simply cannot see the way ahead, and no amount of abstract reasoning will light our way. In such times we should not pretend to have light as the younger Lewis did in writing about pain; instead, we should encourage one another that light may yet be found, and then seek it in the company of those who will join us, as we see in the example of the later Lewis’s writing on grief. What plans God has for evil I do not know. I hope, and I believe, that God will right the world’s wrongs. But given the reality and gravity of human suffering—that is, given the real pain endured by my neighbors near and far— 236 DAVID L. O’HARA

I am reluctant to try to do what God does not. God gives Job no explanation for Job’s suffering; Jesus himself cries out “why?” from the cross. In neither case does God reply with an argument. We philosophers who love God should hesitate before trying to speak where God Himself is silent. Lewis’s argument is interesting, and it can help our heads catch up to our hearts. It can be a good exercise in reasoning for those in the ivory tower, but we must note its many limitations, especially those to which Lewis calls our attention. It can, perhaps, even encourage the individual who reads it. But The Problem of Pain is not a final answer to the problem of pain, and if we take it as one, we run the risk of becoming another source of pain to those who seek the solace of divine love. As First Corinthians says, knowledge puffs up—and uncertain knowledge that is presented confidently puffs us up even more so. In the end, we should remember Lewis’s own words: “Christianity is not the conclusion of a philosophical debate on the origins of the universe: it is a catastrophic historical event following on the long spiritual preparation of humanity” (Lewis, 1996, p. 14). Lewis knew that The Problem of Pain and its abstract philosophical reasoning cannot save us from suffering. It tells a story, even a plausible story, but in the end it is no substitute for the story of Christ.

NINETEEN

REPLY TO DAVID L. O’HARA

Philip Tallon

I would like to begin by thanking Dave O’Hara for his genial and thoughtful challenge. It is more enjoyable to argue with him than to agree with many others. With opponents like O’Hara, who needs friends? O’Hara’s objections to Lewis’s theodicy are very clearly presented, but I will briefly summarize them before replying to each in turn. His major criticisms are as follows:

(1) Lewis is not (as he admits) a professional theologian, so we should not view him as a final authority on the problem of evil. (2) Lewis’s intellectual response to the problem of pain is of limited value for three reasons: (a) As Lewis admits, in the midst of suffering, a little courage, sympathy, and the love of God are of far more help than mere intellectual understanding. (b) The only real value of theodicy is to remove an obstacle to faith. (c) Theodicy can give us a false sense of intellectual satisfaction and can impede efforts to stop suffering. (3) Many key assumptions in Lewis’s theodicy are too speculative to provide a solid foundation for his argument. (4) Suffering is personal, but an intellectually-based theodicy is impersonal, and thus often repels those it seeks to comfort. The heart, rather than the head, is better equipped to perceive God when we are in the midst of pain.

Before responding to these objections, I must point out that O’Hara does not pinpoint any major flaw or fallacy in Lewis’s thinking. O’Hara says that he does not think Lewis is “wrong,” though he does say he is “not right.” As O’Hara acknowledges, Lewis’s position has “strong support in the broad Christian tradition.” This is important because many of the objections raised against Lewis could also be applied to influential recent theodicists like Alvin Plantinga or Richard Swinburne, or to weighty theologians of old such as Aquinas, Boethius, or Augustine. Thus, O’Hara’s issues are not always “Lewis-specific.” His problems with The Problem of Pain are mostly complaints about the way theologians have approached the problem of evil throughout the Christian centuries. 238 PHILIP TALLON

It is helpful for the consistency of my response that I disagree with all four of O’Hara’s objections, even if I think that all touch upon valuable truths. In my view, O’Hara’s main objections to Lewis’s work all appear to stem from a desire to see greater modesty from Lewis about the scope and value of his theodicy. This desire for intellectual (and spiritual) modesty is grounded in what O’Hara takes the proper Christian response to the problem of pain: “a response of awe toward God and the life we have been given, and of love and compassion toward our neighbors,” rather than “trying to construct a system of theology in which God must make sense.” I respect this gesture toward humility, but I believe that it unhelpfully limits one’s ability to speak faithfully about God (drawing on biblical resources) in the midst of evil. I also think that it undercuts the greatest benefit Lewis has to offer us in his theodicy: a deeper understanding of God’s love, seen through the tears of suffering and grief. I will return to this point again. First, however, let me turn to O’Hara’s specific objections.

1. Is It Important that Lewis is Not a Professional Theologian?

O’Hara rightly notes, as I also did in my essay, that Lewis prefaces his work with self-deprecatory remarks about his own credentials. He admits to being a layman in the Church of England and an amateur in theology. In addition, his work is presented to laypeople rather than to the academic community. In a deft turn of phrase, O’Hara writes that it “would be inconsistent and unwise to take Lewis as an authority on philosophical theology but not as an authority on the limitations of his own studies.” That is a fair point, and Lewis’s modesty helps us to understand both his context and his intention in writing. It should not, however, lead us to underrate his contributions to twentieth- century academic theodicy, much of which has not been produced by professional theologians. What Lewis’s comments should do is to focus our attention on the quality of his ideas. As I noted in my opening essay, many important figures in late-twentieth-century academic theodicy (for example, Alvin Plantinga, John Hick, and Richard Swinburne) develop and defend ideas similar to Lewis’s. This makes clear that The Problem of Pain was, and is, a very significant contribution to contemporary philosophical theology. O’Hara makes another interesting point that invites a response. He writes: “Lewis effectively urges us to take his voice as also being just one among many. His is not the orthodox view.” I agree, because with theodicy there really is no orthodox view. Reflection on the problem of evil falls in the cut- and-thrust of the philosophical and theological conversation, as Christian thinkers attempt to speak meaningfully about God’s goodness in the face of great suffering and pain. In fact, in theology as a whole, active discussion is occurring in almost every area of Christian belief. It is true that Lewis is no Reply to David L. O’Hara 239 more the final word on theodicy than Saint Augustine was, but that says little about the value of his work.

2. Is Lewis’s Theodicy of Limited Value?

I agree with Lewis and O’Hara that theodicy is of limited value, but only in the sense that all theology is of limited value when compared to larger, more ultimate concerns. It is of limited value for me to hold true beliefs about my wife’s favorite foods. It is far more valuable to combine this knowledge with action in order to cook dinner for her; then we have true beliefs and good actions. The same point applies to virtually any aspect of theology. “What availeth thee to dispute highly of the Trinity if thou lack meekness and thereby thou displeaseth the Trinity?” (Thomas à Kempis, 1943, p. 130). Biblical theology is of small value compared with its right application in practice. But that does not mean that theodicy (or theology in general) is of small value. Quite the opposite. A point Lewis was fond of making with regard to the natural loves—that the highest does not stand without the lowest—is relevant here. Lewis is right that courage, sympathy, and Divine love are of more value than knowledge is when we must endure suffering; that does not mean, however, that knowledge is unimportant; for in the final analysis, the moral does not stand without the intellectual. As Lewis himself writes: “I am not asking anyone to accept Christianity if his best reasoning tells him the weight of evidence is against it. That is not the point at which Faith comes in” (Lewis, 1996, p. 140). For Lewis, as O’Hara notes, faith is the “art of holding on to things your reason has once accepted” (Lewis, 1996, p. 140). Faith is thus more than an intellectual conviction, but it is not less. In this sense, theodicy can play a crucial role in forming the foundation for trusting obedience to God. As someone who once felt strongly the force of the logical problem of evil, it was necessary for me to intellectually resolve that problem with the help of religious thinkers like Lewis. But working through issues in theodicy did more than remove an intellectual obstacle; it helped me to see additional aspects of God’s character, and increase my trust in, and love of, God. I see no reason to bracket theodicy from the rest of one’s Christian’s life. O’Hara points out that too often we do not seek true answers, but merely satisfying ones. I agree. But I also think that many of the key ideas in Lewis’s theodicy are not merely satisfying, but true. They are grounded in scripture, tested by tradition, logically coherent, and fit with our experience. When Christians experience suffering, they often try to make sense of it based on everything they believe. One result is that Christians frequently accept very inadequate explanations of why God allows suffering. Many pastors I know have to help their parishioners to abandon non-Biblical views of God, such as the idea that God took a child’s life because the child was 240 PHILIP TALLON

“needed” in heaven. In this sense, a theodicy like Lewis’s can be helpful in dislodging semi-satisfying but unhelpful answers, and by spurring believers to seek deeper truths. I would like to hear more from O’Hara about why he thinks theodicy might be an impediment to “work and worship.” I agree that all sorts of bad theodicies could do so. (For instance, I think that Calvinist responses to the problem of evil will present such problems.) But I do not see why theodicy as such would do so. Lewis tries to show why a high view of God’s providence does not undermine a Christian’s efforts to live rightly: “you will certainly carry out God’s purpose, however you act, but it makes a difference to you whether you serve like Judas or like John” (Lewis, 1996, p. 99). A strong intellectual belief in God’s sovereignty does not mean that we must become passive. As Lewis remarks in Mere Christianity, “If you read history you will find that the Christians who did the most for the present world were just those who thought most of the next” (Lewis, 2001, p. 134).

3. Are Lewis’s Arguments Too Speculative?

O’Hara points to a number of places in The Problem of Pain where Lewis indulges in speculative storytelling or admits to the limitations of our intellect in understanding the ways of God. I will discuss some of those passages, but I want to begin by pointing out that the two central pieces of Lewis’s response to the problem of evil—his appeal to free will and to the value of soul- making—are well supported by biblical texts and by solid moral reasoning. The story of scripture centers round God’s ongoing relationship with humanity. God’s main conflict with humans involves his attempt to create a people to be his community on earth, a plan that is continually being thwarted by human sin. As such, the role of free will is not peripheral to the witness of scripture. The notion that God’s love for us requires our free cooperation is also a deep moral conviction shared by religious believers throughout the ages. That libertarian free will is necessary for moral responsibility informs our thinking in common life and in our legal system. Not everyone, of course, agrees with this, but it is a position with a rich tradition and much evidential support. Likewise, the idea that a loving God can use (or even send) some measure of suffering for our moral transformation is found throughout scripture (for instance, in Romans 5:3-5). The Christian theologian who suggests that God uses suffering for our growth is not speculating but is instead relying on scripture, tradition, and reflective experience. As such, I think that Lewis’s rebuttal of the logical problem of evil stands even if the speculative elements of The Problem of Pain were removed. With regard to these conjectural elements, Lewis is up front about when he departs from settled theological tradition and allows his imagination to run free. This points to his intellectual rigor, and it also allows him to explore Reply to David L. O’Hara 241 beyond what is clearly defensible from scripture. His discussions in The Problem of Pain of the fall of humanity, animal pain, and the nature of heaven are among my favorites in the book for their willingness to offer fruitful avenues of thought without over-committing. In this respect, Lewis is a good model for theologians who want to ground their work in scripture, but also wish explore the many ways that God’s goodness and wisdom may exceed our limited grasp.

