William L. Rowe on philosophy of religion : selected writings 9781351872812, 1351872818

Table of contents :
Acknowledgements --
Introduction --
I: The Philosophical Theology of Paul Tillich --
1. Tillich's Concept of God --
2. Religious Symbols --
II: The Problem of Evil --
3. The Problem of Evil and Some Varieties of Atheism --
4. The Empirical Argument from Evil --
5. Evil and Theodicy --
6. Paradox and Promise: John Hick's Solution to the Problem of Evil --
7. Ruminations about Evil --
8. William Alston on the Problem of Evil --
9. The Evidential Argument from Evil: A Second Look --
10. In Defense of "The Free Will Defense" --
11. Grounds for Belief in God Aside, Does Evil Make Atheism More Reasonable than Theism? --
12. Friendly Atheism, Skeptical Theism, and the Problem of Evil --
III: Replies to Critics on the Evidential Problem of Evil --
13. Evil and the Theistic Hypothesis: A Response to Wykstra --
14. Response to Lynda Zagzebski 15. Reply to Plantinga --
16. Skeptical Theism: A Response to Bergmann --
17. Reply to Howard-Snyder and Bergmann --
IV: Divine Freedom --
18. The Problem of Divine Perfection and Freedom --
19. Evil and God's Freedom in Creation --
20. Can God Be Free? --
V: The Cosmological Argument --
21. Two Criticisms of the Cosmological Argument --
22. The Cosmological Argument --
23. Cosmological Arguments --
VI: The Ontological Argument --
24. The Ontological Argument --
25. Modal Versions of the Ontological Argument --
VII: Other Writings in the Philosophy of Religion --
26. Augustine on Foreknowledge and Free Will --
27. Religious Experience and the Principle of Credulity --
28. The Rationality of Religious Belief --
29. Religion within the Bounds of Naturalism: Dewey and Wieman --
30. Religious Pluralism --
William L. Rowe: A Bibliography

Citation preview

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William Rowe is one of the leading thinkers in contemporary philosophy of religion. Although he is best known for his contributions to the problem of evil, he has produced innovative and influential work across a wide array of subjects at the interface between philosophy and religion. He has, for example, written extensively on the existentialist theologian, Paul Tillich, on the challenging problem of divine freedom, and on the traditional arguments in support of the existence of God. His work in these areas is distinguished by its clarity, rigour, originality, and sensitivity towards the claims of his theistic opponents. Indeed, Rowe's work has played a pivotal role in the remarkable revival of analytic philosophy of religion since the 1970s. The present collection brings together for the first time Rowe's most significant contributions to the philosophy of religion. This diverse but representative selection of Rowe's writings will provide students, professional scholars as well as general readers with stimulating and accessible discussions on such topics as the philosophical theology of Paul Tillich, the problem of evil, divine freedom, arguments for the existence of God, religious experience, life after death, and religious pluralism.

This page intentionally left blank

ASHGATE CONTEMPORARY THINKERS ON RELIGION: COLLECTED WORKS General Editor: John Hinnells

William L. Rowe on Philosophy of Religion

ASHGATE CONTEMPORARY THINKERS ON RELIGION: COLLECTED WORKS General Editor: John Hinnells Other titles in this series: Zoroastrian and Parsi Studies: Selected Works of John R. Hinnells JOHN R. HINNELLS, University of Derby, UK ISBN 978-0-7546-1501-9 Neusner on Judaism: Volume 1: History JACOB NEUSNER, Bard College, New York, USA ISBN 978-0-7546-3598-7 Williams on South Asian Religions and Immigration: Collected Works RAYMOND BRADY WILLIAMS, Wabash College, USA ISBN 978-0-7546-3856-8 Beck on Mithraism: Collected Works with New Essays ROGER BECK, University of Toronto at Mississauga, Canada ISBN 978-0-7546-4081-3 Neusner on Judaism: Volume 2: Literature JACOB NEUSNER, Bard College, New York, USA ISBN 978-0-7546-3599^ Neusner on Judaism: Volume 3: Religion and Theology JACOB NEUSNER, Bard College, New York, USA ISBN 978-0-7546-3600-7 Drummer on Meaning and the Christian Faith: Collected Writings of Vincent Brummer VINCENT BRUMMER, University of Utrecht, The Netherlands ISBN 978-0-7546-4028-8 Thiselton on Hermeneutics: The Collected Works and New Essays of Anthony Thiselton ANTHONY C. THISELTON, University of Nottingham and University of Chester, UK ISBN 978-0-7546-3925-1 Ninian Smart on World Religions: Selected Works Edited by JOHN J. SHEPHERD, St Martin's College, UK ISBN 978-0-7546-4080-6

ASHGATE CONTEMPORARY THINKERS ON RELIGION: COLLECTED WORKS General Editor: John Hinnells

William L. Rowe on Philosophy of Religion Selected Writings

William L. Rowe Purdue University, USA Edited by Nick Trakakis Monash University, Australia

1

3 Routledge ^ Taylor & Francis Croup

LONDON AND NEW YORK

First published 2007 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business Copyright © Nick Trakakis 2007. For copyright of individual articles refer to the Acknowledgements. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data Rowe, William L., 1931William L. Rowe on philosophy of religion : selected writings. - (Ashgate contemporary thinkers on religion: collected works) 1. Religion - Philosophy I. Title II. Trakakis, Nick 210 ISBN 978-07546-5558-9 Library of Congress Cataloging-in-Publication Data Rowe, William L., 1931William L. Rowe on philosophy of religion : selected writings / William L. Rowe; edited by Nick Trakakis. p. cm. - (Ashgate contemporary thinkers on religion) "William L. Rowe: a bibliography"- P. Includes bibliographical references and index. ISBN 978-0-7546-5558-9 (hardcover: alk. paper) 1. Religion - Philosophy. I. Trakakis, Nick. II. Title. III. Title: On philosophy of religion. BL51.R5993 2007 210-dc22 2006030604 ISBN 9780754655589 (hbk)

Contents Acknowledgements Introduction

ix xiii

I: The Philosophical Theology of Paul Tillich 1 2

Tillich's Concept of God Religious Symbols

3 35

II: The Problem of Evil 3 4 5 6 7 8 9 10 11 12

The Problem of Evil and Some Varieties of Atheism The Empirical Argument from Evil Evil and Theodicy Paradox and Promise: Hick's Solution to the Problem of Evil Ruminations about Evil William Alston on the Problem of Evil The Evidential Argument from Evil: A Second Look In Defense of 'The Free Will Defense' Grounds for Belief Aside, Does Evil Make Atheism More Reasonable than Theism? Friendly Atheism, Skeptical Theism, and the Problem of Evil

61 69 91 105 119 139 163 187 193 207

III: Replies to Critics on the Evidential Problem of Evil 13 14 15 16 17

Evil and the Theistic Hypothesis: A Response to Wykstra Response to Linda Zagzebski Reply to Plantinga Skeptical Theism: A Response to Bergmann Reply to Howard-Snyder and Bergmann

223 229 233 241 249

IV: Divine Freedom 18 19 20

The Problem of Divine Perfection and Freedom Evil and God's Freedom in Creation Can God Be Free?

255 267 281

V: The Cosmological Argument 21

Two Criticisms of the Cosmological Argument

303

viii 22 23

CONTENTS The Cosmological Argument Cosmological Arguments

323 337

VI: The Ontological Argument 24 25

The Ontological Argument Modal Versions of the Ontological Argument

353 363

VII: Other Writings in the Philosophy of Religion 26 27 28 29 30

Augustine on Foreknowledge and Free Will Religious Experience and the Principle of Credulity The Rationality of Religious Belief Religion within the Bounds of Naturalism: Dewey and Wieman Religious Pluralism

371 379 387 401 421

William L. Rowe: A Bibliography

433

Index

439

Acknowledgements The chapters in this volume are reproduced from the sources listed below. Grateful acknowledgement is made to the copyright holders of these works for permission to reproduce them in this volume. 1. "Tillich's Concept of God," in William L. Rowe, Religious Symbols and God: A Philosophical Study of Tillich's Theology (Chicago, IL: The University of Chicago Press, 1968), pp. 11^2. © 1968 The University of Chicago. Used by permission of the publisher, The University of Chicago Press. 2. "Religious Symbols," in William L. Rowe, Religious Symbols and God: A Philosophical Study of Tillich's Theology (Chicago, IL: The University of Chicago Press, 1968), pp. 171-94. © 1968 The University of Chicago. Used by permission of the publisher, The University of Chicago Press. 3. "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16 (1979): 335^1. © 1979 North American Philosophical Publications Inc. 4. "The Empirical Argument from Evil," in Robert Audi and William J. Wainwright (eds), Rationality, Religious Belief, and Moral Commitment: New Essays in the Philosophy of Religion (Ithaca, NY: Cornell University Press, 1986), pp. 227-47. Reproduced with kind permission of the editors, Robert Audi and William J. Wainwright. 5. "Evil and Theodicy," Philosophical Topics 16 (1988): 119-32. © 1988 University of Arkansas Press. 6. "Paradox and Promise: Hick's Solution to the Problem of Evil," in Harold Hewitt, Jr (ed.), Problems in the Philosophy of Religion: Critical Studies of the Work of John Hick (London: Macmillan, 1991), pp. 111-24. Reproduced with permission of Palgrave Macmillan. 7. "Ruminations about Evil," Philosophical Perspectives, 5, Philosophy of Religion, 1991, edited by James E. Tomberlin, pp. 69-88. Reprinted by permission of Ridgeview Publishing Company. 8. "William Alston on the Problem of Evil," in Thomas D. Senor (ed.), The Rationality of Belief and the Plurality of Faith: Essays in Honor of William P. Alston (Ithaca, NY: Cornell University Press, 1995), pp. 71-93. © 1995 by Cornell University. Used by permission of the publisher, Cornell University Press.

X

ACKNOWLEDGEMENTS

9. "The Evidential Argument from Evil: A Second Look," in Daniel Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington, IN: Indiana University Press, 1996), pp. 262-85. Reproduced with permission of Indiana University Press, Bloomington & Indianapolis. 10. "In Defense of 'The Free Will Defense'," International Journal for Philosophy of Religion 44 (1998): 115-20. © 1998 Kluwer Academic Publishers, reprinted with kind permission of Springer Science and Business Media. 11. "Grounds for Belief Aside, Does Evil Make Atheism More Reasonable than Theism?/' in William Rowe (ed.), God and the Problem of Evil (Maiden, MA: Blackwell, 2001), pp. 124r-37. Reproduced with permission of Blackwell Publishing. 12. "Friendly Atheism, Skeptical Theism, and the Problem of Evil," International Journal for Philosophy of Religion 59 (2006): 79-92. © 2006 Kluwer Academic Publishers, reprinted with kind permission of Springer Science and Business Media. 13. "Evil and the Theistic Hypothesis: A Response to Wykstra," International Journal for Philosophy of Religion 16 (1984): 95-100. © 1984 Martinus Nijhoff Publishers, reprinted with kind permission of Springer Science and Business Media. 14. "Response to Linda Zagzebski," in Harold Hewitt, Jr (ed.), Problems in the Philosophy of Religion: Critical Studies of the Work of John Hick (London: Macmillan, 1991), pp. 130-33. Reproduced with permission of Palgrave Macmillan. 15. "Reply to Plantinga," Nous 32 (1998): 545-52. Reproduced with permission of Blackwell Publishing. 16. "Skeptical Theism: A Response to Bergmann," Nous 35 (2001): 297-303. Reproduced with permission of Blackwell Publishing. 17. "Reply to Howard-Snyder and Bergmann," in William Rowe (ed.), God and the Problem of Evil (Maiden, MA: Blackwell, 2001), pp. 155-58. Reproduced with permission of Blackwell Publishing. 18. "The Problem of Divine Perfection and Freedom," in Eleonore Stump (ed.), Reasoned Faith: Essays in Philosophical Theology in Honor of Norman Kretzmann (Ithaca, NY: Cornell University Press, 1993), pp. 223-33. © 1993 by Cornell University. Used by permission of the publisher, Cornell University Press. 19. "Evil and God's Freedom in Creation," American Philosophical Quarterly 36 (1999): 101-13. Copyright © 1999 North American Philosophical Publications Inc.

ACKNOWLEDGEMENTS

xi

20. "Can God Be Free?/' Faith and Philosophy 19 (2002): 405-24. © 2002 The Society of Christian Philosophers. 21. "Two Criticisms of the Cosmological Argument," The Monist 54 (1970): 441-59. Copyright © 1970 The Monist: An International Quarterly Journal of General Philosophical Inquiry, Peru, Illinois, U.S.A. 61354. Reprinted by permission. 22. "The Cosmological Argument," Nous 5 (1971): 49-61. Reproduced with permission of Blackwell Publishing. 23. "Cosmological Arguments," in William E. Mann (ed.), The Blackwell Guide to the Philosophy of Religion (Oxford: Blackwell, 2005), pp. 103-116. Reproduced with permission of Blackwell Publishing. 24. "The Ontological Argument," in Joel Feinberg (ed.), Reason and Responsibility: Readings in Some Basic Problems of Philosophy, third edition (Encino and Belmont, CA: Dickenson Publishing Company, 1975), pp. 8-17. © 1975, reprinted with permission of Wadsworth, a division of Thomson Learning: www.thomsonrights.com. Fax 800 730-2215. 25. "Modal Versions of the Ontological Argument," in Louis P. Pojman (ed.), Philosophy of Religion: An Anthology, first edition (Belmont, CA: Wadsworth Publishing Company, 1987), pp. 69-73. © 1987, reprinted with permission of Wadsworth, a division of Thomson Learning: www.thomsonrights.com. Fax 800 730-2215. 26. "Augustine on Foreknowledge and Free Will," The Review of Metaphysics 18 (1964): 356-63. © 1964 The Review of Metaphysics. 27. "Religious Experience and the Principle of Credulity," International Journal for Philosophy of Religion 13 (1982): 85-92. © 1982 Martinus Nijhoff Publishers, reprinted with kind permission of Springer Science and Business Media. 28. "The Rationality of Religious Belief," Contemporary Philosophy 14 (1992): 3-9. © 1992 REALIA. 29. "Religion within the Bounds of Naturalism: Dewey and Wieman," International Journal for Philosophy of Religion 38 (1995): 17-36. © 1995 Kluwer Academic Publishers, reprinted with kind permission of Springer Science and Business Media. 30. "Religious Pluralism," Religious Studies 35 (1999): 139-50. © 1999 Cambridge University Press.