4. Is Theodicy Cold Comfort for Sufferers?

O’Hara indicates that he sees his last criticism—the claim that “the argument easily lends itself to bad pastoral theology”—as the most serious. I take it that O’Hara sees a danger in “impersonal,” purely intellectual explanations of suffering, such as Lewis’s much-quoted claim that pain is God’s “megaphone to rouse a deaf world” (Lewis, 1996, p. 83). An example of a more personal treatment of pain is found A Grief Observed, where Lewis chronicles and tries to come to terms with his own terrible experience of loss. As I said in my opening salvo, I think A Grief Observed is a wonderful book that has brought consolation and healing to many. But it is a mistake to pit Lewis’s later work against his earlier writings. For one thing, he does not retract anything from The Problem of Pain in A Grief Observed or, for that matter, in any of his later writings. In A Grief Observed, Lewis does say that “probably half the questions we ask—half our theological and metaphysical problems . . . are [unanswerable]” (Lewis, 1961, p. 55). But, significantly, he does not say which problems are unanswerable. And Lewis had already acknowledged the existence of unsolvable problems in The Problem of Pain! It is also worth noting that Lewis engages in intellectual theorizing in A Grief Observed, as when he talks about the necessity of our having some sense of the divine nature in order to meaningfully talk about God’s goodness (Lewis, 1961, p. 28). This brings me to my last major point. O’Hara presses the case, in a very Pascalian way, that we perceive God’s goodness primarily by the heart, not by the rational intellect. Theodicy, as he acknowledges, can help to remove obstacles to belief, but it is often, at best, an attempt to lasso the head so it can get caught up with the heart. I think this is often true, but not quite in the way that O’Hara suggests. First, Lewis’s picture of the interaction between the head and the heart is very dynamic. Often, as he writes in The Problem of Pain, understanding precedes emotional response:

But God wills our good, and our good is to love Him (with that responsive love proper to creatures) and to love Him we must know Him; and if we know Him, we shall in fact fall on our faces. . . .Yet the call is 242 PHILIP TALLON

not only to protestation and awe, it is to a reflection of the Divine life (Lewis, 1996, p. 48).

God’s love is, in some sense, mediated to us through the understanding. This is not offered merely as a retort to O’Hara’s picture of the relationship between the head and heart, but is meant to get to a deeper point. Many aspects of Lewis’s theodicy are not intended to provide novel arguments to buttress faith in God, but simply to unpack core concepts of Christian belief. Understanding more fully the nature of God’s love for us brings with it a deeper grasp of why God allows suffering. Thus, I take Lewis’s theodicy to be largely a more searching exploration of basic truths about God the Christian already understands. The heart which experiences God’s love already knows, if only dimly, many of the ideas that Lewis unfolds in his argument. Here again we see that theodicy and sound theology are not ultimately separable. One further minor point. Unlike O’Hara, I do not find Lewis’s talk of “games” in theodicy to be problematic. I think it is a helpful analogy, at least as applied to the relationship between human decision making and a stable natural environment, as Lewis does (Lewis, 1996, p. 30). But I also recognize that pastoral theology is an art. Context matters in discussing theodicy. Theodicists should be careful about their words and take O’Hara’s warning (and all his insights) as helpful input. I know I will. I genuinely look forward to O’Hara’s response—and am glad to give him the last word in this book—because there is no “last word” in theology this side of the eschaton; all of us now see through a glass, darkly (I Cor. 13:12). Neither I, nor Lewis, intend to foreclose the theodicy discussion; all our suggestions are provisional and open to correction. But this does not mean that Christians should not faithfully (and with a certain confidence tempered by humility) attempt to speak about how we fit into the story told by God, who works all for the good of those who love Him.

TWENTY

REPLY TO PHILIP TALLON

David L. O’Hara

What is the task of Christian apologetics? “Apologetics” harks back to that sentence of St. Peter, who wrote that Christians should always be prepared to give an apologia for the hope that is within them. The word means “a speech given in defense,” for instance, the kind of speech one might give in a courtroom, or in a forum where laws and policies are decided. The task of apologetics begins right here, with the attempt to offer a defense of Christian hope. Like a courtroom speech, apologetics can serve several purposes, and it can proceed in various ways. Its purpose may be to defend the innocent from punishment or other unjust attacks; to establish the legitimacy of certain beliefs; to cast doubt upon accusations; to hearten allies and entice the curiosity of opponents; or even to prepare the ground for future appeals. One of the things I most admire in both Lewis and Tallon is the example of how they, in their apologetics, demonstrate love for their neighbors. Each, in his own way, reminds me of a child who stands up to the playground bully, a David placing himself before Goliath to defend both the honor of God and the faith of those who have trusted in God. The powerful apologias Lewis and Tallon offer, together with the example of their lives, is Christian apologetics at its finest. God needs no defense; surely God can fight God’s own battles. But those who believe often find themselves harassed and bullied by those who claim such belief is indefensible. Tallon and Lewis show us that Christian belief may be thoughtfully defended by powerful minds. It is with this in mind that I offer my reply to Tallon, not intending to attack either him or Lewis, but hoping to join them in the work of apologetics, even if I take a different angle on the question at hand. As I said in my first reply, I am offering a critique, not a refutation, and my purpose is to try to ensure that the apologia that we offer is as strong and as helpful as possible. In reading Tallon’s two chapters together, I want to say “Tallon wins.” This is not mere humility; it is a recognition of three things: Tallon is plainly brilliant; we are at the edge of our understanding when we discuss these matters, and I cannot be entirely confident in my own position; and nothing matters more than grace and charity. But the grace we show one another can be to take one another’s arguments seriously. Without wishing to polemicize, let me reply instead with these three points: First, just as the task of Christian apologetics may vary with circumstances, so we may need different kinds of arguments at different times 244 DAVID L. O’HARA in our lives. Second, arguments may be of different kinds of value. As Alvin Plantinga once pointed out, an argument may be valuable and valid even if it is not universally agreed to be sound. Third, even if an argument does not prove what it sets out to prove, the example of the arguer or of the arguer’s life may be an intellectually and spiritually fortifying story. The value of The Problem of Pain is not so much in its conclusion as in its example, especially when taken together with A Grief Observed.

1. We Need Different Arguments at Different Times

In a sense, I do not want to argue against Lewis so much as against the potential misuse of Lewis’s arguments, and against the abstraction of Lewis’s arguments from the life and texts in which they occur. Particular arguments do not live outside of particular contexts. The logical form of argumentation, like the form of a syllogism, is probably something that endures throughout the ages. But the arguments themselves—the flesh and living blood of actual arguments, and not just their skeletal logic—appear in times and places, and we can make good or bad use of them. Look at Aquinas’ “five ways” of proving the existence of God as something that stands for all time and all places and you do it an injustice by making it stand still. I doubt that St. Thomas, were he alive today, would be content with his own arguments. When he wrote, he tried to square knowledge of God with the best knowledge we have of the world. The enduring legacy of St. Thomas does not lie in stale prose but in an eternal example of loving God with all of one’s being. St. Thomas does not show us God, in the end. Rather, he shows us a faithful life of seeking God with the best tools we have, and his example of seeking is more powerful than his proofs; after all, the proofs may be refuted by disputing the premises, but the life of the saint is not subject to refutation. Although I disagree with Lewis sometimes, I think it is plain that Lewis has earned a rank among the great Christian thinkers of history. Like Aquinas, Lewis illustrates the perpetual struggle for the truth, even while knowing that for now we see it only “in a glass, darkly.” In the end what matters most is not the argument but what we do with it. We must live lives of charity. If an argument helps you to do that, then it is valuable to just that degree. While I am concerned about the pastoral consequences of most of the arguments that try to explain pain and suffering, I recognize that sometimes it is helpful to see that we can think reasonably about God and the suffering. In my early years as a Christian, apologists like Lewis helped me to see my way past intellectual challenges to faith. Later in life I find I have grown to where the challenges to faith are less intellectual in nature. I have to be careful here: I do not mean to imply that intellectual obstacles to faith are a sign of immaturity. Not at all! Each of us finds different problems at different times. Lewis knew this to be true not just of Reply to Philip Tallon 245 individuals but even of different periods of history. “Every age has its own outlook. It is specially good at seeing certain truths and specially liable to make certain mistakes” (Lewis, 1970, p. 202). That is why I do not say that the only real value of theodicy is to remove obstacles to faith, but that this is its chief value. An argument that helps us begin to trust God so that we can love God and our neighbor has done good work. Such arguments can give us courage to continue to seek, and they can demonstrate that belief is intellectually defensible even if they do not give us a flawless conclusion.

2. An Argument May Be Helpful Even If Its Conclusion Is Not Certain

This brings me to my second point: an argument may be useful even if it is not universally recognized as supporting a certain conclusion. This is, after all, how a legal defense almost always works, giving the best reasons we can muster in favor of our position, even while the opposition gives the best reasons they can muster against our position. When all the speeches have been given, the outcome still must be decided, often with incomplete information. Alvin Plantinga famously concluded his modal version of the Ontological Argument by pointing out that his argument did not prove that God exists, but only that believing in God’s existence is intellectually defensible (Plantinga, 1974, p. 112). An argument may be valuable even if it does not do the work it apparently sets out to do at first. To put it differently, and maybe to take a slight liberty with Plantinga’s point: much may depend on who we are and where we are in our lives, since those will change what we are willing to believe. As Lewis wrote, “What you see and hear depends a good deal on where you are standing: it also depends on what sort of person you are” (Lewis, 1994, p. 148). I think that Lewis would agree that virtue and character matter tremendously, while what one believes about matters that the scriptures leave uncertain matters far less. This is why he wrote, in The Problem of Pain, that he was only working on an intellectual problem, and one that was of relatively small importance. Lewis also knew that in the end his words would fall short, especially if they were aiming to communicate simple facts. Words are often better suited to communicating matters of the heart. “One of the most important and effective uses of language is the emotional,” Lewis wrote. “It is also, of course, wholly legitimate. We do not talk only to reason or to inform” (Lewis, 1960, p. 215). Lewis aimed at something more than a logical proof in his writing about suffering; he hoped to inculcate courage. As I said earlier, this is one of the tasks of Christian apologetics. So I would reword very slightly one of Tallon’s characterizations of my position. He hears me saying that “Many key assumptions in Lewis’s theodicy are too speculative to provide a solid foundation for his argument.” That is pretty close to my position but not exactly it. In fact, I do think myths and stories can provide solid foundations 246 DAVID L. O’HARA for arguments—depending, of course, on what we mean by “a solid foundation.” As Lewis knew, myths and stories often provide us with reasons to act and to believe, and those reasons will feel solid enough. To appeal to the title of Lewis’s essay once again, “sometimes fairy stories may say best what’s to be said” (Lewis, 1975, p. 35). We just need to remember that when we rest our belief on a story, other people may not trust the story in the same way that we do.

3. A Positive Argument Ad Hominem

I will add that it is possible to think someone is not wrong while at the same time thinking that their position is not right. If this sounds like a logical contradiction, let me defer to a greater logician than I, Charles Sanders Peirce. When asked whether God was capable of growth, Peirce replied that “its implications concerning God will be partly disavowed, and yet held to be less false than their denial would be” (Peirce, 1958, p. 6.466). In other words, it would be false to say that God is capable of growth, but even more false to say that God is not capable of growth. The first proposition undercuts what we believe about the eternity of God; the second undercuts what we believe to be true about the limitless power of God. If we had to choose one or the other, the question would not be which one was right, but which was less false. I think the theodicy Lewis offers in The Problem of Pain is not finally and completely right. At the same time, I think he is not wrong to affirm, with the tradition, that God is good, that God will somehow make right all wrongs and dry all tears. The scriptures make these promises, and the Christian tradition affirms this hope. I just want to make sure that we never become so quietly content with the affirmation of that hope and that promise that we cease to live lives of love for God and our neighbor. We know so little. Remember the conclusion of the Book of Job: Job longs for answers, but God does not give the kind of answer Job wants, even while commending Job for speaking what he knows. Let me close by saying again that I think Lewis is more right in his later writing than in his earlier writing. My argument, in the end, is not against Lewis the man any more than it is against Tallon the man. I am for them both. I am honored to call both of them my teachers and my brothers. They, and others like them, are a gift to all who need an example of a great mind working together with a loving heart. My argument is against taking Lewis’s earlier book as complete solution to the problem named in its title. Furthermore, I think Lewis himself would agree with me, as evidenced by the fact that in his last years he returned to the problem with an entirely different approach. St. Augustine, in the late evening of his intellectual life, looked back over his life and writings and reconsidered his earlier positions. In his book The Retractions, he compiled a list of the ways in which he thought that he had Reply to Philip Tallon 247 been mistaken as a younger man, and of the things he would have said differently. All of us should be so blessed as to have the opportunity to review our mistakes and to make them known so that others do not fall into our errors. That book is one of the greatest illustrations of the life of a Christian thinker. We know that we see only in part, that at our very best we do not know as fully as we are known, and that our greatest ratiocinations are but playful imitations of the mind of God. We offer them up not as perfections, but with the joy of being allowed to play in the Master’s workshop. We are like Aulë in Tolkien’s Silmarillion, who, when he made the dwarves without Eru’s permission, defended his action by saying, “The making of things is in my heart from my own making by thee; and the child of little understanding that makes a play of the deeds of his father may do so without thought of mockery, but because he is the son of his father” (Tolkien, 1977, p. 41). We are not wrong, but we should be reluctant to say too quickly that our conclusions are right. Our rightness is in our seeking and in the joyful, loving, and rigorous exercise of the gifts we have been given. Lewis’s A Grief Observed is a kind of revision of The Problem of Pain. Lewis does not reject his earlier work, but he returns to the same subject and, in a time of personal grief, he offers a very different angle on pain. Neither book ends theological investigation, just as neither book taken alone gives us the authoritative Lewis. But taken together, they remind us of two things: first, that even a singular mind like Lewis’s may find that The Problem of Pain is not enough. Second, the faithful Christian life is one of following, not of stopping to defend one’s earlier views. Once again, remember Lewis’s statement that he would write about his grief until he ran out of notebooks. Taken at face value, this was the limit he set on his book. But is it not also a kind of metaphor for the life of seeking God? The “notebooks” we all have are the days and hours we may dedicate to thinking, talking, and praying. Our seeking could, and should, run for as long as our lives are given more pages on which to write. In Easter of 1945, Lewis gave a talk to an assembly of Anglican priests in Carmarthen. His talk was about apologetics, and he concluded with a warning about the dangers and limitations of apologetics:

I have found that nothing is more dangerous to one’s own faith than the work of an apologist. No doctrine of that Faith seems to me so spectral, so unreal as one that I have just successfully defended in a public debate. For a moment, you see, it has seemed to rest on oneself: as a result, when you go away from that debate, it seems no stronger than that weak pillar. That is why we apologists take our lives in our hands and can be saved only by falling back continually from the web of our own arguments, as from our intellectual counters, into the Reality—from Christian apologetics into Christ himself. That is also why we need one another’s 248 DAVID L. O’HARA

continual help—oremus pro invicem. [Let us pray for one another.] (Lewis, 1970, p. 103)

And this, in the end, is Lewis’s legacy. He reminds us that we must continue to try to give as helpful, clear, and charitable a defense of the faith as we can; and at the same time, we must never rest upon ourselves. Knowing that we will be wrong, we must always rest upon the one of whom Augustine wrote in the Confessions that “our hearts are restless until they find their rest in Thee.” And we must depend upon one another for correction, never trusting too much in our own successful defenses of our beliefs. If this book we have written has been of any value, it is because Lewis lived—and wrote—such an example in his own books that we, the authors of these chapters, find ourselves moved to imitate his example and to continue his work; just as we now invite those who may chance upon these chapters not to take us as giving the final word on Lewis, or on God. Rather, pick up the argument where we have left it, and join us in this joyful, convivial pursuit.

WORKS CITED

Introduction Gregory Bassham

Addison, Joseph, Richard Steele, and Others (1907) The Spectator, vol. 2. London: Dent. Anscombe, G. E. M. (1981) Metaphysics and the Philosophy of Mind. Minneapolis: University of Minnesota Press. Aquinas, Thomas. (1945) Basic Writings of Saint Thomas Aquinas, vol. 1, ed. Anton C. Pegis. New York: Random House. ———. (1956b) On the Truth of the Catholic Faith, Book, 3, Part I. Translated by Vernon J. Bourke. Garden City, N.Y: Hanover House. ______. (1956b) On the Truth of the Catholic Faith: Summa Contra Gentiles. Book Two. Translated by James F. Anderson. Garden City, N.Y..: Hanover House. Audi, Robert. (2011) Epistemology: A Contemporary Introduction to the Theory of Knowledge. New York: Routledge. Baggett, David and Jerry L. Walls. (2011) Good God: The Theistic Foundations of Morality. New York: Oxford University Press. Balfour, Arthur J. (1879) The Foundations of Belief. London: Macmillan. ______. (1915) Theism and Humanism. New York: Hodder and Stoughton. Bedau, Mark A. and Paul Humphreys, eds. (2008) Emergence: Contemporary Readings in Philosophy and Science. Cambridge, Mass.: MIT Press. Beversluis, John. (1985) C. S. Lewis and the Search for Rational Religion. Grand Rapids, Mich.: Eerdmans. Blackburn, Simon. (2005) Truth: A Guide. New York: Oxford University Press. Bradley, F. H. (1883) The Principles of Logic. London: Kegan Paul, Trench, and Co. Brazier, P. H. (2012) C. S. Lewis: Revelation and the Christ, vol. 2. Eugene, Oreg.: Wipf and Stock. Chesterton, G. K. (1959) Orthodoxy. New York: Image Books. ______. (1993) The Everlasting Man. San Francisco: Ignatius Press. Clayton, Philip and Paul Davies, eds. (2006) The Re-Emergence of Emergence: The Emergentist Hypothesis from Science to Religion. Oxford: Oxford University Press. Copleston, F. C. (1964) “Bertrand Russell and F. C. Copleston: A Debate on the Existence of God.” Reprinted in The Existence of God, ed. John Hick. New York: Macmillan. Originally broadcast by the BBC in 1948. Davis, Stephen T. (2004) “The Mad/Bad/God Trilemma: A Reply to Daniel Howard- Snyder,” Faith and Philosophy, 21:4. Descartes, René. (1958) Meditations on First Philosophy. In Philosophical Writings. Selected and translated by Norman Kemp Smith. New York: Modern Library. Dickerson, Matthew. (2011) The Mind and the Machine: What It Means to Be Human and Why It Matters. Grand Rapids, Mich.: Brazos Press. Evans, C. Stephen. (2010) Natural Signs and Knowledge of God: A New Look at Theistic Arguments. New York: Oxford University Press. 250 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Ficino, Marsilio. (1948) Five Questions Concerning the Mind. In The Renaissance Philosophy of Man, eds. Ernest Cassirer, P. O. Kristeller, and J. H. Randall. Chicago: University of Chicago Press. Goetz, Stewart and Charles Taliaferro. (2008) Naturalism. Grand Rapids, Mich: Eerdmans. Guthrie. W. K. C. (1965) A History of Greek Philosophy, vol. 2. Cambridge: Cambridge University Press. Haldane, J. B. S. (1929) Possible Worlds and Other Essays. London: Chatto and Windus. Hasker, William. (1999) The Emergent Self. Ithaca, N.Y.: Cornell University Press. Holyer, Robert. (1988) “The Argument from Desire,” Faith and Philosophy, 5(1), pp. 61-71. Hooker, Richard. (1845) Works, vol. 1. Oxford: Oxford University Press. Horner, David A. “Aut Deus Aut Malus Homo: A Defense of C. S. Lewis’s ‘Shocking Alternative.’” In David Baggett, Gary R. Habermas, and Jerry l. Walls, eds. C. S. Lewis as Philosopher: Truth, Goodness, and Beauty. Downers Grove, Ill.: InterVarsity Press. Howard-Snyder, Daniel. (2004) “Was Jesus Mad, Bad, or God? . . . Or Merely Mistaken?” Faith and Philosophy 21:4, pp. 456-79. Hume, David. (1977) An Inquiry Concerning Human Understanding. Indianapolis: Hackett. ______. (1978) A Treatise of Human Nature, ed. P. H. Nidditch. Oxford: Clarendon Press. James, William. (1950) Principles of Psychology. New York: Dover. Joad, C. E. M. (1933) Guide to Modern Thought. London: Faber. Kant, Immanuel. (1956) Critique of Practical Reason. Translated by Lewis White Beck. Indianapolis: Bobbs-Merrill. ______. (1961) A Critique of Pure Reason. Translated by Norman Kemp Smith. New York: St. Martin’s Press. ______. (1993) Grounding for the Metaphysics of Morals. Translated by James W. Ellington. Indianapolis: Hackett. Kierkegaard, Soren. (1968) Concluding Unscientific Postscript. Translated by David F. Swenson. Princeton: Princeton University Press. Kim, Jaegwon. (2011) Philosophy of Mind. (3rd ed.; 1st ed. 1996). Boulder, Colo.: Westview Press, 2011. Knox, Ronald. (1958) The Belief of Catholics. New York: Image Books. Kreeft, Peter and Ronald Tacelli. (1994) A Handbook of Christian Apologetics. Downers Grove, Ill.: InterVarsity Press. Lewis, C. S. (1947) Miracles: A Preliminary Study. London: G. Bles. ______. (1949) The Weight of Glory and Other Addresses. New York: Macmillan. ______. (1956) Surprised by Joy. New York: Harcourt Brace. ______. (1965) “Is Theology Poetry?” In Screwtape Proposes a Toast. London: Collins. ______. (1967) “De Futilitate.” In Walter Hooper ed., The Seeing Eye and Other Selected Essays from Christian Reflections. New York: Ballantine Books. ______. (1970) “Bulverism.” In God in the Dock: Essays on Theology and Ethics, ed. Walter Hooper. Grand Rapids, Mich.: Eerdmans. ______. (1977) The Pilgrim’s Regress. Grand Rapids, Mich.: Eerdmans. Works Cited 251

______. (1987) “On Living in an Atomic Age.” In Present Concerns, ed. Walter Hooper. San Diego: Harcourt, Brace, Jovanovich. ______. (2001a) Mere Christianity. San Francisco: HarperSanFrancisco. ______. (2001b) Miracles: A Preliminary Study. San Francisco: HarperSanFrancisco. ______. (2001c) The Problem of Pain. San Francisco: HarperSanFrancisco. ______. (2004) The Collected Letters of C. S. Lewis, vol. 2, ed. Walter Hooper. San Francisco: HarperSanFrancisco. ______. (2007) The Collected Letters of C. S. Lewis, vol. 3, ed. Walter Hooper. San Francisco: HarperSanFrancisco. Liddon, H. P. (1867) The Divinity of Our Lord and Savior Jesus Christ. London: Rivingtons. Linville, Mark D. (2009) “The Moral Argument.” In The Blackwell Companion to Natural Theology, eds. William Lane Craig and J. P. Moreland. Oxford: Blackwell, pp. 391-448. Mackie, J. L. (1977) Ethics: Inventing Right and Wrong. Harmondsworth: Penguin Books. McDowell, Josh. (1979) Evidence That Demands a Verdict. San Bernardino, Calif.: Here’s Life Publishers. McGrath, Alister. (2014) The Intellectual World of C. S. Lewis. Malden, Mass.: Wiley- Blackwell. Nagel, Thomas. (2012) Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature is Almost Certainly False. Oxford: Oxford University Press. Newman, John Henry. (1979) An Essay in Aid of a Grammar of Assent. Notre Dame, Ind.: University of Notre Dame Press. Paul, Richard. (1993) Critical Thinking (3rd ed.; 1st ed. 1990). Santa Rosa, Calif.: Foundation for Critical Thinking. Plantinga, Alvin. (2011) Where the Conflict Really Lies: Science, Religion, and Naturalism. New York: Oxford University Press. Reppert, Victor. (2003) C. S. Lewis’s Dangerous Idea: In Defense of the Argument from Reason. Downers Grove, Ill.: InterVarsity Press. ______. (2009) “The Argument from Reason.” In The Blackwell Companion to Natural Theology, eds. William Lane Craig and J. P. Moreland. Oxford: Blackwell, pp. 344-390. Ritchie, Angus. (2012) From Morality to Metaphysics: The Theistic Implications of our Ethical Commitments. Oxford: Oxford University Press. Robb, David and John Heil. (2013) “Mental Causation.” Stanford Encyclopedia of Philosophy. Available at: http://plato.stanford.edu/entries/mental-causation/ (accessed February 24, 2014). Ruse, Michael. (1986) Taking Darwin Seriously: A Naturalistic Approach to Philosophy. Oxford: Blackwell. Russell, Bertrand. (1927) “The Validity of Inference.” In The Basic Writings of Bertrand Russell, eds. Robert E. Egner and Lester E. Denonn. New York: Simon and Schuster, 1961. ______. (1948) Human Knowledge: Its Scope and Limits. New York: Simon and Schuster. Smilde, Arend. (2014) “Horrid Red Things: A New Look at the ‘Lewisian Argument from Desire’—and Beyond,” The Journal of Inkling Studies, 4:1, pp. 33-92. 252 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Wielenberg, Erik. J. (2008) God and the Reach of Reason: C. S. Lewis, David Hume, and Bertrand Russell. Cambridge: Cambridge University Press. Willey, Basil. (1970) Cambridge and Other Memories. London: Chatto & Windus. Young, Frances. (1977) “A Cloud of Witnesses.” In The Myth of God Incarnate, ed. John Hick. Philadelphia: Westminster Press. Zagzebski. Linda. (2004) “Does Ethics Need God?” Faith and Philosophy, 4, pp. 294– 303.