This page intentionally left blank

Introduction William Leonard Rowe is widely regarded as one of the most important philosophers of religion today, having played a leading role in the remarkable revival in philosophy of religion since the 1970s, at least as this discipline is practised in the Anglo-American analytic tradition. In a recent collection of essays dedicated to Rowe, editors Jeff Jordan and Daniel Howard-Snyder note that, "At mid-century philosophical critics dismissed philosophy of religion as an uninteresting area of which Hume and Kant had already had the last say/'1 But, as even the most causal observer can attest, things are now vastly different: Today philosophy of religion is a vibrant and exciting area of philosophical research with four academic journals and at least three professional societies devoted to the field... Indeed, many of the most influential philosophers at work today publish regularly on topics in this particular field.2

Jordan and Howard-Snyder do not omit to add that, "A good part of the reason that philosophy of religion is today a vivacious area is due to the work of Bill Rowe/'3 The present collection, it is hoped, supports this view by providing a representative sample of the many insightful and influential contributions Rowe has made during his long and distinguished career. William Rowe was born on 26 July 1931 in Detroit, Michigan. Although he is now best known for his powerful philosophical challenges to theistic belief, Rowe in fact began his journey as an evangelical Christian. After a conversion experience at the age of sixteen, he decided to devote his life to Christian service, and therefore enrolled at the Detroit Bible Institute after completing his secondary schooling. But, disappointed with the Bible Institute's handling of the dissenting views of his favourite teacher, Rowe soon left for Wayne State University. He found a firm sparring partner in one of his philosophy teachers at Wayne State, George Nakhnikian, an atheist who would engage in hours of dialogue with his religiously-conservative student, thus forging a mentoring relationship that developed into a lifelong friendship.4 Rowe was awarded the Bachelor of Arts degree (with distinction) from Wayne State University in 1954, and then moved to Chicago to study for a Bachelor of Divinity degree at the Chicago Theological Seminary. The critical and liberal environment at the Chicago seminary made a great impact on Rowe's fundamentalist faith. No longer confident of the existence of a supernatural being who created the world, Rowe became interested in the work of two liberal theologians, Henry Nelson Wieman and Paul Tillich. In his Bachelor of Divinity thesis Rowe examined 1 Jeff Jordan and Daniel Howard-Snyder (eds), Faith, Freedom, and Rationality: Philosophy of Religion Today (Lanham, MD: Rowman & Littlefield Publishers, 1996), p. ix. 2 Ibid. 3 Ibid., p. xiv. 4 See Nakhnikian's account of this relationship in Jordan and Howard-Snyder (eds), Faith, Freedom, and Rationality, pp. xiii-xiv.

xiv

INTRODUCTION

the naturalistic conception of God developed by Wieman, but unconvinced that such a conception could form the focus of religious commitment, Rowe turned to the then immensely popular figure of Paul Tillich. (Tillich even appeared on the cover of Time magazine on March 16,1959.) Having met Tillich in Chicago and still in search of some kind of divine reality in which he could believe, Rowe enrolled at the University of Michigan to study with William P. Alston. At Michigan, Rowe wrote his doctoral dissertation on the philosophical theology of Paul Tillich (the dissertation later forming the basis of his first published book) and was awarded the PhD in 1962. By this time, Rowe had come a long way from his early Protestant fundamentalism, but his route to atheism was a slow and very personal one, informed not so much by any specific philosophical or scientific arguments, but (as he put it in a recent interview) by "the lack of experiences and evidence sufficient to sustain my religious life and my religious convictions."5 In 1962, Rowe joined the relatively young philosophy faculty at Purdue University, where he became Professor of Philosophy in 1969 and served as Head of the Philosophy Department from 1981 to 1991. He has also held visiting appointments at Wayne State University (1963-64) and the University of Michigan (1970), and has held fellowships from the National Endowment for the Humanities, the National Humanities Centre, the Institute for Advanced Studies in the Humanities (at the University of Edinburgh), and the Guggenheim Foundation. Further, in 1986-87 he was elected to the prestigious position of President of the Central Division of the American Philosophical Association. After a 43-year-long teaching career at Purdue, during which time he authored five books, edited two volumes and published over seventy journal papers, Rowe retired in 2005 (with a three-day conference at Purdue on the problem of evil held in his honour). This collection of Rowe's writings is divided into seven broad sections, the first six covering a specific area within the philosophy of religion where Rowe has made a significant contribution, and the final section encompassing a range of other topics that Rowe has explored in his philosophical study of religion. Section I presents two excerpts from Rowe's study of Tillich's philosophical theology taken from Rowe's first, and well-received, monograph, Religious Symbols and God, published in 1968. The first excerpt is from the opening chapter of the book, where Rowe takes on the difficult task of clarifying the content of Tillich's dark sayings that "God is that which concerns man ultimately" and "God is being-itself". The problem of interpreting such statements in non-symbolic (or literal) terms is again taken up in the second excerpt (culled from the first half of chapter 6 of Rowe's book), where Rowe uncovers the fundamental problem in Tillich's theory of religious symbolism: it appears impossible on Tillich's theory to break out of the web of symbols. Section II contains ten papers spanning more than two decades on the subject of the problem of evil, the problem of reconciling the existence of a loving God with the innumerable evils littering our planet. Rowe's contribution to this debate is undoubtedly the finest moment in his philosophy of religion. His 1979 paper, "The Problem of Evil and Some Varieties of Atheism", has become a widely anthologised classic, spurring 5 See Nick Trakakis, "Interview with William Rowe/' Philosophy Now, issue 47, August-September 2004, p. 16.

INTRODUCTION

xv

a series of counter-arguments and theistic defences or theodicies. Rowe's post-1979 work on the problem of evil is largely an attempt to meet some of the more forceful of these objections, and in the process he has modified, sometimes in quite radical ways, his original formulation of the evidential argument from evil. Also included in Section II, however, is a paper in support of a theistic defence against the problem of evil, more specifically, the so-called 'logical problem of evil', which (in Rowe's view) can be resolved by appeal to Plantinga's famous 'free will defence'. The interplay between Rowe and his critics on the problem of evil is continued in Section III, which includes five short replies Rowe has made to criticisms of his evidential argument(s). Interestingly, most of these criticisms revolve around Rowe's rejection of 'sceptical theism', the idea that even if God did have good reasons for permitting evil, we should not expect to know what these reasons are. Although the papers Rowe is responding to have not been included, the care that Rowe takes in fairly explicating his opponent's position goes some way in making up for this omission. Also, to overlook these short rejoinders of Rowe's would be to miss out on many of the innovative turns his thinking has taken on the relationship between God and evil. Section IV offers three of Rowe's papers on the topic of divine freedom, which in many ways are forerunners to Rowe's recently published book, Can God Be Free?. As these papers indicate, Rowe's 'natural atheology' draws not only upon evidential arguments from evil, but also upon attacks on the coherence of the theistic worldview. Specifically, Rowe contends that the attribution of libertarian free will and moral unsurpassibility to God results in an incoherent conception of the deity. He argues, for example, that if there is no single, morally best world that God could create (so that for any world creatable by God there is another creatable world that is morally better), then it would be possible for there to be a being morally better than God - in which case, God is not (essentially) morally unsurpassable. On the other hand, if there is a single, morally best world God could create, then God qua morally perfect being must create that world - in which case, God lacks the freedom not to create the world in question. As William Hasker succinctly puts the matter, "According to William Rowe, if God is unsurpassably good He cannot be free; if He is free [i.e., free to select from an endless number of creatable worlds], He cannot be unsurpassably good."6 Rowe thus attempts to cast doubt on the very coherence and hence the plausibility of the traditional theistic view of God. In addition to advancing arguments against theism, Rowe has also published many historically-informed but also philosophically subtle critiques of the traditional arguments in support of theism, particularly the cosmological and ontological arguments. Section V provides a selection of Rowe's work on the cosmological argument. However, the first essay in this selection, entitled "Two Criticisms of the Cosmological Argument" (later incorporated into chapter 3 of Rowe's 1975 book-length study, The Cosmological Argument), is actually a defence of a version of the cosmological argument adapted from Samuel Clarke. More precisely, Rowe defends the argument from two standard criticisms: firstly, that it makes no sense to apply the notion of 'cause' or 'explanation' to the totality of 6

453.

William Hasker, "Can God Be Free?: Rowe's Dilemma for Theology/' Religious Studies 41 (2005):

xvi

INTRODUCTION

things, and secondly, the Humean view that in giving an explanation for every single member of a collection, one has thereby given an explanation for the collection itself. By contrast, the second paper in Section V, entitled "The Cosmological Argument" and originally presented as the lead symposium paper at the Western Division of the American Philosophical Association in 1971, takes a more negative view of the cosmological argument in the light of difficulties in establishing the truth of the Principle of Sufficient Reason. The third selection, a survey article commissioned for The Blackwell Guide to the Philosophy of Religion, provides a nice summary of Rowe's current assessment of various cosmological arguments. Still sceptical of the ability of such arguments to substantiate the existence of the theistic God, Rowe ends his article on the somewhat surprising note that "it must be acknowledged that the emergence of the Big Bang theory of the origin of the universe has given new weight to an argument for the existence of some sort of creator." Section VI turns to Rowe's critique of the ontological argument, or some of the many versions of that argument, both old and new. In "The Ontological Argument," which first appeared in the third edition of Joel Feinberg's collection, Reason and Responsibility (published in 1975), Rowe provides a perspicuous reconstruction of Anselm's Proslogium II argument, critically surveys several standard objections, and then offers a unique objection of his own. The following selection, originally commissioned for the first (1987) edition of Louis Pojman's Philosophy of Religion: An Anthology, turns to modal ontological arguments inspired by Proslogium III, such as that offered by Plantinga. These modal arguments, Rowe concedes, are not obviously fallacious or unsound, though they fall short of Plantinga's aim of establishing the rational acceptability of theism. Section VII presents a selection of Rowe's less well-known, but no less interesting, papers on a variety of topics within the philosophy of religion. The opening paper is an early piece by Rowe, published in a 1964 issue of The Review of Metaphysics, which takes as its theme the infamous problem of the compatibility of divine foreknowledge and human free will. Rowe outlines an argument, drawn from Augustine's treatise On Free Will, against the compatibility of foreknowledge and free will, then rejects Augustine's response to that argument, and finally offers a diagnosis as to where the argument goes astray. The second piece provides an excellent discussion of theistic arguments from religious experience that depend upon 'the principle of credulity' - the principle that our experiences should be treated, epistemically speaking, innocent until proven guilty (so that one's experience of x should be treated as veridical unless there is good reason to think otherwise). Rowe argues that such arguments fail because the principle of credulity, when suitably recast, cannot be applied to religious experience as we lack criteria for distinguishing veridical from delusive religious experiences. The epistemic value of religious experience also looms large in "The Rationality of Religious Belief." Rowe there examines the contention of Reformed epistemologists that belief in God can be 'properly basic', that is, entirely rational even if not held on the basis of arguments or (propositional) evidence. Although Rowe concedes that belief in God is a basic belief for an enormous number of people, he argues that theistic beliefs cannot rationally be held in the basic way, at least by a mature, questioning theist. As before, Rowe highlights the absence of criteria for separating genuine from counterfeit religious

INTRODUCTION

xvii

experiences, and argues that this creates an important disanalogy between sensory experience and religious experience, thus preventing experiences of the latter sort (but not of the former sort) in rationally justifying the basic beliefs they produce. In "Religion within the Bounds of Naturalism: Dewey and Wieman," Rowe looks at some attempts at rendering belief in God compatible with naturalism, the central idea of which is the thesis that all that exists is either physical or causally dependent on physical things. On the face of it, no two ideas could be further apart, but Rowe shows how John Dewey and Henry Nelson Wieman attempted to forge a naturalistic conception of God and religious faith. In the end, however, Rowe concludes that these naturalistic theologies are religiously inadequate. Dewey's God - defined as "the unity of all ideal ends arousing us to desire and actions", or whatever natural forces support these ideal ends - is ultimately a human creation, and so lacks creative power. And Wieman's God - identified as a force or process of 'creative interchange' operative in human communities and leading to the best humans can achieve if only they submit to its workings in their lives - may be capable of exerting creative power, and may even be worthy of our care and attention, but it is not clear how it could command our love, devotion and commitment. The collection closes with Rowe's elucidation and critique of John Hick's defence of 'religious pluralism'. According to Hick, the world's great religious traditions have not only been equally fruitful in producing saintliness, but they can also be seen as conceptualising in diverse but equally valid ways the human encounter with the transcendent divine Reality. Hick adds, however, that the Real in itself is ineffable insofar as the nature of the Real is beyond the scope of the network of human concepts. Rowe takes issue with the claim to ineffability, arguing that there is no way to avoid ascribing to the Real one or the other of two contradictory properties - e.g., the Real is either personal or non-personal, good or non-good, etc. As is evidenced by the Bibliography at the end of the volume, there is much more in Rowe's philosophy of religion, and philosophy in general, than what is included here. But the present selection will have served its purpose if it has highlighted the originality, clarity, rigour, and sensitivity toward religious faith that has become the hallmark of Rowe's rich corpus in the philosophy of religion.

This page intentionally left blank

PARTI THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

This page intentionally left blank

CHAPTER ONE

Tillich's Concept of God

TTJAUL TILLICH'S doctrine of God is at once both reliJLgious and philosophical. The religious aspect is expressed by the statement "God is that which concerns man ultimately* (ST, 1:211). The philosophical aspect of his doctrine of God is expressed by the statement "God is being-itselT (ST, 1:235). The task of elucidating Tillich's concept of God is largely one of explaining these two basic theistic statements. It is best, I believe, to begin with the statement "God is that which concerns man ultimately/' The key notion in the statement expressing the religious aspect of Tillich's doctrine of God is 'ultimate concern/ Tfllich describes ultimate concern as the abstract translation of the great commandment: "The Lord, our God, the Lord is one; and you shall love the Lord your God with all your heart, and with all your soul, and with all your mind, and with all your strength" (ST, 1:11). From this and other remarks he makes, it would seem that ultimate concern is roughly equivalent to 'utter devotion* or 'complete commitment/ 'Devotion* or 'commitment* expresses the element of love mentioned in the commandment, litter* or 'complete* embraces two ideas. First, it means that the

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

4

commitment is not divided between God and something else; it is not partial. Second, as expressed in the commandment, the commitment is complete in the sense that all man's dimensions participate in the commitment—the mind, the heart, the soul, etc. Tillich elsewhere expresses this by saying that ultimate concern is a "centered act/' "Faith as ultimate concern is an act of the total personality. It happens in the center of the personal life and includes all its dimensions.... It is not a movement of a special section or a special function of man's total being. They are all united in the act of faith."1 Thus to be ultimately concerned about x is to regard x as infinitely important, to be completely committed to x. However, the notion of complete commitment does not quite capture all that Tillich packs into the concept of ultimate concern. To be ultimately concerned about x is not simply to be deeply committed to x; it is also to experience x as holy. "What concerns one ultimately becomes holy" (DF, 12-13). What is it to experience or regard x as holy? Refusing to identify holiness with moral perfection, Tillich turns to Rudolf Otto's The Idea of the Holy, pointing out that for Otto the awareness of the holy is the awareness of a presence "which remains mysterious in spite of its appearance, and it exercises both an attractive and a repulsive function on those who encounter it" (DF, 13). The holy object is experienced with awe, mystery, dread, and yet one is fascinated by it, irresistibly attracted to it. Tillich contends that these feelings are not peculiar to some one religion, but can be found in i Paul Tillich, Dynamics of Faith (New York: Harper & Bros., 1957), p. 4. Hereafter references to this work will appear in the text abbreviated asDF. 12

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

5

all religions ". . . because they are the way in which man always encounters the representation of his ultimate concern" (DF, 13). To be ultimately concerned about x, then, is not only to be absolutely committed to x, but to experience x as divine or holy (in the sense explained above). Unless I am mistaken, ultimate concern about x as including absolute commitment to x and certain feelings about x (awe, mystery, etc.), summed up under "experiencing x as holy or divine," is for Tillich a necessary and sufficient condition for a: to be a sacred object (religious symbol). I see nothing essentially wrong with this procedure. To identify an object as a religious object is not to identify some property or properties that the object possesses independently of the way in which men react to that object. An object becomes sacred by virtue of becoming the content of ultimate concern. What must God be for Tillich if He is that about which men are ultimately concerned? The appropriate answer would seem to be "Almost anything." For, given our explication of 'ultimate concern/ it seems reasonable to believe that almost anything could be an object of ultimate concern. At least it is true that human beings and movements (e.g., communism) qualify as objects of ultimate concern. But this creates a difficulty in our attempt to understand what 'God' means for Tillich. For Tillich wishes to equate God with whatever is the object of man's ultimate concern. But, as we shall see, Tillich is not satisfied with a view which ultimately identifies God with a human being or a political movement. This means either (a) Tillich is mistaken in speaking of God as that about which we are ultimately concerned, or (b) he must have a more complex view than I suggested of what it is to be ultimately 13

6

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

concerned. The second point b is the more fruitful alternative to explore. Point b allows several possibilities: (1) Tillich could have a special sense of 'ultimate concern' which is such that one is ultimately concerned only about being-itself. The point here is that 'ultimate concern* would be so defined that €x is ultimately concerned' would entail *x is ultimately concerned about being-itself/ (2) Tillich could be holding that, while we can be ultimately concerned about almost anything, the only satisfactory or appropriate object of ultimate concern is being-itself. If this is his view, then God need not be the only object of ultimate concern. (3) Tillich could be maintaining that ultimate concern— which is defined in terms of certain feelings and attitudes focused on specific objects—is basically directed toward being-itself. Point 3 differs from 1 in that it makes the connection between ultimate concern and being-itself a metaphysical connection, rather than an analytic necessity following from a special definition of 'ultimate concern/ It is quite difficult to determine which of these (if any) is Tillich's view. Point 2, more than 1 or 3, fits in best with some of Tillich's statements about idolatry. However, there are passages strongly suggesting 1 or possibly 3: ... for that which concerns us ultimately must belong to reality as a whole; it must belong to being. Otherwise we could not encounter it, and it could not concern us. Of course, it cannot be one being among others; then it would not concern us infinitely. It must be the ground of our being, that which determines our being or not-being, the ultimate and unconditional power of being (ST, 1:21).

Here Tillich seems to be denying that we can be ultimately concerned about anything less than the ultimate, being14

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

7

itself. ('Ground of being/ 'power of being* are both synonymous with 'being-itself [ST, 1:236].) Anything less than ultimate can be a matter of concern for us, but not of ultimate concern. The clearest expression of this point is as follows: The unconditional concern . . . is the concern about the unconditional. The infinite passion . . . is the passion for the infinite. . . . The ultimate concern is concern about what is experienced as ultimate (DF, 9).

These two quotations seem to make the same general point. We can be ultimately concerned only about what is ultimate, beyond all finite limitations, for only in this way could it have the power of determining our being or nonbeing. It seems, then, that ultimate concern is a kind of total commitment or complete devotion to the ultimate, being-itself. But immediately a difficulty emerges. What is it like to be totally committed to being-itself? It may make sense to speak of someone's being totally committed to a person (e.g., Napoleon) or a political movement (e.g., communism), but what sense does it make to speak of someone's being totally committed to being-itself? To clear this up we must explore Tillich's claim that the concrete content of this total commitment (ultimate concern) to beingitself is the religious symbol. In making this point Tillich seems to recognize our quandary over what it could mean to be committed to being-itself. The phrase "being ultimately concerned" points to a tension in human experience. On the one hand, it is impossible to be concerned about something which cannot be encountered concretely, be it in the realm of reality or in the realm of the imagination. . . . On the other hand, ultimate concern must transcend every pre15

8

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

-

liminary finite and concrete concern. It must transcend the whole realm of finitude in order to be the answer to the question implied in finitude. But in transcending the finite the religious concern loses the concreteness of a being-to-being relationship. It tends to become not only absolute but also abstract, provoking reactions from the concrete element. This is the inescapable inner tension in the idea of God (ST, 1:211).

This means that we must amend our statement that ultimate concern is total commitment to being-itself, the ultimate. For the focus of our ultimate concern (total commitment) is something concrete. Without some concrete content we cannot be ultimately concerned. Yet Tillich wants to say also that we cannot be ultimately concerned about what is not the ultimate—and anything that is concrete is not the ultimate. This is beginning to look like a contradiction. Consider these three statements: (1) every concrete thing is less than the ultimate; (2) we can be ultimately concerned only about something concrete; (3) we are ultimately concerned about the ultimate. We cannot assert all three statements without contradiction. Tillich wants to assert something like these three statements, but avoid any contradiction. Let us focus our attention on 2 and 3.1 have suggested that the former is necessary if we are to make sense out of ultimate concern as total commitment or complete devotion. To be ultimately concerned is to be totally committed to something. The point here is that the object of total commitment (ultimate concern) must make some demands on us, must be able to promise some sort of fulfillment. In Dynamics of Faith Tillich says that that which is our ultimate concern demands total surrender, requires the sacrifice of every other claim to it, promises total fulfillment. It is easy to see how one could love, be devoted and committed to something that can demand total sur16

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

render and promise total fulfillment. But what sense does it make to speak of loving, being devoted to being-itself ? How does being-itself promise total fulfillment? Napoleon can demand total surrender. Communism can promise total fulfillment. But is it not a mistake to think of beingitself doing these things? Only something concrete can perform these functions and be a direct object of love or devotion. Hence, if we understand ultimate concern as total devotion or complete commitment, we can see the necessity of arguing, as Tillich does, that we can be ultimately concerned only about something concrete. We must now ask what leads Tillich to the view that we are ultimately concerned about the metaphysical ultimate, being-itself. One possibility, suggested by Alston, is that Tillich is misled by the verbal ambiguity of 'ultimate/ Thus, in his penetrating essay "Tillich's Conception of a Religious Symbol/' Alston remarks: . . . as Tillich explained 'ultimate concern/ the ultimacy is psychological; it consists in the supremacy of that concern in the psychic structure of the individual. It is in a quite different way that being-itself is thought by Tillich to be ultimate. It is ontologically ultimate by virtue of the fact that it is the ultimate ground of all being. Once this distinction is made, we can see that there is no reason to suppose that (psychologically) ultimate concern must be concern directed to what is (ontologically) ultimate. But the verbal identity may make the transition seem obvious.2

Alston is certainly correct in pointing out that the psychological considerations mentioned in defining ultimate concern do not necessitate that the metaphysical ultimate be 2 William P. Alston, "Tillich's Conception of a Religious Symbol," Religious Experience and Truth, ed. Sidney Hook (New York: New York University Press, 1961), pp. 20-21.

17

•

9

10

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

the object of such concern. If one did think so, he most likely would have been misled by the use of 'ultimate' both for what is psychological and for what is metaphysical. However, it is not clear that Tillich is being misled here, and even if he is he has other reasons for holding that we are ultimately concerned about the metaphysical ultimate. This being so, pointing out that he may be misled by the ambiguity of 'ultimate* will not dispense with his claim unless we also show either that his reasons are not good ones or that they rest on this same ambiguity (which is just a special case of showing that they are not good reasons). Hence, it is necessary to examine the considerations leading Tillich to claim that we are ultimately concerned about being-itself. A complete account of the considerations which lead Tillich to the view that we are ultimately concerned about being-itself would necessitate a careful discussion of the role of being-itself in his ontology—a discussion that must be left for the next chapter. However, as a preliminary account, we may note the following two claims that Tillich makes: (a) we are ultimately concerned about what determines our being or nonbeing—only being-itself can do this; (ft) we are ultimately concerned about that which can provide an answer to the question implied in finitude— only being-itself can do this since everything else is bound by the conditions of finitude. What a and b come to, I think, can be set forth in a series of statements. 1. Man is infinitely concerned (anxious) about his being. 2. The source of his anxiety is nonbeing, for nonbeing is what threatens his being. 3. This infinite concern (anxiousness) about his being produces an 18

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

11

infinite quest, longing, for that which can overcome this threat to his being. 4. That which can overcome the threat of nonbeing is being-itself, for only being-itself is not exposed to that threat. 5. Man, therefore, is seeking for being-itself, for some vital contact with that reality which possesses the power of overcoming the existential threat of nonbeing. 6. Since man can encounter being-itself only through the concrete, his infinite quest for being is focused on something concrete through which the power of being is experienced.

Concerning 1, it is important to recognize that by a man's "being* Tillich does not mean simply his existence in time and space. Rather, when a man is concerned or anxious about his being he is basically concerned about the significance or meaning of his life. Thus Tillich says, *... the term "being' means the whole of human reality, the structure, the meaning, and the aim of existence" (ST, 1:14). The source of this human concern about one's being is, as stated in 2, nonbeing. The concept of nonbeing is a fundamental one in Tillich's existentialist ontology of man, as well as in existentialist literature generally. In The Courage to Be Tillich endeavors to explain his use of this concept and to suggest the various ways in which nonbeing is experienced as a threat to one's being. Nonbeing, he remarks, threatens our being in three ways: Nonbeing threatens man's ontic self-affirmation, relatively in terms of fate, absolutely in terms of death. It threatens man's spiritual self-affirmation, relatively in terms of emptiness, absolutely in terms of meaninglessness. It threatens man's moral self-affirmation, relatively in terms of guilt, absolutely in terms of condemnation. The awareness of this threefold threat is anxiety appearing in three forms, that of fate and death (briefly, the anxiety of death), that of emptiness and loss of meaning (briefly, the anxiety of meaninglessness), that of guilt and condemnation (briefly, the anxiety of 19

12

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION condemnation). In all three forms anxiety is existential in the sense that it belongs to existence as such and not to an abnormal state of mind as in neurotic (and psychotic) anxiety.3

It is not my purpose here to analyze Tillich's concept of nonbeing or to evaluate his claims concerning the ways in which man experiences nonbeing as a threat to his being. My aim is to set forth in logical order the steps by which Tillich comes to the conclusion that being-itself is the object of man's ultimate concern. The first step, as we have seen, is his claim that man is ultimately concerned (anxious) about his own being. The source of this anxiety is nonbeing. The anxiety or concern about one's own being produces in man a quest or longing for that which can overcome the threat of nonbeing, for that which determines our being or nonbeing. It is important to note that Tillich uses the same expression, 'ultimate concern/ to cover both our anxiety concerning our own being and our longing for and commitment to that which can overcome the threat to our being. Thus he asserts: "Man is ultimately concerned about his being and meaning'" (ST, 1:14) and "Man is infinitely concerned about the infinity to which he belongs, from which he is separated, and for which he is longing'' (ST, 1:14). It appears, then, that Tillich uses 'ultimate concern' in at least two different senses. Alston draws attention to this ambiguity. First, "man is ultimately concerned about his being and meaning" in the sense of 'concern* in which it means something like *being worried about' or 'being anxious about/ ("I am concerned about his 3 Paul Tillich, The Courage to Be (New Haven: Yale University Press, 1952), p. 41.

20

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

13

state of health/') But this is a quite different sense from that which Tillich has given to the phrase 'ultimate concern/ Surely Tillich is not suggesting that we are worried about the fate or condition of being-itself!4

The fact that Tillich uses 'concern' not only in his special sense of 'commitment to/ but also in the more ordinary sense of 'anxious about/ does not in itself signify any error in his reasoning. However, an ambiguity in a crucial expression can easily result in faulty reasoning, for a statement containing the ambiguous expression may be true in one sense of that expression but false in a second sense of that expression. Thus it is important to note the fact that Tillich uses 'ultimate concern' in two different senses. However, I do not think that this particular ambiguity leads Tillich into any serious mistakes. Tillich holds—for reasons to be investigated later—that being-itself is that which can overcome the threat of nonbeing, for only being-itself is not exposed to that threat. Since man is longing for that which can overcome this threat, Tillich concludes that man is seeking for beingitself, for some vital contact with that reality which possesses the power of overcoming the existential threat of nonbeing. However, since man can encounter being-itself only through the concrete, his infinite quest becomes focused on something concrete through which the power of being is manifested. The preceding paragraphs represent a brief, general account of the steps by which Tillich is led to claim that man is ultimately concerned about being-itself. However, there is more than one way of picturing the logical progres* Alston, Religious Experience and Truth, p. 20. 21

14

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

sion of Tillich's thought in connection with this claim. Instead of taking man's infinite concern about his own being as the starting point, we can just as easily construe Tillich's thought as beginning with a phenomenological description of the relationship of the religious man to some concrete, sacred object. In characterizing the way the religious man experiences and responds to the sacred object, Tillich introduces the concept of ultimate concern. "Ultimate concern' is defined in terms of certain human attitudes and feelings focused on sacred objects. Having described what goes on in profound religious experience, Tillich then, quite naturally, seeks to explain why men become ultimately concerned about finite, limited things. Why is it that such a vast variety of finite objects (human beings, animals, statues of stone, etc.) have become objects of man's ultimate concern? Psychological and sociological answers to this question have been given.5 Tillich, I think, is trying to answer this question in such a way that religion in all its forms is not deprived of objective validity. It is at this point that his elaborate ontology swings into play. Man is striving for that reality which can overcome the threat of nonbeing. Since being-itself is that reality, Tillich concludes that what man is striving for is some vital contact with being-itself. Given this metaphysical background, the phenomenon of ultimate concern focused on sacred objects can be explained as the way in which man's striving for being-itself expresses itself in actual life. Ultimate concern, Tillich concludes, is really directed at beingitself through the concrete, sacred object. Sacred objects, 5 See Tillich's discussion in "The Religious Symbol," The Journal of Liberal Religion, 2 (Summer, 1940): 1S-33.