Chapter One Peter S. Williams

Barkman, Adam. (2009) C. S. Lewis and Philosophy as A Way of Life. Allentown, Pa.: Zossima Press. Camus, Albert. (1959) The Myth of Sisyphus. Translated by Justin O’Brien. New York: Vintage Books. Carnell, Corbin Scott. (1974) Bright Shadow of Reality: Spiritual Longing in C. S. Lewis. Grand Rapids, MI: Eerdmans. Chalmers, Thomas. (1943) Natural Theology, Book II. Quoted in T. C. Hammond, Reasoning Faith: An Introduction to Christian Apologetics. London: IVP. Cottingham, John. (2009) Why Believe? London: Continuum. Davis, Ashley M. (2012) “A Natural Signs Approach to the Argument from Desire,” http://beardocs.baylor.edu/xmlui/bitstream/handle/2104/8377/FinalThesis_Beard ocs_ADavis_PDFA.pdf?sequence=1. (accessed Dec. 20, 2014). Dawkins, Richard. (2006) The God Delusion. London: Bantam. Dougherty, Trent. (2005) “Argument from Desire,” http://prosblogion.ektopos.com/archives/2005/11/argument_from_d.html (accessed Dec. 20, 2014). Evans, C. Stephen. (1996) Why Believe? Reason and Mystery as Pointers to God. Downers Grove, Ill.: IVP. Geisler, Norman L. (1999) Baker Encyclopedia of Christian Apologetics. Grand Rapids, Mich.: Baker. Grahame, Kenneth. (1993) The Wind in the Willows. London: Everyman’s Library Children’s Classics. Grayling, A. C. (2011) The Good Book. London: Bloomsbury. Habermas, Gary R. and J. P. Moreland. (1998) Beyond Death: Exploring the Evidence for Immortality. Wheaton, Ill.: Crossway. Haldane, John. (2007) “Philosophy, the Restless Heart and the Meaning of Theism.” In The Meaning of Theism, ed. John Cottingham. Oxford: Blackwell. Holyer, Robert. (1988) “The Argument from Desire,” Faith and Philosophy, vol. 5: 1, pp. 61-71. Joad, C. E. M. (1946) Guide to Philosophy. London: Victor Gallancz. ______. (1952) The Recovery of Belief. London: Faber. Keller, Timothy. (2008) Reason for God: Belief in an Age of Skepticism. London: Hodder & Stoughton. Kreeft, Peter. (1989) Heaven: The Heart’s Deepest Longing. San Francisco: Ignatius. Lewis, C. S. (1977a) The Pilgrim’s Regress. London: Fount. ______. (1977b) The Problem of Pain. London: Fount. ______. (1985) “Meditation in a Toolshed.” In First and Second Things. London: Fount. Works Cited 253

———. (1997) Mere Christianity. London: Fount. ______. (1998) Surprised by Joy. London: Fount. ______. (2002) “The Weight of Glory.” In C. S. Lewis Essay Collection: Faith, Christianity and the Church, ed. Lesley Walmsley. London: HarperCollins. Lovell, Steven. (2003) Philosophical Themes from C. S. Lewis. University of Sheffield, myweb.tiscali.co.uk/annotations/content/phd_thesis.pdf . McGrath, Alister. (1992) Bridge-Building: Communicating Christianity Effectively. Leicester: IVP. ______. (2012) Mere Apologetics. Grand Rapids, Mich.: Baker. ______. (2013) C. S. Lewis: A Life. London: Hodder & Stoughton. ______. (2014) The Intellectual World of C. S. Lewis. Oxford: Wiley-Blackwell. Moreland, J. P. (2009) The God Question: An Invitation to a Life of Meaning. Eugene, Oreg.: Harvest House. Morris, Thomas V. (1999) Philosophy for Dummies. Foster City, Calif.: IDG Books. Pascal, Blaise. (1941) Pensées and Provincial Letters. New York: Modern Library. Plantinga, Alvin. (2000) Warranted Christian Belief. New York: Oxford University Press. Puckett, Joe, Jr. (2013) The Apologetics of Joy: A Case for the Existence of God from C. S. Lewis’s Argument from Desire. Cambridge: Lutterworth Press. Purtill, Richard. (2004) C. S. Lewis’s Case for the Christian Faith. San Francisco: Ignatius. Reppert, Victor. (2006) “The Bayesian Argument from Desire” http://dangerousidea.blogspot.com/2006/09/bayesian-argument-from- desire.html#comments (accessed Dec. 20, 2014). Scarry, Elaine. (1999) On Beauty and Being Just. Princeton: Princeton University Press. Sheiman, Bruce. (2009) An Atheist Defends Religion: Why Humanity Is Better Off with Religion Than Without It. New York: Alpha. Swinburne, Richard. (2010) Is There A God? rev. ed. Oxford: Oxford University Press, 2010. Tolkien, J. R. R. (1977) The Silmarillion. London: Allen & Unwin. Weatherhead, Leslie D. (1955) Psychology, Religion and Healing. London: Hodder & Stoughton.

Chapter Two Gregory Bassham

Aquinas, Saint Thomas. (1956) On the Truth of the Catholic Faith: Summa Contra Gentiles. Book 2. Translated by Vernon J. Bourke. Garden City, N.Y.: Hanover House. Aristotle. (1948) Politics. Translated by Ernest Barker. Oxford: Clarendon Press. Bassham, Gregory. (2013) “The Meaning of Life.” https://www.academia.edu/6375759/The_Meaning_of_Life (accessed December 24, 2014). Beversluis, John. (2007) C. S. Lewis and the Search for Rational Religion (rev. ed.). Amherst, N.Y.: Prometheus Books. Camus, Albert. (1959) The Myth of Sisyphus and Other Essays. Translated by Justin O’Brien. New York: Vintage Books. 254 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Dougherty, Trent. (2005) “Argument from Desire” http://prosblogion.ektopos.com/archives/2005/11/argument_from_d.html (accessed December 24, 2014). Feinberg, Joel. (1994) “Absurd Self-Fulfillment.” In Freedom and Fulfillment: Philosophical Essays. Princeton: Princeton University Press. Hare. R. M. (1972) “Nothing Matters.” Reprinted in Life, Death, and Meaning: Key Philosophical Readings on the Big Questions, ed. David Benatar. Lanham, Md.: Rowman & Littlefield. Hooker, Richard. (1845) Works. Vol. 1. Oxford: Oxford University Press. James, William. (1967) “The Dilemma of Determinism. In The Writings of William James, ed. John J. McDermott. New York: Random House. Kreeft, Peter. (1989) Heaven: The Heart’s Deepest Longing. San Francisco: Ignatius. Kreeft, Peter and Ronald K. Tacelli. (1994) Handbook of Christian Apologetics. Downers Grove, lll.: InterVarsity Press. Lewis, C. S. (1949) The Weight of Glory and Other Addresses. New York: Macmillan. ———. (1958) The Pilgrim’s Regress. Grand Rapids, Mich.: Eerdmans. ______. (1998) Surprised by Joy. San Diego: Harcourt Brace & Co. ______. (2001a) The Problem of Pain. San Francisco: HarperSanFrancisco. ______. (2001b) Mere Christianity. San Francisco: HarperSanFrancisco. Lovell, Steven. (2003) Philosophical Themes from C. S. Lewis. University of Sheffield, myweb.tiscali.co.uk/annotations/content/phd_thesis.pdf (accessed March 13, 2015). McGrath, Alister. (2014) The Intellectual World of C. S. Lewis. Oxford: Wiley- Blackwell. Morris, Thomas V. (1999) Philosophy for Dummies. Foster City, Calif.: IDG Books. Nagel Thomas. (1971) “The Absurd,” Journal of Philosophy, Vol. 68:20, pp. 716-727. Plantinga, Alvin. (2000) Warranted Christian Belief. New York: Oxford University Press. Puckett, Joe. (2013) The Apologetics of Joy: A Case for the Existence of God from C. S. Lewis’s Argument from Desire. Cambridge: Lutterworth Press. Schick, Theodore and Lewis Vaughn. (2005) How to Think about Weird Things (4th ed.). New York: McGraw-Hill. Swinburne, Richard. (1979) The Existence of God. Oxford: Clarendon Press. Wielenberg, Erik. (2008) God and the Reach of Reason. New York: Cambridge University Press. Williams, Peter S. (2013) C. S. Lewis vs. the New Atheists. Milton Keynes, U.K: Paternoster.

Chapter Three Peter S. Williams

Alston, William P. (1991) Perceiving God: The Epistemology of Religious Experience. Ithaca: Cornell University Press. Behe, Michael. (2006) Darwin’s Black Box: The Biochemical Challenge to Evolution, 10th Anniversary Edition. New York: Free Press. ______. (2007) The Edge of Evolution. New York: Free Press. Budziszewski, J. (n.d.) “Question 91, Article 2: Whether There Is in Us a Natural Law?” www.undergroundthomist.org/sites/default/files/related-documents/1- 2Q91Art2.pdf (accessed September 24, 2014). Works Cited 255

Cottingham, John. (2003) On the Meaning of Life. London: Routledge. Craig, William Lane. (2008) Reasonable Faith: Christian Truth and Apologetics (2nd ed.). Wheaton, Ill.: Crossway. ______. and J. P. Moreland, eds. (2009) The Blackwell Companion To Natural Theology. Oxford: Wiley-Blackwell. ______. and Joseph E. Gorra. (2013) “Is Life Absurd without God?” In A Reasonable Response: Answers to Tough Questions on God, Christianity and the Bible. Chicago: Moody Publishers. Downes, Stephen M. (2014) "Evolutionary Psychology," The Stanford Encyclopedia of Philosophy, Edward N. Zalta (ed.), http://plato.stanford.edu/archives/sum2014/entries/evolutionary-psychology/ (accessed September 24, 2014). Elliotta, Kyle H., Robert E. Ricklefs, Anthony J. Gaston, Scott A. Hatch, John R. Speakman, Gail K. Davoren. (2013) “High Flight Costs, but Low Dive Costs, in Auks Support the Biomechanical Hypothesis for Flightlessness in Penguins,” Proceedings of the National Academy of Sciences of the United States of America (June 4th), www.pnas.org/content/early/2013/05/16/1304838110.abstract#cited- by (accessed December 24, 2014). Endo, Hideki, Daishiro Yamagiwa, Yoshihiro Hayashi, Hiroshi Koie, Yoshiki Yamaya and Junpei Kimura. (1999) “Role of the Giant Panda’s Psuedothumb,” Nature, January 28, Vol. 347 # 6717, pp. 309-310. Gauger, Ann, Douglas Axe, and Casey Luskin. (2012) Science and Human Origins. Seattle: Discovery Institute Press. Geisler, Norman L. (2013) Christian Apologetics (2nd ed.). Grand Rapids, Mich.: Baker. Gordon, Jeffery. (2013) “Is The Existence of God Relevant?” In Exploring the Meaning of Life: An Anthology and Guide, ed. Joshua W. Seachris. Chichester: Wiley-Blackwell. Groothuis, Douglas. (2011) Christian Apologetics: A Comprehensive Case for Biblical Faith. Downers Grove, Ill.: IVP Academic/Apollos. Haldane, John. (2007) “Philosophy, the Restless Heart and the Meaning of Theism.” In The Meaning of Theism, ed. John Cottingham. Oxford: Blackwell. Holyer, Robert. (1988) “The Argument from Desire,” Faith and Philosophy, vol. 5: 1, pp. 61-71. Keener, Craig S. (2011) Miracles: The Credibility of the New Testament Accounts. Grand Rapids, Mich.: Baker Academic. Layman, Stephen C. (2007) Letters to Doubting Thomas: A Case for the Existence of God. New York: Oxford University Press. Lewis, C. S. (1937) “The Sagas and Modern Life: Morris, Mr. Yeats and the Originals,” a review of The Works of Morris and of Yeats in Relation to Early Saga Literature, by Dorothy M. Hoare, The Times Literary Supplement (May 29, 1937), p. 409. ______. (1977a) Pilgrim’s Regress. London: Fount. ______. (1977b) The Problem of Pain. London: Fount. ______. (1991) “The Empty Universe” in Present Concerns. London: Fount. ______. (1997) Mere Christianity. London: Fount. ______. (1998a) Surprised by Joy. London: Fount. ______. (1998b) Miracles. London: Fount. 256 C. S. LEWIS’S CHRISTIAN APOLOGETICS

______. “The Weight of Glory” (2002) In C. S. Lewis Essay Collection: Faith, Christianity and the Church, ed. Lesley Walmsley. London: HarperCollins. ______. (2013) “On Living in an Atomic Age.” Reprinted in Exploring the Meaning of Life: An Anthology and Guide, ed. Joshua W. Seachris. Chichester: Wiley- Blackwell. Luskin, Casey. (2012) “Junk No More: ENCODE Project Nature Paper Finds ‘Biochemical Functions for 80% of the Genome,’” Evolution News & Views, www.evolutionnews.org/2012/09/junk_no_more_en_1064001.html (accessed September 24, 2014). McGrath, Alister. (2014) The Intellectual World of C. S. Lewis. Chichester: Wiley- Blackwell. Mavrodes, George I. (1986) “Religion and the Queerness of Morality.” In Rationality, Religious Belief, and Moral Commitment, eds. Robert Audi and William J. Wainwright. Ithaca, N.Y.: Cornell University Press. Menuge, Angus. (2004) Agents under Fire: Materialism and the Rationality of Science. Lanham, Md.: Rowman & Littlefield. Meyer, Stephen C. (2013) Darwin’s Doubt: The Explosive Origin of Animal Life and the Case for Intelligent Design. New York: HarperOne. Moreland, J. P. (1987) Scaling the Secular City: A Defense of Christianity. Grand Rapids, Mich.: Baker. ______. and Kai Nielson (1993) Does God Exist? Amherst, N.Y.: Prometheus. ______. and William Lane Craig, eds. (2000) Naturalism: A Critical Analysis London: Routledge. ______. (2001) “Intelligent Design Psychology and Evolutionary Psychology: A Comparison of Rival Paradigms,” Journal of Psychology and Theology, 29:4, pp. 361-377. ______. (2002) “Intelligent Design Psychology and Evolutionary Psychology on Consciousness: Turning Water into Wine,” Journal of Psychology and Theology, 30:1, pp. 51-67. Morris, Tom. (1999) Philosophy for Dummies. New York: IDG. Nagel, Thomas. (2012) Mind and Cosmos: Why The Materialist Neo-Darwinian Conception Of Nature Is Almost Certainly False. New York: Oxford University Press. Nelson, Paul. (1997) “Jettision the Arguments, Or the Rule? The Place of Darwinian Theological Themata In Evolutionary Reasoning,” www.discovery.org/a/104 (accessed September 24, 2014). Oddie, Graham. (2005) Value, Reality and Desire. Oxford: Clarendon Press. O’Hear, Anthony. (2001) Philosophy in the New Century. London: Continuum. Peels, Rik. (2014) “The Presumption of Theism.” Unpublished paper presented at the Philosophy Network of the European Leadership Forum. Pinnock, Clark. (1980) Reason Enough: A Case for the Christian Faith. Exeter: Paternoster. Plantinga, Alvin. (2000) Warranted Christian Belief. New York: Oxford University Press. Pruss, Alexander R. (2010) “The Ontological Argument from Desire,” http://alexanderpruss.blogspot.co.uk/2010/04/ontological-argument-from- desire.html (accessed October 3, 2014). Rana, Fazale. (2008) The Cell’s Design. Grand Rapids, Mich.: Baker. Works Cited 257

Reppert, Victor. (2003) C. S. Lewis’s Dangerous Idea: In Defence of the Argument from Reason. Downers Grove, Ill.: IVP. ______. (2006) “Lovell and the Argument from Desire,” September 12, 2006, http://dangerousidea.blogspot.co.uk/2006/09/lovell-and-argument-from- desire.html (accessed September 24, 2014). Schaeffer, Francis A. (1994) A Christian View of Philosophy and Culture (2nd ed.). Wheaton, Ill.: Crossway Books. Scruton, Roger. (1999) An Intelligent Person’s Guide to Modern Culture. London: Duckworth. ______. (2009) Beauty: A Very Short Introduction. New York: Oxford University Press. Smith, Michael. (2006) “Is That All There Is?” The Journal of Ethics, 10 (1-2), pp. 75- 106. Swinburne, Richard. (2010) Is There A God? rev. ed. Oxford: Oxford University Press. Tallis, Raymond. (2011) Aping Mankind: Neuromania, Darwinitis and the Misrepresentation of Humanity. Durham: Acumen. Taylor, Charles. (2007) A Secular Age. Cambridge, Mass.: Harvard University Press. Wells, Jonathan. (2011) The Myth Of Junk DNA. Seattle: Discovery Institute Press. West, John G. (2012) “Darwin in the Dock: C. S. Lewis’s Critique of Evolution and Evolutionism.” In The Magician’s Twin: C. S. Lewis on Science, Scientism, and Society, ed. John G. West. Seattle: Discovery Institute Press. Williams, Peter S. (1999) The Case for God. Crowborough: Monarch. ______. (2004) I Wish I Could Believe In Meaning: A Response To Nihilism. Southampton: Damaris. ______. (2013a) “Truth, Faith and Hope in Life of Pi—A Philosophical Review,” Dialogue Australasia Journal, May Issue, www.bethinking.org/truth/truth-faith- and-hope-in-life-of-pi (accessed September 24, 2014). ______. (2013b) A Faithful Guide to Philosophy: A Christian Introduction to the Love of Wisdom. Milton Keynes: Paternoster. ______. (2013c) C. S. Lewis vs. The New Atheists. Milton Keynes: Paternoster.

Chapter Four Gregory Bassham

Gregory, T. Ryan. (2012) “A Slightly Different Response to Today’s ENCODE Hype,” http://www.genomicron.evolverzone.com/2012/09/a-slightly-different- response-to-todays-encode-hype/ (accessed September 30, 2014). James, William. (1956) The Will to Believe and Other Essays in Popular Philosophy. New York: Dover. Kurland, Philip B and Ralph Lerner. (1986) The Founder’s Constitution, http://- presspubs.uchicago.edu/founders/documents/a7s3.html (accessed September 30, 2014). Martin, Michael. (1986) “The Principle of Credulity and Religious Experience,” Religious Studies, 22:1, pp. 79-93. Raven, Peter H. and George B. Johnson. (1999) Biology (5th ed.). New York: McGraw-Hill.

258 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Chapter Five Victor Reppert

Anscombe, G. E. M. (1981) Metaphysics and the Philosophy of Mind. Minneapolis: University of Minnesota Press. Beversluis, John. (1985) C. S. Lewis and the Search for Rational Religion. Grand Rapids, Mich.: Eerdmans. Dawkins, Richard. (1986) The Blind Watchmaker. New York: Norton. Dennett, Daniel. (1987) The Intentional Stance. Cambridge, Mass: MIT Press. ______. (1995) Darwin’s Dangerous Idea: Evolution and the Meanings of Life. New York: Simon & Schuster. Fodor, Jerry A. (1990) A Theory of Content and Other Essays. Cambridge, Mass.: MIT Press. Hasker, William. (1999) The Emergent Self. Ithaca: Cornell University Press. Lewis, C. S. (1956) Surprised By Joy. San Diego: Harcourt Brace. ______. (1967) Christian Reflections. Grand Rapids, Mich.: Eerdmans. ______. (1978) Miracles: A Preliminary Study. New York: Macmillan. ______. (2001) Mere Christianity. San Francisco: HarperSanFrancisco. Reppert, Victor. (2003) C. S. Lewis’s Dangerous Idea: In Defense of the Argument from Reason. Downers Grove, Ill.: InterVarsity Press. Swinburne, Richard. (1986) The Evolution of the Soul (rev. ed.). Oxford: Clarendon Press.

Chapter Six David Kyle Johnson

Bourget, David and David Chalmers. (2013) “What Do Philosophers Believe?” Philosophical Studies, 3, pp. 1-36. The results of the study can be found at: http://philpapers.org/archive/BOUWDP. Dvorsky, George. (2013) “Scientific evidence that you probably don’t have free will.” io9. 14 January, 2013. http://io9.com/5975778/scientific-evidence-that-you- probably-dont-have-free-will (accessed December 27, 2014). Johnson-Laird, Philip. (2008) How We Reason. New York: Oxford University Press. Kasparov, Garry. (2008) How Life Imitates Chess: Making the Right Moves from the Board to the Boardroom. New York: Bloomsbury. Kripke, Saul. (1980) Naming and Necessity. Cambridge, Mass.: Harvard University Press. Lewis, C. S. (1947) Miracles: A Preliminary Study. London: G. Bles. ______. (2001) Miracles: A Preliminary Study. San Francisco: HarperSanFrancisco. McLaughlin, Brian and Karen Bennett. (2014) "Supervenience," The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, URL = (accessed December 27, 2014).

Chapter Seven Victor Reppert

Dennett, Daniel. (1994) “Self-Portrait.” In A Companion to the Philosophy of Mind, ed. Samuel Guttenplan. Oxford: Blackwell. Works Cited 259

Hasker, William. (1999) The Emergent Self. Ithaca, N.Y.: Cornell University Press. Feser, Edward. (2005) Philosophy of Mind: A Beginner’s Guide. Oxford: Oneworld Publications. Melnyk, Andrew. (2007) “Naturalism, Free Choices, and Conscious Experiences,” http://infidels.org/library/modern/andrew_melnyk/against-dualism.html (accessed October 16, 2014).

Chapter Eight David Kyle Johnson

Bourget, David and David Chalmers. (2013) “What Do Philosophers Believe?” http://philpapers.org/archive/BOUWDP (accessed November 7, 2014). Descartes, René. (1985) Discourse on the Method of Rightly Conducting One’s Reason and Seeking the Truth in the Sciences. In The Philosophical Writings of Descartes, vol. I. Translated by John Cottingham, Robert Stoothoff, and Dugald Murdoch. Cambridge: Cambridge University Press. Johnson, David Kyle. (2013) “Do Souls Exist?” Think 12 (35), pp. 61-75.