22

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

then, are the media through which the power of being is communicated to men. In giving a preliminary account of the considerations which lead Tillich to the view that we are ultimately concerned about the metaphysical ultimate, being-itself, we have distinguished between (1) Tillich's phenomenological description of religious experience and (2) the elaborate ontology he employs in explaining the phenomena of religious experience. In my view Tillich is at his best when he is describing the complex phenomena of religious experience. He has, I think rightly, drawn our attention to the similarities between the attitudes and feelings the "religious man" directs at sacred objects and the attitudes and feelings that men direct at what would normally be regarded as non-religious objects—the state, success, etc. In characterizing these latter attitudes, feelings, and their objects as "religious," as instances of "ultimate concern," Tillich undoubtedly deviates from the common usage of "religion" and "religious." But in so doing he makes us aware of fundamental similarities that our ordinary use of language may conceal. However, it is one thing to describe brilliantly the phenomena of religious experience and to point out the pervasive character of religious attitudes and feelings, and it is quite another thing to endeavor to explain these phenomena in terms of an elaborate metaphysical system involving, as it does, such complicated philosophical concepts as being-itself, nonbeing, participation, etc. It is one thing to point out the vast variety of finite objects which have been the focus of religious attitudes (ultimate concern), and quite another thing to explain this phenomenon as the way in which man, in actual life, ex-

15

16

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

presses his ultimate concern about being-itself. For the explanation of the phenomena of religion is given in terms of a particular metaphysics, and the question of the meaningfulness and truth of the metaphysics is not answered by any description, however accurate, of religious experience. In part, the purpose of this book is to single out and critically examine those aspects of Tillich's ontology which bear directly on his doctrine of God. Of chief concern, both in this chapter and the next, is the concept of being-itself and the use Tillich makes of this metaphysical concept in developing his doctrine of God. We began with two statements about God: (1) God is that which concerns us ultimately, and (2) God is beingitself. In exploring 1 it appeared at first that any finite object might be an object of ultimate concern. But a more careful examination of what Tillich says about the situation of ultimate concern disclosed that the finite object is to be understood as the focus of our ultimate concern about being-itself. The finite object expresses or symbolizes our ultimate concern, which is directed toward the ultimate, being-itself. Hence, 1 and 2 are not incompatible. We must now consider Tillich's assertion that God is being-itself. Before embarking on the difficult task of determining what Tillich could mean by the expression ^being-itself/ it is necessary to raise the question of the logical status of the statement 'God is being-itself/ Tillich holds that all religious statements about God are symbolic (ST, 1:9). Does the statement 'God is being-itself have the same logical status as religious statements about God? Or is this statement one which Tillich regards as literal? It is not difficult to find an answer to this question in Tillich's writings. The difficulty is to find a consistent answer. For it seems to me 24

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

that Tillich has shifted his view on this question at least twice in print. If the first two volumes of his Systematic Theology were consistent on this question, we could take the answer they give as revealing Tillich's mature opinion on the matter. But, as I shall show, one of the two shifts occurs between volumes 1 and 2 of the Systematic Theology. Hence it is necessary to set forth the development of Tillich's thought on this topic and to indicate the shifts that have occurred before we can decide the matter, if we can decide it at all. In an essay appearing in 1940 in the Journal of Liberal Religion, entitled "The Religious Symbol," Tillich made two assertions that drew serious criticism from the pen of Wilbur M. Urban. In his discussion Tillich argued that the distinguishing feature of the religious or mythical symbol is its reference to the transcendent. Urban takes the same position.6 However, Tillich asserted that the referent of the religious symbol transcends even being. "The thing referred to in the mythical symbol is the unconditioned transcendent, the source of both existence and meaning, which transcends being-in-itself as well as being-for-us."7 Tillich's critics understood ^being-in-itself as referring to the metaphysical ultimate. Urban claimed Tillich's view was contradictory. Aubrey expressed the criticism as follows: Dr. Urban's objection to the phrase applied to the unconditioned— "which transcends being-in-itself"—I share. The idea appears to come from Boehme's Ungrund and indicates a negation of all 6 See Wilbur M. Urban, Language and Reality (New York: Macmillan, 1939), chap. 12. 7

Tillich, Journal of Liberal Religion, 2 (Summer, 1940): 26. 25

17

18

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION descriptions. But since the phrase "being-in-itself" is usually employed to connote the ultimate, the notion that there could be that which transcends the ultimate strikes me as a logical impossibility.8

In a reply to both Urban and Aubrey published in the Journal of Liberal Religion (1940, pp. 202-6), Tillich accepted this criticism and suggested that all he wished to prevent was the identification of God with some special being. The second, and related point, on which Tillich was criticized by Urban concerned Tillich's assertion that "all knowledge of God has a symbolic character."9 Tillich's point is that while one can and does experience the unconditioned as the ground and abyss of everything conditioned, there is no positive, literal characterization of the unconditioned that can be given. Hence all our statements about God are necessarily symbolic. Urban characterized Tillich's view as pan-symbolism—"the view, namely, that all knowledge of God has a symbolic character. As I have sought to point out, the notion of symbolic knowledge (and symbolic truth) is meaningless except in contrast with nonsymbolic knowledge/'10 Now if we take Urban to be saying that it is meaningless to assert that all statements about God are symbolic (as Tillich apparently understood Urban), I do not think Tillich accepted this criticism in his 1940 reply to Urban and Aubrey. At least he nowhere in the reply suggests that we can make any literal statement about God. But by 1951, with the publication of volume 1 of the Systematic Theol8 Aubrey, Journal of Liberal Religion, 2 (Spring, 1941): 202. » Tillich, Journal of Liberal Religion, 2 (Summer, 1940): 28. i° Urban, Journal of Liberal Religion, 2 (Summer, 1940): 35-36.

26

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

19

ogy, Tillich took the position that there is exactly one statement about God that is literal—the statement that God is being-itself. The statement that God is being-itself is a nonsymbolic statement. It does not point beyond itself. It means what it says directly and properly; if we speak of the actuality of God, we first assert that he is not God if he is not being-itself. Other assertions about God can be made theologically only on this basis (ST, 1:239).

That Tillich credits Urban's criticism with showing him the necessity of at least one statement about God being nonsymbolic is clear from the following: An early criticism by Professor Urban of Yale forced me to acknowledge that in order to speak of symbolic knowledge one must delimit the symbolic realm by an unsymbolic statement. I was grateful for this criticism, and under its impact I became suspicious of any attempts to make the concept of symbol all-embracing and therefore meaningless. The unsymbolic statement which implies the necessity of religious symbolism is that God is beingitself, and as such beyond the subject-object structure of everything that is.ii 11

Charles W. Kegley and Robert W. Bretall, eds., The Theology of Paul Tillich (New York: Macmfflan, 1959), p. 334. In his book, Paul Tillich and the Christian Message (New York: Scribner's, 1962), Father Tavard is careless in his quotation of this passage and, as a result, misunderstands what Tillich is saying. He writes: "What then is the unsymbolic statement that makes symbolic knowledge possible? It is 'that God is being-itself and as such beyond the subject-object structure of everything that is/ It is interesting to note that contrary to TiUich's impression, this is a beautiful instance of symbolic language. The word *beyond* as applied to the Unconditional is obviously a symbol, . . /' (pp. 55-56). But Father Tavard neglects to note that there is a comma between 'being-itself and 'and* in the statement he quotes from p. 334 of The Theology of Paul Tillich— thus ^giving the impression that the unsymbolic statement is: "God is beingitself and as such beyond the subject-object structure of everything that is." The correct quotation is "that God is being-itself, and as such beyond the subject-object structure of everything that is/' This makes it clear that 'beyond* is not part of the unsymbolic statement. The unsymbolic statement is simply "God is being-itself'—this is certainly made clear on p. 239 of Systematic Theology 1.

27

20

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

Hence, it seems that one shift Tillich has made is from the position that all statements about God are symbolic to the position of volume 1 of the Systematic Theology that there is exactly one statement about God that is nonsymbolic, namely, the statement that God is being-itself.12 The second shift in Tillich's thought seems to occur in the Introduction to volume 2 of the Systematic Theology.1* I say "seems to occur" for on the very next page Tillich continues to speak of "defining God as being-itself/'14 Be this as it may, Tillich makes the point in the Introduction that "everything religion has to say about God, including his qualities, action, and manifestation, has a symbolic character .. ." (ST, 2:9). He then raises the question whether there is a point at which a nonsymbolic assertion about God must be made, and says: "There is such a point, namely, the statement that everything we say about God 12 1 fail to see the force of Urban's criticism. Apparently he is arguing that 'symbolic* is a polar concept and is unintelligible without its polar opposite, 'nonsymbolic' or literal.' Hence, to say "all knowledge is symbolic" is senseless because it is an essential feature of the meaning of 'symbolic knowledge* that it be in contrast with 'nonsymbolic knowledge/ The same sort of argument has been given in connection with other pairs of words such as, 'large* and 'small, Vague' and 'clear,' 'animate' and 'inanimate' (see N. Malcolm, "Moore and Ordinary Language," in The Philosophy of G. E. Moore [Evanston: Northwestern University Press, 1942], pp. 264-65). I do not wish to quarrel with this general point. Nor do I wish to argue that 'symbolic' and 'nonsymbolic' are not instances of this general point. The point which Urban's argument overlooks is that Tillich's original claim is with reference only to God. It is one thing to claim that (1) all statements about God are symbolic, and another thing to claim (hat (2) all statements about anything whatever are symbolic; 2 may be meaningless, but the fact that 2 is meaningless does not entail, by itself, that 1 is meaningless. 13 14

See Introduction above, note 1.

This remark assumes that Tillich would count a definition of 'God* as a statement about God.

28

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

21

is symbolic. Such a statement is an assertion about God which itself is not symbolic. Otherwise we would fall into a circular argument" (ST, 2:9). If I understand this passage correctly, Tillich is here implying that 'God is being-itself is not the straightforward, nonsymbolic statement he took it to be in the first volume. He still holds that not every statement about God can be symbolic. But now he says that the one nonsymbolic assertion about God is the statement, (1) Every statement about God is symbolic (I take 1 to be equivalent to (2) Everything we say about God is symbolic.) If this is so then 'God is being-itself' can no longer be the simple nonsymbolic statement of volume 1. Tillich suggests as much: If we say that God is the infinite, or the unconditional, or beingitself, we speak rationally and ecstatically at the same time. These terms precisely designate the boundary line at which both the symbolic and the non-symbolic coincide (ST, 2:10).

In fact, once we examine Tillich's new position that (1) Every statement about God is symbolic, it begins to look like the original position that Urban criticized with one small difference—Tillich now wishes to call 1 a statement about God which is nonsymbolic. But is not this addition a mistake? Statement 1 is not a statement about God but a statement about statements about God, namely, that all of them are symbolic. Tillich's original position was that no positive, literal characterization could be given of God as the unconditioned. To give a positive, literal characterization of something is to attach a positive predicate to that thing (e.g., loves/ 'created heaven and earth,' 29

22

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

'first cause*) and to claim that the resulting assertion is to be taken literally. But to say 'Every statement about God is symbolic' is not to attach a positive predicate to God at all. It is to say something about statements about God. Tillich could reply that even though 1 does not have the same form as a statement like 'God is the first cause5 or 'God created heaven and earth'; nevertheless, it does tell us something about God, it does provide some information concerning the unconditioned. What it tells us is that God is incomprehensible in the sense that no literal, positive predicate can be meaningfully applied to him. Thus while 1 appears to be about theological statements and not about God, it really is about God since it tells us that he is incomprehensible. But even if this were true, 1 turns out to be just another way of saying that we can give no positive, literal characterization of God—and this is precisely Tillich's original position as set forth in the 1940 essay "The Religious Symbol/' Hence 1 implies that the statement 'God is being-itself must be taken symbolically rather than literally. I have discussed Tillich's shifts concerning the question of the logical status of the statement 'God is being-itself since it is of considerable importance in elucidating Tillich's concept of God, to determine whether TDeing-itself applies symbolically or literally to God. As we have seen, Tillich's writings do not yield a consistent answer to this question. But before we proceed, some disposition of this question must be made. As far as this study is concerned, die doctrine of volume 1 will be accepted; namely, that 'God is being-itself' is a literal (nonsymbolic) identity statement. This has the consequence that the ontological status of God and the ontological status of being-itself 30

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

are not two different questions but one and the same question. If being-itself were not identical with God but simply a symbol for God, these two questions would be distinct. I take the doctrine of volume 1 because it is not clear that Tillich means to be denying it in the passages quoted from volume 2. I think the denial is implied, but it is not clear that Tillich is aware of this. I say this is not clear because in restating some aspects of his doctrine of God (set forth in volume 1), Tillich makes a point of saying "in none of these cases has the substance of my earlier thought changed, but formulations have proved to be inadequate in clarity, elaboration, and emphasis" (ST, 2:5). Be this as it may, having accepted as Tillich's position the view expressed in volume 1, we may now proceed with our discussion of the question: What meaning does Tillich attach to the expression 'God? The meaning of 'God' is expressed by Tillich in the phrases T^eing-itself,' 'the ground of being/ or 'the power of being' (ST, l:235-36).15 Of course, this alone will not suffice. The objection to explaining 'God* by these phrases is that the phrases themselves must be explained if we are to discover what 'God' means in Tillich's thought. A second objection is that 'God' is primarily a religious notion, whereas the phrases used in explaining the meaning of 'God' are primarily philosophical. Is there any reason to suppose that in using the former, religious people are referring to what Tillich means by the latter? Even if a !5 However, see Kegley and Bretall, The Theology of Paul Tillich, p. 335, where in reply to a criticism by Randall, Tillich rejects the literal equivalence of the phrases and takes the position that 'power* and 'ground* are symbols for being-itself.

31

23

24

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

reply can be given to the first objection—that is, even if we can explicate satisfactorily the notion of being-itself— what grounds does Tillich have for identifying a basically religious notion with a basically philosophical notion?16 Is the move from the religious level to the philosophical level anything more than arbitrary? I wish to take up both these objections—but in the reverse order in which I have presented them. It should be noted that, in identifying a philosophical concept (being-itself) and a religious concept (God), Tillich is not doing anything new in theology. Aquinas did substantially the same thing when he argued for the existence of a first efficient cause and then added "to which everyone gives the name of God." We might want to ask in the case of Aquinas whether this is what people actually mean when they use the expression 'God/ Even more, in the case of Tillich, we wonder whether l>eingitself is what people actually mean when they use the expression 'God/ It is some such puzzlement as this which lies behind the objection to the procedure of telling us what the expression 'God' means by referring us to a philosophical concept such as 'first cause' or TDeing-itself/ Perhaps we should say that Tillich, unlike Aquinas, does not claim to be relating his discussion to what 'everyone' means by the expression 'God/ Perhaps he only intends to be telling us what he means by it. But this move does not reflect Tillich's concern to relate his characterization 16 By a religious expression I mean an expression which has a place in religious activity—prayer, worship, reciting creeds, etc. Thus 'being-itself is not (or, at least, not primarily) a religious expression, whereas 'God' obviously is.

82

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

25

of 'God' to the way in which the expression is used in actual religions. My suggestion is that, in identifying God with being-itself, Tillich is trying to develop a doctrine of God which adequately accounts for what he regards as a basic tension in the way in which 'God' is used and God is experienced in actual religions. If we clarify this tension in the actual use of the expression 'God/ we can get some idea of the problem which the notion of beingitself is set up to deal with. In discussing the meaning of 'God/ Tillich points out: A phenomenological description of the meaning of "God" in every religion, including the Christian, offers the following definition of the term "god." Gods are beings who transcend the realm of ordinary experience in power and meaning, with whom men have relations which surpass ordinary relations in intensity and significance (ST, 1:211-12).

Tillich proceeds to discuss each element of this description of the meaning of 'God/ It is in this discussion that the tension in the actual use of the expression 'God' becomes clear. The tension Tillich points to is between the "tendency toward concreteness" and the "tendency toward ultimacy." As religious people use the expression 'God/ it refers to a being who exhibits human qualities and limitations. God is influenced by human action and in turn influences human actions. "Gods are substances, caused and causing, active and passive, remembering and anticipating, arising and disappearing in time and space" (ST, 1:212). God is spoken of as located in a burning bush, in a tabernacle, in pieces of bread and drops of wine. All this is very familiar in a religion. It is to these features of the use of the expression 'God' that Tillich is 33

26

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

pointing when he speaks of the "tendency toward concreteness" in the idea of God. But there is also a strong tendency toward ultimacy. . . . not only do the images of the gods bear all the characteristics of finitude—this makes them images and gives them concreteness —but they also have characteristics in which categorical finitude is radically transcended. Their identity as finite substances is negated by all kinds of substantial transmutations and expansions, in spite of the sameness of their names. Their temporal limitations are overcome; they are called "immortals" in spite of the fact that their appearance and disappearance are presupposed. Their spatial definiteness is negated when they act as multi- or omnipresent, yet they have a special dwelling place with which they are intimately connected. Their subordination to the chain of causes and effects is denied, for overwhelming or absolute power is attributed to them in spite of their dependence on other divine powers and on the influence finite beings have on them (ST, 1:212-13).

Tillich is making two points here. First, he is arguing that there is another tendency in the use of the expression 'God/ Gods are called immortals, invisible, possessors of unlimited power, possessors of perfect knowledge, etc. In speaking of God, religious people deny that he is subject to any empirical limitations, that he is bound by space or time, that there is anything he does not know. God is felt to be mysterious, incomprehensible, absolutely unconditioned. It is to these features that Tillich is pointing when he speaks of the "tendency toward ultimacy" in the idea of God. The second point he is making is that there is a genuine tension or struggle between the two tendencies of ultimacy and concreteness. It can be objected here that not all religious people experience this tension in the decisive way Tillich suggests. Does the mystic refer to God as having a spatial location, dependent on other divine powers, etc.? However, in spite of negative in34

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

27

stances, as a general characterization of the idea of God as it occurs in various religions, Tillich's description in terms of a tension between two diverse tendencies is quite acceptable. Of course, a given religion or even a given period in the history of a religion may move in one direction more than another. Consequently, in discussing religions Tillich develops various types ranging from forms of polytheism (emphasis on concreteness) to mystical monotheism (emphasis on ultimacy). I have suggested that Tillich's identification of God and being-itself is not an arbitrary move but is an attempt to develop a doctrine of God which adequately accounts for a basic tension in the concept of God as it is employed in a religion. This tension refers to our speaking about God both as a concrete being, subject to various limitations and as absolutely transcendent, beyond all limitations. I wish now to make some general remarks concerning the way in which Tillich tries to account for this tension by means of identifying God with being-itself. The rationale for doing this is that it will provide some direction to the task of determining what Tillich means by T^eingitself.' Tillich's discussion of being-itself moves around two fundamental assertions. The first assertion is that every being is subject to limitations; whereas, being-itself is beyond all limitations. The second assertion is that every being participates in being-itself. Both assertions are made in the following statement: As the power of being, God transcends every being and also the totality of beings—the world. Being-itself is beyond finitude and infinity; otherwise it would be conditioned by something other than itself, and the real power of being would lie beyond both it and 35

28

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION that which conditioned it. Being-itself infinitely transcends every finite being. . . . On the other hand, everything finite participates in being-itself and in its infinity. Otherwise it would not have the power of being. It would be swallowed by nonbeing, or it would never have emerged out of nonbeing (ST, 1:237).

Tillich uses the first assertion to account for the aspect of ultimacy in the idea of God. The argument here is that religious statements about God must be statements about either a being or being-itself. But every being is subject to limitations which the tendency toward ultimacy in the idea of God denies. Hence, in developing a doctrine of God, it is a mistake for a theologian to identify God with a being. God must be thought of as being-itself, for only being-itself can account adequately for the aspect of ultimacy. But if being-itself can account for the aspect of ultimacy by virtue of being beyond all limitations, how is Tillich able to accommodate the aspect of concreteness in the idea of God? For if being-itself is free of all limitations, it cannot be concrete. Tillich's solution here is interesting. He appeals to the second assertion about being-itself, namely, that every being participates in being-itself. The argument here is that, by virtue of the participation of every being in being-itself, every being has the potential to express the concrete element in the idea of God. That is, Tillich is arguing that the relationship of participation between being-itself and every being makes it possible for a particular being to become a focal point through which God is disclosed. (Apparently he holds that, in order for a being to mediate or manifest God to man, that being must stand in some ontological relationship to God. This is a necessary, but not a sufficient, condition of revelation.) 36

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

When any being becomes such a focal point—a sacred object or religious symbol—Tillich views that being as an expression of the tendency toward concreteness in the idea of God. Hence Tillich thinks he can account for the concrete element in religious experience and religious discourse about God by means of the presence and disclosure of the divine in the concrete, made possible by the fact that concrete beings participate in being-itself (God). The preceding paragraphs give rise to two critical questions: First, Can we make any sense out of Tillich's notion of being-itself? Clearly only to the extent that it is possible to do so can we say that Tillich has given an intelligible, let alone adequate, account of the aspect of ultimacy in the idea of God. Second, Does Tillich's theory of religious symbols provide a coherent and adequate account of the concrete element in the idea of God? Although these questions appear to be logically independent of one another, they are not. For, as we shall see, Tillich makes use of the concept of being-itself in the development of his theory of religious symbols. In the rest of this chapter, as well as the next, I shall endeavor to answer the first of these two questions. The second question will be explored in later chapters. We have seen that the philosophical concept of beingitself is introduced by Tillich in order to give an adequate account of certain tensions present in religious discourse about God. This indicates that Tillich takes the view that religious statements about God are to be elucidated, or at least partly explained, by translating them into metaphysical statements—in this case into metaphysical statements about being-itself. However, it should be obvious that the value of the proposed elucidation will depend on the ex37

29

30

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION

tent to which we can understand what Tillich means by the expression "being-itself/ Thus in order for us to understand Tillich's use of the concept God and his analysis of theological statements (statements with 'God' as the subject expression—e.g., 'God is the creator of heaven and earth') we must be able to give some answer to the question, What is the meaning of *being-itself ? What does Tillich mean by 'being-itself? Rather than give an unclear answer to this question, I propose to raise and answer a prior question; namely, is it possible to answer this question, and, if it is possible, what kind of answer can be given? There are at least three ways in which one might try to answer such a question. If one is asked what he means by a certain term, one way in which he can answer is by giving a definition of the term in question. However, this way is not open to Tillich. It is impossible to define being-itself, Tillich argues, because it is presupposed in every definition.17 If one is unable to give a definition of the term in question, he still might be able to provide a more familiar expression equivalent in meaning to the term in question.18 Tillich does propose 'power of being' and 'ground of being' as substitutes for 'beingitself.' However, there are two reasons why these will not 17 Paul Tillich, Biblical Religion and the Search for Ultimate Reality (Chicago: University of Chicago Press, 1955), p. 19. 18 One might speak of such an expression as a synonymous definition in contrast, say, to definition per genus et differentia. If so, then Tillich's claim that it is impossible to define being-itself must not be extended to synonymous definitions, for he seems to hold that there are expressions identical in meaning with being-itself. Thus he says, "Many confusions in the doctrine of God and many apologetic weaknesses could be avoided if God were understood first of all as being-itself or as the ground of being. The power of being is another way of expressing the same thing in a circumscribing phrase (ST, 1:235-36).

38

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

31

help us here. First, 'power of being' and 'ground of being' are as much in need of elucidation as the original expression, T)eing-itself.' If we understood the substitutes it is unlikely that we would have been in the dark concerning the original. Second, when pressed on these matters, Tillich withdraws 'power of being' and 'ground of being' as literal equivalents and suggests that they are symbols for God or being-itself .19 Hence, these two substitutes will not do as attempts to explain the literal meaning of 'beingitself.' Even if one is prevented from giving a definition or providing an expression which is equivalent in meaning, perhaps he might still give a partial answer by indicating that what he is talking about is of the same general sort as other entities with which we are acquainted. Thus, if one introduces the term 'citrine/ one can indicate what is meant by pointing out that citrine is a color—in. this way indicating what sort of entity citrine is. Now is this way a possibility for Tillich? I am not sure what the right answer is to this question. It is clear that we cannot say what sort of entity being-itself is if this involves saying that beingitself is an x where Y designates a class of beings. The point here is that Tillich holds that every being is subject to the structure of being. This implies that of every class of beings it is the case that every member is subject to the structure of being. This case is analogous to saying that since every animal is a living thing, it will be true that of every class of animals—man, canine, etc.—every member is a living thing. Now if we are to say what sort of thing a is by saying a is a man, we cannot deny that a is a living thing. If we deny initially that a is a living thing we can19

Kegley and Bretall, The Theology of Paul Tillich, p. 335. 39

32

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION

not say what we mean by V by saying a is a man. Similarly, since Tillich holds that only being-itself is not subject to the structure of being, he cannot say what he means by T^eing-itself by saying that being-itself is an x where V designates a class of beings. The real question here is, I think, whether on Tillich's theory, we can literally predicate any general terms of an entity without presupposing that the entity is subject to the structure of being. If we cannot, then there is no way of making a literal positive statement about being-itself. Hence, in saying what Tillich means by being-itself we could choose only what seem to be appropriate metaphorical or symbolic statements. Tillich's position is not clear on this point. Certain remarks he makes suggest that he does hold that no literal statement can be made about being-itself. As we already have seen, God as being-itself is the ground of the ontological structure of being without being subject to this structure himself. . . . Therefore, if anything beyond this bare assertion is said about God, it no longer is a direct and proper statement, no longer a concept. It is indirect, and it points to something beyond itself. In a word, it is symbolic (ST, 1:239).

Notice that it is because God is being-itself that no positive, literal assertion can be made about him. Much the same point is made concerning the attachment of concepts like cause to God or being-itself. For Tillich, to say x is a cause in the literal sense of 'cause' is to imply that x is a part of the causal series, that x is an effect as well as a cause. 'Cause* gets its meaning in application to entities, subject to the structure of being. If we say that God or being-itself is the first cause, this means "that the category of causality is being denied while it is being used. In other 40

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

33

words, causality is being used not as a category but as a symbol" (ST, 1:238). Statements such as these suggest that Tillich is holding that general terms and predicate expressions get their meaning by being attached to entities subject to the structure of being. Hence, they cannot be applied literally beyond the structure of being. When we do so apply them, they must be taken symbolically, not literally. If this were the case with all our general terms (and Tillich gives us ho reason for thinking it is not) then every statement about being-itself would be metaphorical or symbolic. In fact, Tillich says at one point, "Every assertion about being-itself is either metaphorical or symbolic/'20 If this is so, and if 'x is ineffable' is understood as 'no positive, literal assertion can be made about x? we can say that on Tillich's theory being-itself is ineffable.21 The main difficulty we have thus far uncovered in trying to explicate Tillich's concept of God can be expressed as follows: (1) to explicate Tillich's concept of God we must first understand what ^being-itself' means; (2) being-itself is ineffable. Two points follow from 1 and 2. First, there is something wrong with Tillich's proposed elucidation of our admittedly analogical or symbolic discourse about God in terms of ontological statements about being-itself. For on his own account being-itself is ineffable and consequently, statements about being-itself must be construed as symbolic or analogical. Thus the ontological statements 20 The Courage to Be, p. 179. 21 There seems to be a difficulty here. Tillich argues elsewhere that "in order to speak of symbolic knowledge one must delimit the symbolic realm by an unsymbolic statement" ("Reply to Interpretation and Criticism," in Kegley and Bretall, The Theology of Paul Tillich, p. 334). Now he seems to be saying that no unsymbolic statement can be made about being-itself.

41

34

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

—at least those whose subject term is "being-itself—seem as much in need of elucidation as the theological statements they were to elucidate. Second, there does not seem to be any clear answer to the question "What does Tillich mean by 'God?" For any answer to this question must involve an elucidation of what is ineffable, being-itself. In view of this, the best that can be done, it seems to me, is to compare and contrast the role that being-itself plays in Tillich's philosophical theology with the role of ineffable concepts in more fully worked out metaphysical systems. Thus, for example, one can compare and contrast beingitself in Tillich with The Good in Plato and The One in Plotinus. This, of course, will be profitable only if The Good or The One is more intelligible than being-itself and, hence, may serve to illuminate the latter. This, it seems to me, is the direction in which one must proceed if he hopes to get a philosophically adequate interpretation of Tillich's concept of God, if such an interpretation is possible. It is in this direction that we shall proceed.

42

CHAPTER TWO

Religious Symbols

In his writings on symbolism Tillich has endeavored to distinguish various 'types' or levels' of religious symbols. There appear to be two basic distinctions he is anxious to make. One distinction is between two levels of religious symbols, a 'transcendent level' and an 'immanent level/ "There are two fundamental levels in all religious symbols: the transcendent level, the level which goes beyond the empirical reality we encounter, and the immanent level, the level which we find within the encounter with reality" (TC, 61). A second distinction is between two kinds of religious symbols, 'primary* and 'secondary/1 If I understand Tillich, primary religious symbols are those which 'represent' the ultimate, being-itself; whereas secondary 1 Tillich, "The Meaning and Justification of Religious Symbols/' Religious Experience and Truth, ed. Sidney Hook (New York: New York University Press, 1961), p. 8.