Chapter Nine David Baggett

Adams, Robert M. (1983) “Divine Necessity,” Journal of Philosophy, 80, pp. 741-752. ______. (1999) Finite and Infinite Goods: A Framework for Ethics. Oxford: Oxford University Press. Anscombe, G. E. M. (1958) “Modern Moral Philosophy,” Philosophy, 33:124, pp. 1- 16. Baggett, David. (2008) “Is Divine Iconoclast as Bad as Cosmic Sadist?” In C. S. Lewis as Philosopher: Truth, Goodness, and Beauty, eds. David Baggett, Gary R. Habermas, and Jerry L. Walls. Downers Grove, Ill: IVP Academic. ______. and Jerry L. Walls. (2011) Good God: The Theistic Foundation of Morality. New York: Oxford University Press. ______. (2013) “On Thomas Nagel’s Rejection of Theism,” Harvard Theological Review, 106:2, pp. 227-238. Beversluis, John. (1985) C. S. Lewis and the Search for Rational Religion. Grand Rapids, Mich.: Eerdmans. ______. (2007) C. S. Lewis and the Search for Rational Religion, rev. & updated. Amherst, N.Y.: Prometheus Books. Evans, C. Stephen. (2010) Natural Signs and Knowledge of God: A New Look at Theistic Arguments. Oxford: Oxford University Press. ______. (2013) God and Moral Obligation. Oxford: Oxford University Press. Ewing, A. C. (1973) Value and Reality: The Philosophical Case for Theism. London: George Allen & Unwin. Joyce, Richard. (2001) The Myth of Morality. Cambridge: Cambridge University Press. ______. (2006) The Evolution of Morality. Cambridge, Mass.: MIT Press. Leftow, Brian. (1990) “Is God an Abstract Object?” Nous, 24, pp. 581-598. Lewis, C. S. (1947) Miracles: A Preliminary Study. New York: MacMillan. ______. (1952) Mere Christianity. New York: Macmillan. ______. (1995) “The Poison of Subjectivism.” In Christian Reflections. Grand Rapids, Mich.: Eerdmans. 260 C. S. LEWIS’S CHRISTIAN APOLOGETICS

______. (2001) The Abolition of Man. New York: HarperCollins. ______. (2002) The Problem of Pain. New York: HarperCollins. Lovell, Steven. (2003) “Philosophical Themes from C. S. Lewis.” PhD dissertation, University of Sheffield, http://etheses.whiterose.ac.uk/6054/1/398641.pdf (accessed March 13, 2015). Marks, Joel. (2012) Ethics without Morals: In Defence of Amorality. London: Routledge. Mavrodes, George. (1986) “Religion and the Queerness of Morality.” Reprinted in Ethical Theory: Classical and Contemporary Readings (2nd ed.), ed. Louis P. Pojman. New York: Wadsworth, 1995. Moreland, J. P. (2009) The Recalcitrant Imago Dei: Human Persons and the Failure of Naturalism. London: SCM Press. Morris, Thomas V. (1986) The Logic of God Incarnate. Ithaca, N.Y.: Cornell University Press. Nagel, Thomas. (2012) Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature is Almost Certaintly False. Oxford: Oxford University Press. Plantinga, Alvin. (1974) The Nature of Necessity. Oxford: Clarendon Press. ______. (1980) Does God Have a Nature? Milwaukee: Marquette University Press. ______. (1982) “How to be an Anti-Realist” In Proceedings and Addresses of the American Philosophical Association, 56 (1). New York: State University of New York Press. ______. (2011) Where the Conflict Really Lies: Science, Religion, and Naturalism. New York: Oxford University Press. Rist, John M. (2002) Real Ethics: Rethinking the Foundations of Morality. Cambridge: Cambridge University Press. Russell, Bertrand. (1986) “Why I Am Not a Christian.” Reprinted in Bertrand Russell on God and Religion, ed. A. Seckel. New York: Prometheus. Wielenberg, Erik J. (2005) Value and Virtue in a Godless Universe. New York: Cambridge University Press. ______. (2008) God and the Reach of Reason: C. S. Lewis, David Hume, and Bertrand Russell. New York: Cambridge University Press.

Chapter Ten Eric J. Wielenberg

Adams, Robert Merrihew. (1983) “Divine Necessity,” Journal of Philosophy, 80, pp. 741-752. Adams, Robert Merrihew. (1999) Finite and Infinite Goods: A Framework for Ethics. Oxford: Oxford University Press. Chalmers, David. (1996) The Conscious Mind: In Search of a Fundamental Theory. Oxford: Oxford University Press. Enoch, David. (2007) “An Outline of an Argument for Robust Metanormative Realism” In Oxford Studies in Metaethics vol. 2, ed. Russ Shafer-Landau. Oxford: Oxford University Press, pp. 21-50. Evans, C. Stephen. (2004) Kierkegaard’s Ethic of Love: Divine Commands and Moral Obligations. Oxford: Oxford University Press. ______. (2013) God and Moral Obligation. Oxford: Oxford University Press. Works Cited 261

Fales, Evan. (2010) “Divine Commands and Moral Obligation,” Philo, 13:2, pp. 151– 166. Harman, Gilbert. (1977) The Nature of Morality: An Introduction to Ethics. Oxford: Oxford University Press. Huemer, Michael, and Kovitz, Ben. (2003) “Causation as Simultaneous and Continuous,” The Philosophical Quarterly, 53:213, pp. 556-565. Kvanvig, Jonathan. (2008) "Creation and Conservation,” The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta, ed., URL = http://plato.stanford.edu/archives/fall2008/entries/creation-conservation/ (accessed March 13, 2015). Lewis, C. S. (1995) “The Poison of Subjectivism.” In Christian Reflections. Grand Rapids, Mich.: Eerdmans, pp. 72-81. Mackie, J. L. (1977) Ethics: Inventing Right and Wrong. New York: Penguin. Morriston, Wes. (2009) “The Moral Obligations of Reasonable Non-believers: A Special Problem for Divine Command Metaethics,” International Journal for Philosophy of Religion, 65:1, pp. 1-10. Morrow, David. (2009) “Moral Psychology and the ‘Mencian Creature,’” Philosophical Psychology, 22:3, pp. 281-304. Plantinga, Alvin. (1974) The Nature of Necessity. Oxford: Oxford University Press. Prinz, Jesse. (2008) “Resisting the Linguistic Analogy: A Commentary on Hauser, Young, andCushman.” In Moral Psychology Volume 2: The Cognitive Science of Morality: Intuition and Diversity, ed. Walter Sinnott-Armstrong. Cambridge, Mass.: MIT Press, pp. 157-170. Singer, Peter. (2006) “Morality, Reason, and the Rights of Animals.” In Frans de Waal, Primates and Philosophers: How Morality Evolved. Princeton: Princeton University Press, pp. 140-160. Wielenberg, Erik J. (2005) Value and Virtue in a Godless Universe. New York: Cambridge University Press. ______. (2008) God and the Reach of Reason: C. S. Lewis, David Hume, and Bertrand Russell. New York: Cambridge University Press. ______. (2009) “In Defense of Non-Natural, Non-Theistic Moral Realism,” Faith and Philosophy, 26:1, pp. 23-41. ______. (2010) “On the Evolutionary Debunking of Morality,” Ethics, 120, pp. 441- 464. ______. (2014) Robust Ethics: The Metaphysics and Epistemology of Godless Normative Realism. Oxford: Oxford University Press. Zangwill, Nick. (2008) “Moral Dependence.” In Oxford Studies in Metaethics, vol. 3, ed. R. Shafer-Landau. Oxford: Oxford University Press, pp. 109-128.

Chapter Eleven David Baggett

Jordan, Matthew C. (2013) “Divine Commands or Divine Attitudes?” Faith and Philosophy, 30:2 (April 2013), pp. 159-170. Mavrodes, George I. (1986) “Religion and the Queerness of Morality.” In Rationality, Religious Belief, and Moral Commitment: New Essays in Philosophy of Religion, eds. Robert Audi and William J. Wainwright. Ithaca, N.Y.: Cornell University Press, pp. 213-226. 262 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Wielenberg, Erik J. (2009) “In Defense of Non-Natural, Non-Theistic Moral Realism,” Faith and Philosophy, 26:1 (January 2009), pp. 23-41.

Chapter Twelve Eric J. Wielenberg

Janiak, Andrew. (2014) "Newton's Philosophy," The Stanford Encyclopedia of Philosophy, Edward N. Zalta (ed.), URL = http://plato.stanford.edu/archives/sum2014/entries/newton-philosophy/ (accessed November 3, 2014). Jordan, Matthew Carey (2012) “Divine Attitudes, Divine Commands, and the Modal Status of Moral Truths,” Religious Studies, 48:1, pp. 45-60. Miller, Christian. (2009a) “Divine Desire Theory and Obligation.” In New Waves in Philosophy of Religion, eds. Yujin Nagasaw and Erik J. Wielenberg. New York: Palgrave Macmillan, pp. 105-124. ______. (2009b) “Divine Will Theory: Intentions or Desires?” Oxford Studies in Philosophy of Religion 2, ed. Jonathan L. Kvanvig, pp. 185-207. Murphy, Mark C. (1998) “Divine Command, Divine Will, and Moral Obligation,” Faith and Philosophy, 15:1, pp. 3-27. ______. 2011. God and Moral Law: On the Theistic Explanation of Morality. Oxford: Oxford University Press. Quinn, Philip. (1992) “The Primacy of God’s Will in Christian Ethics,” Philosophical Perspectives, 6, pp. 493-513. Wielenberg, Erik J. (2009) “In Defense of Non-Natural, Non-Theistic Moral Realism,” Faith and Philosophy, 26:1, pp. 23-41. ______. (2014) Robust Ethics: The Metaphysics and Epistemology of Godless Normative Realism. Oxford: Oxford University Press.

Chapter Thirteen Donald S. Williams

Bauckham, Richard. (2006) Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony. Grand Rapids, Mich.: Eerdmans. Beversluis, John. (1985) C. S. Lewis and the Search for Rational Religion. Grand Rapids, Mich.: Eerdmans. ______. (2007) C. S. Lewis and the Search for Rational Religion, rev. & updated. Amherst, N.Y.: Prometheus Books. Brazier, P. H. (2012) C. S. Lewis: The Work of Christ Revealed. Eugene, Oreg.: Pickwick. Bruce, F. F. (1960) The New Testament Documents: Are They Reliable? Downers Grove, Ill: InterVarsity Press. Chesterton, Gilbert Keith. (1925) The Everlasting Man. New York: Dodd, Mead, and Co. Davis, Stephen T. (2004) “The Mad/Bad/God Trilemma: A Reply to Daniel Howard- Snyder,” Faith and Philosophy, 21:4, pp. 480-492. Hinten, Marvin D. (2008) “Approaches to Teaching Mere Christianity,” The Lamp- Post of the Southern California C. S. Lewis Society, 30:2, pp. 3-11. Hooper, Walter, ed. (2004) The Collected Letters of C. S. Lewis, vol. 3. San Francisco: HarperSanFrancisco. Works Cited 263

Horner, David A. (2008) “Aut Deus aut Malus Homo: A Defense of C. S. Lewis’s ‘Shocking Alternative.’” In C. S. Lewis as Philosopher: Truth, Goodness, and Beauty, eds. David Baggett, Gary Habermas, and Jerry L. Walls. Downers Grove, Ill.: IVP Academic, pp. 68-84. Howard-Snyder, Daniel. (2004) “Was Jesus Mad, Bad, or God? . . . Or Merely Mistaken?” Faith and Philosophy, 21:4, pp. 456-479. Kreeft, Peter. (1988) Fundamentals of the Faith: Essays in Christian Apologetics. San Francisco: Ignatius. Kreeft, Peter, and Ronald Tacelli. (1994) Handbook of Christian Apologetics. Downers Grove, Ill.: InterVarsity. Lewis, C. S. (1943) Mere Christianity. N.Y.: Macmillan. ______. (1947) Miracles: A Preliminary Study. N.Y.: Macmillan. ______. (1967) “Modern Theology and Biblical Criticism.” Paper given at Westcott House, Cambridge, 11 May 1959. In Christian Reflections, ed. Walter Hooper. Grand Rapids, Mich.: Eerdmans, pp. 152-166. McGrath, Alister. (2013) C. S. Lewis: A Life. Carol Stream, Ill.: Tyndale. Morison, Frank. (n.d.) Who Moved the Stone? Downers Grove, Ill.: InterVarsity Press. Neill, Stephen. (1966) The Interpretation of the New Testament, 1861-1961. New York: Oxford University Press. Purtill, Richard L. (1981) C. S. Lewis’s Case for the Christian Faith. San Francisco: Harper & Row. Wright, N. T. (2007) “Simply Lewis: Reflections on a Master Apologist after 60 Years,” Touchstone: A Journal of Mere Christianity, 20:2, pp. 28-33. Young, Frances. (1977) “A Cloud of Witnesses.” In The Myth of God Incarnate, ed. John Hick. Philadelphia: Westminster, pp. 13-47.