36

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

religious symbols are those which 'support' and/or 'symbolize' symbols on the primary level. Of secondary religious symbols Tillich has little to say. Therefore, I shall not undertake an account of them. It is clear that his main concern is to describe those religious symbols he calls 'primary/ Our first concern must be to relate these two basic sets of distinctions. When Tillich's discussion of primary religious symbols is carefully compared with his discussion of transcendent level and immanent level symbols, it becomes clear that the distinction between two levels of religious symbolism is a distinction within primary religious symbols. Thus, within the class of primary religious symbols—that is, symbols which "point directly to the referent of all religious symbolism"—we are to distinguish two basic sub-classes; namely, a class of transcendent level symbols and a class of immanent level symbols. Primary religious symbols on the transcendent level all concern God. Here Tillich distinguishes (a) God himself as a symbol, (b) the qualities and attributes of God as symbols, and (c) the actions of God as symbols. Primary religious symbols on the immanent level concern the appearances of the divine in the realm of empirical reality. "Here we have first of all the incarnations of the divine, . . . divine beings transmuted into animals or men or any kinds of other beings as they appear in time and space" (TC, 63). Tillich's view that God himself is a symbol is quite difficult to understand. For he seems to hold that the religious idea of God is only partly symbolic. Thus he remarks: But we cannot simply say that God is a symbol. We must always say two things about him: we must say that there is a non-symboHc element in our image of God—namely, that he is ultimate reality, 172

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH being-itself, ground of being, power of being; and the other, that he is the highest being in which everything that we have does exist in the most perfect way. If we say this we have in our mind the image of a highest being, a being with the characteristics of highest perfection. That means we have a symbol for that which is not symbolic in the idea of God-namely, "Being Itself' (TC, 61).

One difficulty this passage raises is how we can render it consistent with Tillich's view, discussed in Chapter I, that God is being-itself and with the view of the ordinary religious man that God is the supremely perfect being. That is, the passage quoted seems to imply that 'God' means both a supremely perfect being and being-itself. Tillich's own view, discussed earlier, seems to be that 'God* means being-itself, and the plain religious man is likely to mean by 'God' no more or less than the supremely perfect being. The solution to this difficulty is, I think, as follows: Tillich does wish to identify God with being-itself, to use the term 'God' to designate solely the unconditionally transcendent. But we are not to construe his view that 'God' means being-itself as a report about the use of the word 'God' in a religious setting, as a report of the meaning ordinarily attached to the word 'God'. Rather, I think we must construe Tillich's "definition" of 'God' as a proposal, the justification for which is its usefulness in solving theoretical problems. Thus, just as the physicist may use ordinary terms—'matter', 'energy', etc.,—in technical ways for theoretical purposes, so the philosophical-theologian may take terms from ordinary religious discourse and give them a technical sense for theoretical purposes. However, it would be a mistake to represent Tillich as introducing a technical meaning for the term 'God' which in his view has no connection with the more familiar notion of God. For, if I understand Tillich, his view is that in the ordinary 173

37

38

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

religious man's conception of God the idea of being-itself is already contained. Thus, even though the plain religious man may have no concept of being-itself, Tillich is, I think, maintaining that as a matter of fact being-itself is an element in his idea of God. This is highly paradoxical and needs explanation. It should be clear by now that on Tillich's theory—involving as it does both a phenomenological description of religious experience and an ontological interpretation of that experience—what is essential and common to religious experience is the encounter with the holy, with beingitself. Since the holy is felt as ultimate and transcendent, and since in trying to conceive of the holy the religious man is bound by his finitude to concrete imagery, he inevitably is led to conceive of the holy as the highest possible object, in short, as the supremely perfect being. This, in brief, is Tillich's explanation for the emergence of the conception of God as the supreme being. The important point, however, is that the idea of a supremely perfect being is the religious man's way of representing what transcends the realm of the concrete altogether, beingitself. The divine beings and the Supreme Being, God, are representations of that which is ultimately referred to in the religious act. They are representations, for the unconditioned transcendent surpasses every possible conception of a being, including even the conception of a Supreme Being.2

We have, then, the following situation. The religious man may quite naturally mean by 'God' the supremely perfect being. But since, on Tillich's view, the notion of a 2

314. 174

Tillich, "The Religions Symbol," Religious Experience and Truth, p.

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

supremely perfect being is the result of the religious man's effort—however unconscious it may be—to form an appropriate religious conception of what is experienced—beingitself—Tillich concludes that the religious man's conception of God has a double meaning, as it were. On the one hand it certainly means a transcendent object, the supremely perfect being. But, on the other hand, it also means the absolutely transcendent, being-itself. But the word "God" involves a double meaning: it connotes the unconditioned transcendent, the ultimate, and also an object somehow endowed with qualities and actions. The first is not figurative or symbolic, but is rather in the strictest sense what it is said to be. The second, however, is really symbolic, figurative. It is the second that is the object envisaged by the religious consciousness. The idea of a Supreme Being possessing certain definite qualities is present in the consciousness. But the religious consciousness is also aware of the fact that when the word "God" is heard, this idea is figurative, that it does not signify an object, that is, it must be transcendent. The word "God" produces a contradiction in the consciousness, it involves something figurative that we really have in mind and that is represented by this idea. In the word "God" is contained at the same time that which actually functions as a representation and also the idea that it is only a representation. It has the peculiarity of transcending its own conceptual content—upon this depends the numinous character that the word has in science and in life in spite of every misuse through false objectification. God as an object is a representation of the reality ultimately referred to in the religious act, but in the word "God" this objectivity is negated and at the same time its representative character is asserted.3

Tillich may be involved in a semantic confusion in holding that the word 'God' has this double meaning. He may also be mistaken on a matter of fact in claiming that the religious consciousness is aware that the word 'God' does not simply signify the highest being. Be that as it may, the principle behind his claim that 'God* has a double mean3 Ibid., p. 315. 175

39

40

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

ing seems to be that since the concept of the supremely perfect being is itself intended as a representation of that which is beyond any concrete conception, we may say that the word 'God' means not only the highest being but also what the concept of the highest being is intended to represent, namely, the holy or being-itself. The material point underlying Tillich's semantic remark about the word 'God' as it is used in religious discourse is that the idea or image of the highest being is the religious man's way of representing or portraying what is not a being at all, but what transcends the realm of objects, namely the holy or beingitself. His view, then, seems to be that the religious idea of God designates not only an object, the highest being, but also being-itself. His own view of God as being-itself—which we may call the philosophical idea of God—amounts to the religious idea of God minus its concrete, symbolic element. From his point of view this is a refinement on the religious idea since it separates out the symbolic content and focuses on what he regards as the essential, common element in the religious encounter. If we ask why the religious conception of God contains the symbolic notion of a highest being, why it does not contain simply the nonsymbolic notion of being-itself, Tillich's answer, presumably, would be that the religious man endeavors to conceive of the ultimate as it is encountered in the religious life. Since it is impossible for man to encounter the ultimate apart from some concrete content, the closest he can come to the ultimate is in terms of that content which most closely approximates the true ultimate, namely, the image of a transcendent being embodying the highest qualities known to man, carried to their highest degree. 176

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

If we adopt Tillich's proposal and simply identify God with the nonsymbolic element in what he takes to be the religious idea of God, I think we can then understand Tillich's view about religious symbols—both linguistic and non-linguistic—and how that view differs from more traditional accounts of religious symbolism. Non-linguistic religious symbols result from the fact that the nonsymbolic element can be encountered only in something concrete. Thus something concrete—a person, a statue, a goal—becomes the focus of ultimate concern and 'represents' the ultimate object of concern—God, being-itself. Linguistic religious symbols arise from the need to talk about God and from the fact that being-itself, God, cannot be described, or otherwise talked about, in nonsymbolic language. The reason God, being-itself, cannot be talked about in literal terms is, as we saw in Chapter I, that language is subject to what Tillich calls "the structure of being'' and can be applied literally only to entities that are subject to that structure. Since being-itself is not subject to the structure of being, it cannot be spoken about literally, but only symbolically. This seems to be the main point of the following passage: As we already have seen, God as being-itself is the ground of the ontological structure of being without being subject to this structure himself. . . . Therefore, if anything beyond this bare assertion is said about God, it no longer is a direct and proper statement, no longer a concept. It is indirect, and it points to something beyond itself. In a word, it is symbolic (ST, 1:239).

We can put Tillich's point in another way. His view is that the primary meaning of a predicate expression is inextricable from the use of that expression to talk about, describe, etc., beings—that is, entities subject to the structure 177

41

42

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

of being. The literal or primary meaning of a predicate expression implies or presupposes that the expression is used to talk about entities subject to the structure of being. A predicate expression therefore cannot be applied literally beyond the structure of being. When a predicate expression is applied beyond the structure of being, as when we talk of God, being-itself, it must be used as a metaphor or a symbol; otherwise we produce either nonsense or contradiction. "Man's ultimate concern must be expressed symbolically, because symbolic language alone is able to express the ultimate" (DF, 41). The radical nature of Tillich's theory of religious myths and symbols—or, more generally, religious discourse about God—is a logical consequence of (1) his view that God is not a being, even a highest being, and (2) his claim that predicate expressions are logically tied (by virtue of their primary meaning) to beings. The radical nature of his theory can be exhibited by contrasting it with a more traditional theory of religious discourse about God. Traditionally, theologians and thoughtful religious men have regarded God as an infinite, immaterial, omnipotent, omniscient, perfectly good being. Accordingly, they have distinguished between predicate expressions which in their primary meaning refer either to parts of a body (e.g., liand', 'face') or to activities impossible without a body (e.g. 'walking', 'drinking') and predicate expressions which in their primary meaning refer to properties of a mind or mental activities (e.g., 'wise', 'good', 'merciful', loving'). Expressions of the first kind cannot be predicated properly (i.e., with their primary meaning) of God. For, since God does not have a body, to apply them properly to him would result either in contradiction or nonsense. Thus, 178

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

such expressions may be predicated of God only metaphorically or symbolically. Expressions of the second kind, it is argued, can be predicated properly of God; however, it is generally pointed out that the properties designated by 'good', 'wisdom', etc., occur in God in the highest possible degree—thus God is said to be infinitely good, infinitely wise, etc. Some philosophers have criticized the traditional view by claiming that the criteria for applying expressions of the second sort involve physical behavior. Since God has no body, it is argued that we have no criteria for applying these expressions (in their primary meaning) to him. Other philosophers have claimed that expressions of the second kind ('good', 'merciful', etc.) involve, as part of their meaning, a reference to bodily behavior. Thus one philosopher recently said that when predicated of God "... all these words lose their meaning if we are told that God does not possess a body.... For what would it be like to be, say, just, without a body? To be just a person has to act justly—he has to behave in certain ways. This is not reductive materialism. It is a simple empirical truth about what we mean by 'just'. But how is it possible to perform these acts, to behave in the required ways without a body?"4 Tillich's rejection of the traditional view, however, does not depend on claiming that expressions of the second sort are logically tied—either by way of their criteria for application or their primary meaning—to bodily behavior. His rejection depends simply on the claim that expressions of the second kind, as well as of the first, can be properly predicated only of beings. Since this is so, it is 4 Paul Edwards, "Some Notes on Anthropomorphic Theology," Religious Experience and Truth, p. 243.

179

43

44

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

clear that the statement "God (being-itself) is merciful" becomes contradictory or nonsense if "merciful" is used in its primary sense. If it is objected that the ordinary religious man means to be talking about a supreme being and not being-itself when he says "God is merciful," Tillich, perhaps, would agree. But he would then argue, as we have seen, that this traditional conception of God is itself a symbol for God as being-itself. This is the meaning of his cryptic remark "God is the symbol of God" (DF, 46). That is, God (conceived as the supreme being) is a symbol for God (conceived as being-itself). Having discussed Tillich's view of (a) God himself as a symbol, we may now give some account of ( b ) the qualities and attributes of God as symbols, and (c) the actions of God as symbols. Perhaps the first point that needs noting is that Tillich here seems to be discussing linguistic religious symbols. That is, he is trying to explain the symbolic character of what I have called theological or religious statements—statements whose subject term is 'God*. These statements are of two sorts, according to whether they ascribe qualities or activities to God. A few examples will facilitate our discussion. 1. God is omnipotent. 2. God is living. 3. God is just. 4. God created the world. 5. God has sent His son. Examples 1 to 3 ascribe certain qualities or attributes to God. Examples 4 and 5 ascribe certain actions to God. Our task here is to explicate and evaluate Tillich's account of the symbolic character of these statements. 180

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

In Chapter I we saw that Tillich's strategy is to translate theological statements about God into ontological statements about being-itself. From the foregoing discussion of God himself as a symbol, it should be clear that on Tillich's view such a translation is not only justified by his philosophical idea of God as being-itself, but also justified by the religious idea of God. For we saw that on his view the religious idea of God contains both a symbolic and a nonsymbolic element. The concept of a supremely perfect being represents the nonsymbolic element, being-itself. The idea of a supreme being is symbolic because (a) it is the religious man's way of representing being-itself and (fo) being-itself—as we have seen—cannot be conceived or talked about in other than symbolic language. It would seem, then, that on Tillich's theory, statements 1 through 5 are symbolic solely by virtue of his claim that their subject term 'God' represents or stands for being-itself. It is because 'God' is taken to stand for being-itself—explicitly by Tillich and implicitly, he maintains, by plain religious men —that statements 1 through 5 are symbolic statements. It is important to distinguish between explaining why 1 through 5 are symbolic statements on Tillich's theory, and explaining what, on his theory, the symbolic statements may be understood to express. The explanation of why 1 through 5 are symbolic is, I think, quite clear. They are symbolic because on Tillich's view they are about beingitself and, as was argued in Chapter I, no literal, positive statement can be made about being-itself. But it is one thing to explain why statements 1 through 5 must be understood to have a symbolic rather than a literal meaning, and it is quite another thing to explain, for each of these statements, what its symbolic meaning is. The really funda-

181

45

46

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

mental, perhaps insoluble, problems in Tillich's account of linguistic religious symbols arise when we ask what is meant by a religious statement on his theory. Indeed, it is Tillich's explanation of why religious statements are symbolic—namely, that they are about being-itself—that creates the problems which arise in explaining what is meant by particular religious statements. In Chapter I, I suggested that there is something wrong with Tillich's proposed elucidation of our admittedly symbolic discourse about God in terms of ontological statements about being-itself. There seems to be something wrong because on his own view being-itself is ineffable and, consequently, statements about being-itself must also be symbolic. But if this is so then the ontological statements about being-itself seem as much in need of elucidation as the ideological statements about God they were to elucidate. In short, if to interpret or explain what is meant by a symbolic or metaphorical statement S is to produce another statement S' which expresses in nonsymbolic terms essentially what S expresses in symbolic terms, it is impossible on Tillich's theory to ever explain or interpret any religious statement about God. For, since 'God* stands for being-itself, the symbolic statement about God would be correctly interpreted only by some ontological statement about being-itself. But, since Treing-itself' is ineffable, every ontological statement about being-itself is necessarily symbolic. The difficulty we have uncovered is, I think, the fundamental problem in Tillich's theory of religious symbolism. It is the source, so far as I can see, of most of the major objections that critics have advanced against his theory. It 182

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

is necessary, therefore, to examine this problem in some detail. Perhaps we should begin by noting that Tillich does offer interpretations of religious statements about God in terms of ontological statements about being-itself. Indeed, he thinks it is the business of theology to do so. "Theology should not weaken the concrete symbols, but it must analyze them and interpret them in abstract ontological terms'" (ST, 1:242). Two examples of his efforts in this direction are as follows: In popular parlance the concept "omnipotence" implies a highest being who is able to do whatever he wants. This notion must be rejected, religiously as well as theologically. . . . It is more adequate to define omnipotence as the power of being which resists nonheing in all its expressions and which is manifest in the creative process in all its forms (ST, 1:273). Life is the actuality of being, or, more exactly, it is the process in which potential being becomes actual being. But in God as God there is no distinction between potentiality and actuality. Therefore, we cannot speak of God as living in the proper or nonsymbolic sense of the word life. We must speak of God as living in symbolic terms. . . . God lives in so far as he is the ground of life (ST, 1:242).