Chapter Fourteen Adam Barkman

Bird, Michael, Craig A. Evans, and Simon Gathercole. (2014) How God Became Jesus: The Real Origins of Belief in Jesus’ Divine Nature—A Response to Bart Ehrman. Greand Rapids, Mich.: Zondervan. Blomberg, Craig. (1997) Jesus and the Gospels. Nashville, Tenn.: Broadman & Holman. Boyd, Gregory. (2013) “Whether or Not There was a Historical Adam, Our Faith Is Secure,” in Four Views on the Historical Adam, ed. Stanley Gundry. Grand Rapids, Mich.: Zondervan. Davis, Stephen T. (2004) “The Mad/Bad/God Trilemma: A Reply to Daniel Howard- Snyder,” Faith and Philosophy, 21:4, pp. 480-492. Howard-Snyder, Daniel. (2004) “Was Jesus Mad, Bad, or God? . . . Or Merely Mistaken?” Faith and Philosophy, 21:4, pp. 456-479. Lewis, C. S. (1999) Mere Christianity. In C. S. Lewis: Selected Books. Long edition. London: HarperCollins. Milton, John. (1957) The Christian Doctrine. In John Milton: Complete Poems and Major Prose, ed. Merritt Y. Hughes. Indianapolis: The Odyssey Press. Robinson, John. (1985) The Priority of John. London: SCM. Swinburne, Richard. (2008) Was Jesus God? Oxford: Oxford University Press.

264 C. S. LEWIS’S CHRISTIAN APOLOGETICS

Chapter Fifteen Donald S. Williams

Gilson, Tom. (2014) “The Gospel Truth of Jesus,” Touchstone: A Journal of Mere Christianity, 27:3, pp. 35-40. Lewis, C. S. (1967) “Modern Theology and Biblical Criticism.” In The Seeing Eye and Other Selected Essays from Christian Reflections. New York: Ballantine Books.

Chapter Sixteen Adam Barkman

Hanson, R. P. C. (2005) The Search for the Christian Doctrine of God: The Arian Controversy, 318-381. Grand Rapids, Mich.: BakerAcademic. Swinburne, Richard. (2010) The Resurrection of God Incarnate. Oxford: Oxford University Press. Wiles, Maurice. (2004) Archetypal Hersey: Arianism through the Centuries. Oxford: Oxford University Press.

Chapter Seventeen Philip Tallon

Alston, William P. (2001) “Some (Temporarily) Final Thoughts on the Evidential Arguments from Evil.” In The Evidential Argument from Evil, ed. Daniel Howard-Snyder. Bloomington, Ind.: Indiana University Press, pp. 311-332. Bassham, Gregory. (2005) “Some Dogs Go to Heaven: Lewis on Animal Salvation.” In The Chronicles of Narnia and Philosophy: The Lion, the Witch, and the Worldview, eds. Gregory Bassham and Jerry L. Walls. Chicago: Open Court, pp. 273-286. Hick, John. (1978) Evil and the God of Love (rev. ed.). New York: Harper & Row. Lewis, C. S. (1955) Surprised by Joy. Orlando: Harcourt. ______. (2001a) Mere Christianity. New York: HarperCollins. ______. (2001b) The Great Divorce. New York: HarperCollins. ______. (2001c) The Problem of Pain. New York: HarperCollins. Murray, Michael. (2009) Nature Red in Tooth and Claw: Theism and the Problem of Animal Suffering. New York: Oxford. O’Rourke, Meghan. (2011) “Ask the Author Live: Meghan O’Rourke on Grief.” NewYorker.com, http://www.newyorker.com/books/ask-the-author/ask-the- author-live-meghan-orourke-on-grief (Feb. 28. 2011) (accessed March 13, 2015). Plantinga, Alvin. (1974) God, Freedom, and Evil. Grand Rapids, Mich.: Eerdmans. Rowe, William L. (2001) “The Problem of Evil and Some Varieties of Atheism.” In The Evidential Argument from Evil, ed. Daniel Howard-Snyder. Bloomington: Indiana University Press, pp. 1-11. Swinburne, Richard. (1998) Providence and the Problem of Evil. Oxford: Clarendon Press. Walls, Jerry L. (2009) “Heaven and Hell.” In The Oxford Handbook of Philosophical Theology, eds. Thomas P. Flint and Michael C. Rea. New York: Oxford, pp. 491- 511. Works Cited 265

Wykstra, Stephen J. (2001) “Rowe’s Noseeum Arguments from Evil.” In The Evidential Argument from Evil, ed. Daniel Howard Snyder. Bloomington: Indiana University Press, pp. 126-150.

Chapter Eighteen David L. O’Hara

Lewis, C. S. (1965) Out Of the Silent Planet. New York: Macmillan. ______. (1980) The Silver Chair. New York: Collier. ______. (1982) On Stories. New York: Harcourt. ______. (1996) The Problem of Pain. San Francisco: HarperSanFrancisco. ______. (2002a) A Grief Observed. In The Complete C. S. Lewis Signature Classics. San Francisco: HarperSanFrancisco. ______. (2002b) Mere Christianity. In The Complete C. S. Lewis Signature Classics. San Francisco: HarperSanFrancisco. ______. (2004) The Lion, the Witch, and the Wardrobe. In The Chronicles of Narnia. New York: HarperCollins. Peirce, C. S. (1955) Philosophical Writings of Peirce, ed., Justus Buchler. New York: Dover. Plato. (2002) Meno. In Plato: Five Dialogues (2nd ed.). Translated by G. M. A. Grube and John M. Cooper. Indianapolis: Hackett. Rorty, Richard and Gianni Vattimo. (2005) The Future of Religion. New York: Columbia. Taliaferro, Charles. (2006) Love, Love, Love and Other Essays. Cambridge, Mass.: Cowley. Wettstein, Howard. (2012) The Significance of Religious Experience. Oxford: Oxford University Press.

Chapter Nineteen Philip Tallon

Kempis, Thomas à. (1943) The Imitation of Christ. In The Consolation of Philosophy, ed. Irwin Edman. New York: Random House. Lewis, C. S. (1961) A Grief Observed. New York: Seabury. ______. (1996) The Problem of Pain. New York: HarperCollins. ______. (2001) Mere Christianity. New York: HarperCollins.

Chapter Twenty David L. O’Hara

Lewis, C. S. (1960) Studies in Words. Cambridge: Cambridge University Press. ______. (1970) God in the Dock. Grand Rapids, Mich.: Eerdmans. ______. (1975) Of Other Worlds: Essays and Stories. New York: Harcourt Brace Jovanovich. ______. (1994) The Magician’s Nephew. New York: HarperCollins. Peirce, C. S. (1958) The Collected Papers of Charles Sanders Peirce, eds. C. Hartshorne, P. Weiss, and A. Burks. Cambridge: Harvard University Press. Plantinga, Alvin (1974) God, Freedom, and Evil. Grand Rapids. Mich.: Eerdmans. Tolkien, J. R. R. (1977) The Silmarillion. New York: Ballantine Books.

THE CONTRIBUTORS

David Baggett is Professor of Philosophy and Apologetics at Liberty Baptist Theological Seminary (Virginia) and Executive Editor of MoralApologetics.com. Among his books are (with Jerry Walls) Good God: The Theistic Foundations of Morality (2011) and (with Gary Habermas and Jerry Walls) C. S. Lewis as Philosopher: Truth, Goodness, and Beauty (2008). He and Walls are wrapping up their new book tentatively entitled God and Cosmos: Moral Truth and Human Meaning. He was the plenary speaker at the 2014 annual meeting of the C. S. Lewis and the Inklings Society at Wesleyan University. Among his articles on Lewis are “Rats in God’s Laboratory: Shadowlands and the Problem of Evil,” in Kimberly Blessing and Paul Tudico (eds.), Movies and the Meaning of Life (2005); “Is Divine Iconoclast as Bad as Cosmic Sadist?” in C. S. Lewis as Philosopher (2008); and “On Thomas Nagel’s Rejection of Theism,” Harvard Theological Review (April 2013).

Adam Barkman (PhD, Free University of Amsterdam) is an Associate Professor of Philosophy and Chair of the Philosophy Department at Redeemer University College (Canada). He is the author of five books, including C. S. Lewis and Philosophy as a Way of Life (2009), and the co-editor of four books, most recently The Philosophy of Ang Lee (2013). He speaks and writes frequently on Lewis.

Gregory Bassham is Professor of Philosophy at King's College (Pennsylvania). His publications include The Chronicles of Narnia and Philosophy (2005), co-edited with Jerry Walls; The Lord of the Rings and Philosophy (2003), co-edited with Eric Bronson; "Lewis and Tolkien on the Power of the Imagination," in David Baggett, Gary R. Habermas, and Jerry L. Walls (eds.), C. S. Lewis as Philosopher (2008); and "Who Could Have Deserved It? C. S. Lewis on Friendship,” in Suzanne Bray and William Gray (eds.), Persona and Paradox: Issues of Identity in C. S. Lewis, His Friends and Associates (2012).

David Kyle Johnson is Associate Professor of Philosophy at King’s College (Pennsylvania). He has published numerous academic articles in journals such as Religious Studies, Sophia, Philo, and Think, and has written extensively for the Blackwell Philosophy and Pop Culture series. His books include Inception and Philosophy: Because It’s Never Just a Dream (2011) and The Myths that Stole Christmas (forthcoming from Humanist Press).

David L. O’Hara is Associate Professor of Philosophy and Classics at Augustana College (South Dakota). His publications include From Homer to Harry Potter (2006); Narnia and the Fields of Arbol: The Environmental 268 Contributors

Vision of C. S. Lewis (2008); and Downstream: Reflections on Brook Trout, Fly-Fishing, and the Waters of Appalachia (2014), all co-authored with Matthew T. Dickerson.

Victor Reppert teaches philosophy at Arizona State University West and Brandman University. He is the author of C. S. Lewis’s Dangerous Idea: In Defense of the Argument from Reason (2003) and has written numerous essays on philosophical themes in Lewis, including an extended defense of the argument from reason in The Blackwell Companion to Natural Theology (2009).

Philip Tallon is Assistant Professor of Theology at Houston Baptist University, where he is Chair of the Department of Apologetics and an instructor in the Honors College. His publications include The Poetics of Evil: Toward an Aesthetic Theodicy (2011), “Evil and the Cosmic Dance: C. S. Lewis and Beauty’s Place in Theodicy,” in David Baggett, Gary Habermas, and Jerry L. Wall (eds.), C. S. Lewis as Philosopher (2008). He is also the co- editor, with David Baggett, of The Philosophy of Sherlock Holmes (2012).

Erik J. Wielenberg is Professor of Philosophy at DePauw University in Greencastle, Indiana. He is the author of Value and Virtue in a Godless Universe (2005), God and the Reach of Reason: C. S. Lewis, David Hume, and Bertrand Russell (2007), and Robust Ethics: The Metaphysics and Epistemology of Godless Normative Realism (2014).

Donald T. Williams is R. A. Forrest Scholar at Toccoa Falls College and president of the International Society of Christian Apologetics. He has published nine books, including Mere Humanity: G. K. Chesterton, C. S. Lewis, and J. R. R. Tolkien on the Human Condition (2006), Inklings of Reality: Essays Toward a Christian Philosophy of Letters (2012), Reflections from Plato’s Cave: Essays in Evangelical Philosophy (2013), and, with Jim Prothero, Gaining a Face: The Romanticism of C. S. Lewis (2014).

Peter S. Williams studied philosophy at Cardiff University (BA), Sheffield University (MA) and at the University of East Anglia in Norwich (MPhil). He is Philosopher in Residence at the UK-based Damaris Trust and Assistant Professor in Communication and Worldviews at Gimlekollen School of Journalism and Communication, NLA University, Norway. His publications include: A Faithful Guide to Philosophy: A Christian Introduction to the Love of Wisdom (2013); C. S. Lewis vs. the New Atheists (2013); Understanding Jesus: Five Ways to Spiritual Enlightenment (2011); A Sceptic's Guide to Atheism (2009); I Wish I Could Believe in Meaning: A Response to Nihilism (2004); The Case for Angels (2002); and The Case for God (1999).

INDEX abductive arguments, 5, 20, 27, 34-37, Camus, Albert, 42, 43, 53, 73 47-48, 55 Carnell, Scott Corbin, 28 Abolition of Man, The, 1, 123, 153, 158 causal closure principle, and naturalism, Adams, Robert M., 121, 137, 138, 139, 15, 18, 88, 98, 108, 115-116, 142-146, 149, 153, 163-167 147, 161 à Kempis, Thomas, 239 Chalmers, Thomas, 28, 38 Alexander, Samuel, 32 Chesterton, G. K., 6, 21, 177, 182 Alston, William, 219 Christianity Today, 1 animal suffering, 211, 224-225 Cloud of Unknowing, The, 234 Anscombe, Elizabeth, 7-9, 14, 28, 81-83, computers, and reasoning, 18, 97, 110, 91-93 117-118 Anselm, St., 199, 230, 234 Copi, Irving, 7 Anthony, Louise, 133 Copleston, F. C., 19 Aquinas, St. Thomas, 2, 6, 40, 52, 136, Cottingham, John, 33, 42-43 138, 237, 244 cumulative-case arguments, in argument from desire, 2-5, 27-73 apologetics, 45 argument from reason, 5-19, 75-120 Aristotle, 40, 52, 64-66, 71-72 Darth Vader photomosaic, 95, 100, 102, Athanasius, 199, 216 106-107, 114 Augustine, 33, 138, 139, 227, 234, 237, Davis, Stephen, 21, 184, 192 239, 246, 248 Dawkins, Richard, 30, 75, 79, 86, 107 Ayer, A. J., 1 deductive argument from desire, 4, 40- 41, 51-52, 64-66 Baggett, David, 20 “De Futilitate,” 5 Balfour, Arthur, 7, 8, 76 Deep Blue (computer), 97, 103, 117-18 Barfield, Owen, 79 Dennett, Daniel, 75, 86, 108, 111 Barkman, Adam, 22 Descartes, Renè, 10, 11-13, 14, 117-118 Bauckham, Richard, 176 determinism, and naturalism, 6-7, 108 Berengar of Tours, 234 divine command theory of ethics, 136, Bernardus Silvestris, 227 142-146, 155-159 Beversluis, John, 4, 7, 48-50, 62, 83, 84, Dostoyevsky, Fyodor, 133 123-126, 173-177, 181, 182, 183 Dougherty, Trent, 39, 48 Bird, Michael, 198 Downes, Stephen, 61 Blomberg, Craig, 196 blue flower motif, 28 Eden, 216 Boethius, 40, 237 Ehrman, Bart, 198 Boyd, Gregory, 199 eliminative materialism, 86, 96, 99 Bradley, F. H., 7-8 emergent materialism, 16, 114-115 Brazier, P. H., 188 Enoch, David, 147 Budziszewski, J., 64-65, 72 Epiphenomenalism, 96, 99, 108, 110- 111, 116, 119 Caiaphas, 195-196 Euthyphro dilemma, 136 Campolo, Tony, 126 Evans, Craig, 188 270 C. S. LEWIS’S CHRISTIAN APOLOGETICS evolutionary psychology, 48-49, 60-63, Jesus of Nazareth, alleged divinity of, 171- 70, 133 210 Ewing, A. C., 130 Joad, C. E. M., 6, 30, 43 Job (Biblical book), 25, 225, 236, 246 Ficino, Marsilio, 2 Johnson, David Kyle, 18-19 Feser, Edward, 108. 116 Jordan, Matt, 156 Flacius, Matthias, 220 Joy, as alleged desire for God or transcendence, 3-4, 27ff., 51-53 Flew, Antony, 1 Joyce, Richard, 133, 159 Fodor, Jerry, 87 Justin Martyr, 227 Four Loves, The, 211 Francis of Assisi, 225 Kant, Immanuel, 2, 10, 19, 76, 121 Franklin, Benjamin, 74 Kasparov, Garry, 97. 98 Keats, John, 222 Gaunilo, 230 Keller, Tim, 39 Gonzalez, Luis, 87, 109, 117 Kierkegaard, Sören, 11 Gordon, Jeffrey, 68 Knox, Ronald, 21 Grayling, A. C., 41 Kreeft, Peter, 21, 38, 41, 52, 62, 176, 182 Great Divorce, The, 217 Grief Observed, A, 23, 211, 212, 233, Last Battle, The, 217 235, 241, 244, 247 Layman, C. Stephen, 58 Groothuis, Douglas, 59 Leibniz, Gottfried, 168 Lewis, Warnie, 3 Habermas, Gary, 38-39 Liddon, H. P., 21 Linville, Mark, 19, 44, 159 Haldane, J. B. S., 6 Lion, Witch, and the Wardrobe, The, 228 Haldane, John, 27, 34, 38, 62, 63 Locke, John, 12 Hare, R. M., 54 Lovell, Steve, 39-40, 42, 50, 53, 134 Harman, Gilbert, 159, 160 Hasker, William, 6, 17, 78, 88, 105 McDowell, Josh, 21 hell, and the problem of evil, 217 MacDonald, George, 217 Hick, John, 174, 222, 238 McGrath, Alister, 27, 28, 31, 35-36, 44, Hinten, Marvin, 173, 180, 182 47, 60, 179 Holyer, Robert, 27, 63 Mackie, J. L., 20-21 Horner, David, 182 Marks, Joel, 133 Hooker, Richard, 2 Maslow, Abraham, 62 Howard-Snyder, Daniel, 22, 180-182, Mavrodes, George, 139, 154 183-187, 192 Melnyk, Andrew, 107, 115 Mere Christianity, 1, 3, 19-20, 37, 77, 119, inferential (inductive) argument from 130, 169-170, 172, 174, 177, desire, 34, 37-40, 48-51, 70-71 185, 186, 197, 207, 210, 228, intentionality, 10, 18, 77, 78, 79, 84-85, 236 97, 99, 100, 103, 108, 112 Milton, John, 199 Irenaeus, 216 mind-body identity theory, 18, 94, 97, Is Theology Poetry?”, 5 106-107, 113-114, 115 Miracles, 1, 5, 7, 9-10, 11, 76, 77, 80, 91, James, William, 50, 72 105, 129, 175, 211 moral argument, the, for God’s existence, 121-170 Index 271

Moreland, J. P., 36-37, 38, 39, 58, 60, 61, Rist, John, 139 129 Robinson, John, 196 Morison, Frank, 186 robust normative realism, 20, 147-148, Morris, Tom, 37-38, 62, 67, 138 166, 167 Murray, Michael, 224 Roosevelt, Franklin, 1 Rorty, Richard, 234 Nagel, Thomas, 129, 133 Rowe, William, 218 Newman, John Henry, 19, 121 Ruse, Michael, 20 Newton, Isaac, 168, 209 Russell, Bertrand, 134

Ockham’s razor, 31, 193 Scarry, Elaine, 31 O’Hara, David, 25 Schaeffer, Francis, 67 O’Hear, Anthony, 59 Scruton, Roger, 59 “On Living in the Atomic Age,” 5 Sehnsucht, 3, 28 O’Rourke, Meghan, 212 Sheiman, Bruce, 31 Out of the Silent Planet, 228 Singer, Peter, 151 skeptical theism, 218-219 Parmenides, 11 Smith, John, 2-3 Pascal, Blaise, 30, 204, 209, 234 Socrates, 74, 234, 235 pastoral theology, and the problem of evil, Socratic Club, 1, 7, 91 233-234, 241-242 Sorley, W. R., 18 Peels, Rik, 58 soul-making theodicy, the, 25, 211, 220- Peirce, C. S., 231, 234, 246 224 Pilgrim’s Regress, The, 1, 3, 28-29, 40, Spectator, The, 2 51, 64 Strauss, Mark, 192 Pinnock, Clark, 57 supervenience, 18, 82, 88, 95, 98, 100- Plantinga, Alvin, 6, 11, 33-34, 53, 68, 138, 101, 102, 106-107, 114-115, 162 148, 214, 237, 244, 245 Surprised by Joy, 3, 28, 34, 213 Plato, 14, 76, 134, 136, 139, 140, 228, Swinburne, Richard, 33, 46, 57-59, 69-70, 232, 235 76, 199, 205, 215, 217, 219, 237, “Poison of Subjectivism, The,” 134, 135, 238 141 prima facie argument from desire, 31-34, Tacelli, Ronald, 21, 175 37, 46, 57-60, 71-72 Tallon, Philip, 25 principle of credulity (Swinburne), 33, 37, Taylor, A. E., 19 48, 57-58, 68-70, 73 Taylor, Charles, 59 problem of evil, Lewis’s response to, 1, Till We Have Faces, 3 23-25, 211-248 Time magazine, 1 Problem of Pain, The, 1, 23-25, 211-248 Tolkien, J. R. R., 34, 51, 63, 225, 247 Property dualism, and naturalism, 95-96, transcendental arguments, 10-11 99-100, 107-109, 115-116 trilemma argument, for Christ’s divinity, Purtill, Richard, 41, 175 20-23, 171-210 Turing test, 118 Rana, Fazale, 65 Rashdall, Hastings, 19 validity, and the argument from reason, 7- reductio argument from desire, 41-44, 53- 9 54, 66-68, 72-73 Reppert, Victor, 6, 18-19, 36, 60, 62 Watson (computer), 97, 103, 117, 118 Ritchie, Angus, 19, 159 Weatherhead, Leslie, 42 272 C. S. LEWIS’S CHRISTIAN APOLOGETICS

“Weight of Glory, The,” 3, 30, 41, 60 Wells, H. G., 41 Wesley, John, 221 Wielenberg, Erik, 20-21, 50, 53, 61-62, 63, 70, 124, 131-132, 134-135, 137 Williams, Bernard, 1 Williams, Donald T., 23 Williams, Peter S., 5, 21 Wind in the Willows, The, 28 Witherington, Ben, 192 Wright, N. T., 171, 173, 174, 177, 192 Wykstra, Steve, 218

Young, Frances, 174, 177

Zangwill, Nick, 149-150