Concerning these and other interpretations Tillich's critics take the view that he intends to be offering nonsymbolic translations of religious statements. For example, Alston remarks: "I do not know how to read this other than as an attempt to translate symbolic language into nonsymbolic language/'5 Another critic, Edwards, argues: Tillich, the dogmatist, does not hesitate to offer translations or what I have called reductions of his "symbolic" statements about God. We can also express literally, for example, what we mean "sym5

Alston, Religious Experience and Truth, p. 25. 183

47

48

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION bolically" when we say that God is living. "God lives," the reduction runs, ''insofar as he is the ground of life."6 . . . although Tillich gives the impression that the metaphors have been eliminated in these and similar cases, this is not so. He never seems to have noticed that even in his basic statement, when elaborated in terms of "ground" and "structure", these words are used metaphorically and not literally.7

Against these criticisms, especially Edward's, several points need to be made. It must be noted that Tillich nowhere says that the interpretation or translation is itself literal. Indeed, to hold such a view would be inconsistent with his explicit claim that every statement about being-itself is symbolic or metaphorical. It is not Tillich but Edwards who gives the 'impression' that Tillich's view is that the translation eliminates the metaphor in the religious statement 'God is living'. It is true, I think, that Tillich would regard the statement (a) 'being-itself is the ground of life' as an essential part of the interpretation of the religious statement (b) 'God is living'. What is not true is that he regards a as a translation of b such that a expresses in literal terms what b expresses symbolically. Having misrepresented Tillich as holding that 'God (being-itself) is the ground of life' is a literal statement, Edwards then correctly points out that 'ground' is being used metaphorically and not literally. As I have presented matters, this is what should be expected. On Tillich's own theory, no statement about being-itself can be nonsymbolic. Hence, on Tillich's theory, it is to be expected that 6Paul Edwards, "Professor Tillich's Confusions/' Mind, 74 (April, 1965): 203-4.

7 Ibid., p. 204. 184

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

the interpretation of the religious statement 'God is living' will itself involve a symbolic statement about being-itself. Edward's claim that Tillich "never seems to have noticed" that in the interpretation 'ground' is being used metaphorically is, I am afraid, simply mistaken. Tillich remarks in his "Reply to Interpretation and Criticism": "... I agree that 'ground' and 'power' of being are symbolic notions, in so far as they use elements of being (power, cause) in order to circumscribe being-itself."8 Alston, unlike Edwards, is careful to acknowledge that Tillich might intend the ontological interpretations of theological statements to be themselves symbolic. Thus, after quoting Tillich's interpretation of theological statements ascribing 'will' and 'intellect' to God as "symbols for dynamics in all its ramifications and for form as the meaningful structure of being-itself," Alston remarks concerning these ontological interpretations: If they were intended to be simply the replacement of one symbol by another symbol, they would be grotesque failures. No one would suppose that "dynamics in all its ramifications" is a better religious symbol, i.e., performs a symbolic function better or more clearly than "will/'9

What I have been arguing is that Tillich's ontological interpretations of religious statements are, on his own theory, symbolic statements. Admitting that point, we must now see what can be said concerning Alston's criticism. First, it must be admitted that "dynamics in all its ramifications" is not a better religious symbol than "will". It is not a better 8

Charles W. Kegley and Robert W. Bretall, eds., The Theology of Paul Tillich (New York: MacmiUan, 1959), p. 335. Also see ST, 1: 156. 9

Alston, Religious Experience and Truth, p. 25.

185

49

50

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

religious symbol because it is not a religious symbol at all. To speak about being-itself in terms of "dynamics in all its ramifications*' is to make an ontological statement, not a religious statement. To add that "dynamics" is being used metaphorically or symbolically is not to imply that the ontological statement is a religious statement. However, Alston goes on to add that no one would suppose that "dynamics in all its ramifications" "performs a symbolic function better or more clearly than 'will'." Certainly this is true of most religious men; indeed it may be true of all those without some training in metaphysics. But I think we have to ask the question, "better for what purpose?" If the purpose is to marshal religious feelings and actions, to provide concrete images to function as foci for ultimate concern, then there is no question but that "will" is a better symbol than "dynamics in all its ramifications". It would be absurd to think otherwise. However, if the purpose is to understand reality, to get a conceptual grasp, a more accurate picture, of the reality lying behind the concrete imagery of religious symbolism, then perhaps "dynamics in all its ramifications" is a better symbol than "will". Tillich, of course, is not suggesting that the symbolic concepts of his ontology are better for religious purposes than the symbolic expressions of religious discourse. He is, I am suggesting, maintaining that they are better if our purpose is to describe the nature of ultimate reality. If Alston intends to claim that they are inferior even for this latter purpose, he needs to support his claim by arguments. These remarks in defense of Tillich leave untouched, however, what I have argued is the central problem in Tillich's account of linguistic religious symbols. On his theory it seems impossible to break out of the web of sym186

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

bols. A religious statement S about God is interpreted by Tillich in terms of an ontological statement S' about beingitself. But, since no nonsymbolic positive statement about being-itself is possible on Tillich's theory, S' must also be symbolic. Now, if to interpret or explain what is meant by a symbolic or metaphorical statement S is to produce another statement S' which expresses in nonsymbolic terms essentially what S expresses in symbolic terms, it follows that it is impossible on Tillich's theory to ever explain or interpret any religious statement about God. This conclusion, of course, is disastrous to Tillich's entire theory of religious symbols. We must now see what can be done to avoid it. There are, I believe, only two things that can be done. One is to abandon the view that being-itself can be described only in symbolic terms. However, since this would result in far reaching changes in Tillich's ontology, as well as in his account of religious symbols, I shall not pursue this possibility. The second is to simply deny the semantic principle that led to the disastrous conclusion. Apart from denying this principle, I see no way in which Tillich's theory can remain intact and avoid the conclusion that it is impossible to interpret or understand any religious statement about God. It is this second possibility that I shall now explore. What I have called the semantic principle may take various forms. One quite strong form of it is as follows: A symbolic, or metaphorical, statement S is meaningful only if what it expresses can be expressed by some nonsymbolic statement S'. As it stands, however, this principle is quite likely false. For it is very doubtful that a literal statement is completely synonymous with the metaphorical state187

51

52

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

ment it translates. What, for example, would be the literal statement which expresses exactly what is expressed by e. e. cummings' metaphorical line, "the sweet small clumsy feet of april came into the ragged meadows of my soul"?10 Metaphorical statements frequently express nuances of feeling, etc., that their literal interpretations fail to express. Hence, if our principle is not to be obviously false, some restriction must be placed on what the literal statement is expected to express in common with the symbolic statement it purports to translate. Perhaps it will suffice if we understand the principle to require only that the cognitive claims made by the metaphorical statement must be expressible by some nonsymbolic statement, or set of such statements. The symbolic statement, for example, "John is a lion" may express more than the literal statement "John is brave," but it can be argued that the latter suffices to express the cognitive import of the former. Thus we may understand the principle to assert that a symbolic statement which purports to make cognitive claims is meaningful only if those cognitive claims can be made by some nonsymbolic statement. If we understand the principle in this fashion, it is clear, I think, that Tillich's theory of religious symbols is inconsistent with it. That this is so can be most easily shown in terms of an example. Earlier, we noted that T)eing-itself is the ground of life' is an ontological statement and is, for Tillich, an essential part, at least, of the correct interpretation of the symbolic, religious statement 'God is living/ Now Tillich is committed to the view that the ontological statement ^being-itself is the ground of life* (a) is symbolic and 10 Quoted in William P. Alston, Philosophy of Language (Englewood Cliffs, N.J.: Prentice-Hall, 1964), p. 96.

188

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

(b) possesses cognitive content. But, given our principle, it would follow that this ontological statement is meaningful only if the cognitive content it purports to express can be expressed by some nonsymbolic statement. But to express precisely that cognitive content the nonsymbolic statement would surely have to be a statement about being-itself. The latter sort of statement—i.e., a nonsymbolic statement about being-itself—is not, however, possible for Tillich. Hence, given our principle it will follow that the ontological statements about being-itself, and the religious statements about God they purport to interpret, are meaningless. But such a conclusion is surely inconsistent with Tillich's view of religious statements about God and ontological statements about being-itself. I have been arguing thus far that Tillich's theory of linguistic, religious symbolism is inconsistent with the semantic principle that a symbolic or metaphorical statement is meaningful only if its cognitive content is expressible by some literal statement. If we like we may call those symbolic statements whose cognitive content can be expressed by literal statements "reducible symbolic statements"; those symbolic statements whose purported cognitive content cannot be expressed by literal statements we may call "irreducible symbolic statements/' Tillich's theory, as I have presented it, implies that there are meaningful, irreducible symbolic statements. Our principle implies that irreducible symbolic statements are not meaningful. If our principle is correct, Tillich's theory is incorrect. If Tillich's theory is correct, our principle is incorrect. A third possibility, of course, is that both the principle and the theory are incorrect. The correctness or incorrectness of the semantic prin189

53

54

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

ciple we have been discussing is an issue that cannot be settled here, involving, as it does, complicated and controversial issues in the philosophy of language. However, some general remarks concerning its employment against Tillich's theory may be helpful. It is evident that a principle of the sort we have been considering is made much of in discussions of religious symbolism. For example, in his admirable book on mysticism, W. T. Stace, in the course of objecting to what he calls "the metaphorical theory" of the mystic's descriptions of mystical experiences, uses—without so much as an argument in its behalf—a rough version of our principle which he expresses as, "metaphorical language is only meaningful and justifiable if it is at least theoretically translatable into literal language; .. ,"n And Edwards, in the course of criticizing Tillich, holds that a metaphorical statement is intelligible only if "the truth-claims made by the sentence in which it occurs can be reproduced by one or more sentences all of whose components are used in literal senses/'12 'Truth-claims' and 'cognitive content' can be taken here as coming to the same thing. Thus he says, "When a sentence contains an irreducible metaphor, it follows at once that the sentence is devoid of cognitive meaning, that it is unintelligible, that it fails to make a genuine assertion."13 In an interesting historical remark Edwards notes that Berkeley's objection to Locke's talk about an unknowable, material substratum that "supports" sense-qualities amounted to the claim that Locke's talk was unintelligible, 11 W. T. Stace, Mysticism and Philosophy (New York: J. B. Lippincott, 1960), p. 293. 12 Edwards, Mind, p. 199. 190

« Ibid., pp. 199-200.

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

involving the irreducible metaphor "supports/'14 Edwards then proceeds to suggest ways in which Locke could have answered Berkeley's objection. The second way Edwards suggests is very pertinent to his claim that Tillich's talk of being-itself involves "irreducible metaphors/' I quote it in full: A second line of defense would begin by admitting that the material substratum would be completely unknowable, if sensory observation were the only method of becoming acquainted with objective realities. In fact, however, it would be said, we possess a "super-sensuous" faculty with which we "experience" such realities as material and spiritual substances. We could, if we wanted, introduce a set of terms as the symbols literally referring to the data disclosed by this super-sensuous faculty and we could exchange information about these with all who share in the possession of the faculty. If we call this the "intellectual language", then, so this defense of Locke would run, sentences with metaphors when containing terms from the "sensory level", can be translated into sentences in the intellectual language which will be free from metaphors.15

Edwards here appears to hold that so long as the referent of our metaphorical descriptions is something we experience, our metaphorical descriptions are not irreducible because, he argues, we could, if we wanted, introduce some terms as literally referring to what is experienced. Hence, we could then translate our metaphorical statements into statements involving these new terms. It is, I think, doubtful that Edwards is right about this point. But what, perhaps, is more important for our purpose is that if he is right, then so far as Tillich's view is concerned it follows that Tillich's metaphorical descriptions of being-itself are not irreducible. For, as we have seen, Tillich surely i* Ibid., p. 200.

16 Hid., pp. 200-201.

191

55

56

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

holds that we do experience the ultimate, being-itself. In fact, it is a rather fundamental claim in his ontology that man has an immediate awareness of what is ontologically ultimate, being-itself. Surprisingly, Edwards seems completely unaware of this rather fundamental claim in Tillich. Indeed, he succeeds in getting Tillich exactly wrong when he attributes to him the view that". .. Being-itself is, even in principle, inaccessible to anybody's observation/'16 In fact, once Tillich's actual position—that being-itself is experienced—is recognized, it follows on Edwards9 view that Tillich's metaphors are not irreducible and, therefore, that Edwards' conclusion is erroneous. But, is Edwards' view that metaphorical descriptions of what is experienced are reducible metaphors correct? Although this question cannot be pursued here, we must at least note that it is not obviously correct; indeed, views opposed to it are currently held by some philosophers of language. Alston, for example, argues that not only God but also many of our inner feelings—which obviously meet Edwards' condition of being experienced—are such that their metaphorical descriptions are irreducible in the sense that what they express cannot be expressed in literal terms.17 So far as statements about God are concerned, Tillich would certainly agree with Alston. And if Alston is right concerning our inner feelings, then we have prima facie grounds, at least, for rejecting not only Edwards' view but also the semantic principle which states that a symbolic or metaphorical statement is meaningful only if its cognitive content is expressible by some literal statement. 16

Ibid., p. 201. It is clear from the context that Edwards does not intend to restrict "observation" to sensory experience. it Alston, Philosophy of Language, pp. 103-6.

192

THE PHILOSOPHICAL THEOLOGY OF PAUL TILLICH

In the past few pages I have been concerned mainly to argue (1) that Tillich's own view is that religious statements about God are to be interpreted in terms of ontological statements about being-itself, (2) that Tillich's own view implies that the ontological statements about beingitself, as well as the religious statements about God they purport to interpret, are irreducible symbolic statements, and (3) that Tillich's theory is, therefore, inconsistent with the principle that a symbolic statement is meaningful only if its cognitive content is expressible by some literal statement. Before we return to our account of primary religious symbols, one final problem needs discussion. Since statements about being-itself are, on Tillich's view, irreducible symbolic statements, and since he proposes to interpret religious statements about God in terms of ontological statements about being-itself, the question arises as to how the ontological statements can perform the job of explanation and interpretation that Tillich assigns to them. Since they too are symbolic it would seem that they are as much in need of elucidation as the theological statements about God they purport to interpret. In discussing a criticism by Alston, I suggested that even though the ontological statements are symbolic, the symbols they employ may serve a different purpose than the symbols employed in the religious statements which the ontological statements interpret. That is, if our purpose is to arrive at a conceptual grasp of the nature of ultimate reality, the symbols occurring in the ontological statements may be better suited to serve this purpose than the symbols occurring in the corresponding religious statements about God. Thus, from the vantage point of this purpose it may not be true that the 193

57

58

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION

symbolic ontological statements are as much in need of elucidation as the religious statements they purport to elucidate. What Tillich is committed to holding is that a proposed explication of a religious symbol or myth need not be completely nonsymbolic in order to be an intelligible interpretation of that symbol or myth. However, it is consistent with this to hold that the interpretation must be less symbolic than the symbol or myth it interprets. And it is important to note that Tillich is sufficiently aware of the problem we are discussing to point out that although his ontological interpretations are symbolic, they are less symbolic than the data they interpret. Thus we have here a second reason why it may not be true that the symbolic, ontological statements are as much in need of elucidation as the religious statements they purport to interpret.

194

PART II THE PROBLEM OF EVIL

This page intentionally left blank

CHAPTER THREE

The Problem of Evil and Some Varieties of Atheism 'T'HIS paper is concerned with three interrelated •*• questions. The first is: Is there an argument for atheism based on the existence of evil that may rationally justify someone in being an atheist? To this first question I give an affirmative answer and try to support that answer by setting forth a strong argument for atheism based on the existence of evil.1 The second question is: How can the theist best defend his position against the argument for atheism based on the existence of evil? In response to this question I try to describe what may be an adequate rational defense for theism against any argument for atheism based on the existence of evil. The final question is: What position should the informed atheist take concerning the rationality of theistic belief? Three different answers an atheist may give to this question serve to distinguish three varieties of atheism: unfriendly atheism, indifferent atheism, and friendly atheism. In the final part of the paper I discuss and defend the position of friendly atheism. Before we consider the argument from evil, we need to distinguish a narrow and a broad sense of the terms "theist/' "atheist," and "agnostic." By a "theist" in the narrow sense I mean someone who believes in the existence of an omnipotent, omniscient, eternal, supremely good being who created the world. By a "theist" in the broad sense I mean someone who believes in the existence of some sort of divine being or divine reality. To be a theist in the narrow sense is also to be a theist in the broad sense, but one may be a theist in the broad sense—as was Paul Tillich—without believing that there is a supremely good, omnipotent, omniscient, eternal being who created the world. Similar distinctions must be made between a narrow and a broad sense of the terms "atheist" and "agnostic." To be an atheist in the broad sense is to deny the existence of any sort

of divine being or divine reality. Tillich was not an atheist in the broad sense. But he was an atheist in the narrow sense, for he denied that there exists a divine being that is all-knowing, all-powerful and perfectly good. In this paper I will be using the terms "theism," "theist," "atheism," "atheist," "agnosticism," and "agnostic" in the narrow sense, not in the broad sense. I

In developing the argument for atheism based on the existence of evil, it will be useful to focus on some particular evil that our world contains in considerable abundance. Intense human and animal suffering, for example, occurs daily and in great plenitude in our world. Such intense suffering is a clear case of evil. Of course, if the intense suffering leads to some greater good, a good we could not have obtained without undergoing the suffering in question, we might conclude that the suffering is justified, but it remains an evil nevertheless. For we must not confuse the intense suffering in and of itself with the good things to which it sometimes leads or of which it may be a necessary part. Intense human or animal suffering is in itself bad, an evil, even though it may sometimes be justified by virtue of being a part of, or leading to, some good which is unobtainable without it. What is evil in itself may sometimes be good as a means because it leads to something that is good in itself. In such a case, while remaining an evil in itself, the intense human or animal suffering is, nevertheless, an evil which someone might be morally justified in permitting. Taking human and animal suffering as a clear instance of evil which occurs with great frequency in

1 Some philosophers have contended that the existence of evil is logically inconsistent with the existence of the theistic God. No one, I think, has succeeded in establishing such an extravagant claim. Indeed, granted incompatibilism, there is a fairly compelling argument for the view that the existence of evil is logically consistent with the existence of the theistic God. (For a lucid statement of this argument see Alvin Plantinga, God, Freedom, and Evil (New York, 1974), pp. 29-59.) There remains, however, what we may call the evidential form—as opposed to the logical form—of the problem of evil: the view that the variety and profusion of evil in our world, although perhaps not logically inconsistent with the existence of the theistic God, provides, nevertheless, rational support for atheism. In this paper I shall be concerned solely with the evidential form of the problem, the form of the problem which, I think, presents a rather severe difficulty for theism.

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

62 336

A M E R I C A N PHILOSOPHICAL QUARTERLY

our world, the argument for atheism based on evil can be stated as follows: 1. There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse.2 2. An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse. 3. There does not exist an omnipotent, omniscient, wholly good being.

What are we to say about this argument for atheism, an argument based on the profusion of one sort of evil in our world? The argument is valid; therefore, if we have rational grounds for accepting its premises, to that extent we have rational grounds for accepting atheism. Do we, however, have rational grounds for accepting the premises of this argument? Let's begin with the second premise. Let st be an instance of intense human or animal suffering which an omniscient, wholly good being could prevent. We will also suppose that things are such that sl will occur unless prevented by the omniscient, wholly good (OG) being. We might be interested in determining what would be a sufficient condition of OG failing to prevent s1. But, for our purpose here, we need only try to state a necessary condition for OG failing to prevent jj. That condition, so it seems to me, is this: Either (i) there is some greater good, G, such that G is obtainable by OG only if OG permits s^, or (ii) there is some greater good, G, such that G is obtainable by OG only if OG permits either ^ or some evil equally bad or worse, or (iii) sl is such that it is preventable by OG only if OG permits some evil equally bad or worse. It is important to recognize that (iii) is not included in (i). For losing a good greater than ^ is

not the same as permitting an evil greater than s^ And this because the absence of a good state of affairs need not itself be an evil state of affairs. It is also important to recognize that sl might be such that it is preventable by OG without losing G (so condition (i) is not satisfied) but also such that if OG did prevent it, G would be loss unless OG permitted some evil equal to or worse than sl. If this were so, it does not seem correct to require that OG prevent sl. Thus, condition (ii) takes into account an important possibility not encompassed in condition (i). Is it true that if an omniscient, wholly good being permits the occurrence of some intense suffering it could have prevented, then either (i) or (ii) or (iii) obtains? It seems to me that it is true. But if it is true then so is premise (2) of the argument for atheism. For that premise merely states in more compact form what we have suggested must be true if an omniscient, wholly good being fails to prevent some intense suffering it could prevent. Premise (2) says that an omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse. This premise (or something not too distant from it) is, I think, held in common by many atheists and nontheists. Of course, there may be disagreement about whether something is good, and whether, if it is good, one would be morally justified in permitting some intense suffering to occur in order to obtain it. Someone might hold, for example, that no good is great enough to justify permitting an innocent child to suffer terribly.4 Again, someone might hold that the mere fact that a given good outweighs some suffering and would be loss if the suffering were prevented, is not a morally sufficient reason for permitting the suffering. But to hold either of these views is not to deny (2). For (2) claims only that if&n omniscient, wholly good being permits intense suffering then either there is some greater good that would have been loss, or some

2 If there is some good. G, greater than any evil, (i) will be false for the trivial reason that no matter what evil E, we pick the conjunctive good state of affairs consisting of G and E will outweigh £and be such that an omnipotent being could not obtain it without permitting E. (See Alvin Plantinga, God and Other Minds [Ithaca, 1967], p. 167.) To avoid this objection we may insert "unreplaceable" into our premises (i) and (2) between "some" and "greater." If E isn't required for G, and G is better than G plus E, then the good conjunctive state of affairs composed of G and E would be replaceable by the greater good of G alone. For the sake of simplicity, however, I will ignore this complication both in the formulation and discussion of premises (i) and (2). 3 Three clarifying points need to be made in connection with (i). First, by "good" I don't mean to exclude the fulfillment of certain moral principles. Perhaps preventing sl would preclude certain actions prescribed by the principles of justice. I shall allow that the satisfaction of certain principles of justice may be a good that outweighs the evil o f s l . Second, even though (i) may suggest it, I don't mean to limit the good in question to something that wouldfollow in time the occurrence of Jr And, finally, we should perhaps not fault OG if the good G, that would be loss were s1 prevented, is not actually greater than sl, but merely such that allowing Jt and G, as opposed to preventing $, and thereby losing G, would not alter the balance between good and evil. For reasons of simplicity, I have left this point out in stating (i), with the result that (i) is perhaps a bit stronger than it should be. 4 See Ivan's speech in Book V, Chapter IV of The Brothers Karamazov.

THE PROBLEM OF EVIL THE P R O B L E M OF EVIL AND SOME VARIETIES OF ATHEISM

equally bad or worse evil that would have occurred, had the intense suffering been prevented. (2) does not purport to describe what might be a sufficient condition for an omniscient, wholly good being to permit intense suffering, only what is a necessary condition. So stated, (2) seems to express a belief that accords with our basic moral principles, principles shared by both theists and nontheists. If we are to fault the argument for atheism, therefore, it seems we must find some fault with its first premise. Suppose in some distant forest lightning strikes a dead tree, resulting in a forest fire. In the fire a fawn is trapped, horribly burned, and lies in terrible agony for several days before death relieves its suffering. So far as we can see, the fawn's intense suffering is pointless. For there does not appear to be any greater good such that the prevention of the fawn's suffering would require either the loss of that good or the occurrence of an evil equally bad or worse. Nor does there seem to be any equally bad or worse evil so connected to the fawn's suffering that it would have had to occur had the fawn's suffering been prevented. Could an omnipotent, omniscient being have prevented the fawn's apparently pointless suffering? The answer is obvious, as even the theist will insist. An omnipotent, omniscient being could have easily prevented the fawn from being horribly burned, or, given the burning, could have spared the fawn the intense suffering by quickly ending its life, rather than allowing the fawn to lie in terrible agony for several days. Since the fawn's intense suffering was preventable and, so far as we can see, pointless, doesn't it appear that premise (i) of the argument is true, that there do exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. It must be acknowledged that the case of the fawn's apparently pointless suffering does not prove that (i) is true. For even though we cannot see how the fawn's suffering is required to obtain some greater good (or to prevent some equally bad or worse evil), it hardly follows that it is not so required. After all, we are often surprised by how things we thought to be unconnected turn out to be intimately connected. Perhaps, for all we know, there is some familiar good outweighing the fawn's suffering to which that suffering is connected in a way we do not see. Furthermore, there may well be unfamiliar goods, goods we haven't dreamed of, to which the fawn's suffering is inextricably connected. Indeed, it would seem to require something like omniscience

63 337

on our part before we could lay claim to knowing that there is no greater good connected to the fawn's suffering in such a manner than an omnipotent, omniscient being could not have achieved that good without permitting that suffering or some evil equally bad or worse. So the case of the fawn's suffering surely does not enable us to establish the truth of (i). The truth is that we are not in a position to prove that (i) is true. We cannot know with certainty that instances of suffering of the sort described in (i) do occur in our world. But it is one thing to know or prove that (i) is true and quite another thing to have rational grounds for believing (i) to be true. We are often in the position where in the light of our experience and knowledge it is rational to believe that a certain statement is true, even though we are not in a position to prove or to know with certainty that the statement is true. In the light of our past experience and knowledge it is, for example, very reasonable to believe that neither Goldwater nor McGovern will ever be elected President, but we are scarcely in the position of knowing with certainty that neither will ever be elected President. So, too, with (i), although we cannot know with certainty that it is true, it perhaps can be rationally supported, shown to be a rational belief. Consider again the case of the fawn's suffering. Is it reasonable to believe that there is some greater good so intimately connected to that suffering that even an omnipotent, omniscient being could not have obtained that good without permitting that suffering or some evil at least as bad? It certainly does not appear reasonable to believe this. Nor does it seem reasonable to believe that there is some evil at least as bad as the fawn's suffering such that an omnipotent being simply could not have prevented it without permitting the fawn's suffering. But even if it should somehow be reasonable to believe either of these things of the fawn's suffering, we must then ask whether it is reasonable to believe either of these things of all the instances of seemingly pointless human and animal suffering that occur daily in our world. And surely the answer to this more general question must be no. It seems quite unlikely that all the instances of intense suffering occurring daily in our world are intimately related to the occurrence of greater goods or the prevention of evils at least as bad; and even more unlikely, should they somehow all be so related, than an omnipotent, omniscient being could not have achieved at least some of those goods (or prevented some of those evils) without permitting the instances of intense suffering that are

64

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

338

AMERICAN PHILOSOPHICAL QUARTERLY

supposedly related to them. In the light of our experience and knowledge of the variety and scale of human and animal suffering in our world, the idea that none of this suffering could have been prevented by an omnipotent being without thereby losing a greater good or permitting an evil at least as bad seems an extraordinary absurd idea, quite beyond our belief. It seems then that although we cannot prove that (i) is true, it is, nevertheless, altogether reasonable to believe that (i) is true, that (i) is a rational belief.5 Returning now to our argument for atheism, we've seen that the second premise expresses a basic belief common to many theists and nontheists. We've also seen that our experience and knowledge of the variety and profusion of suffering in our world provides rational support for the first premise. Seeing that the conclusion, "There does not exist an omnipotent, omniscient, wholly good being" follows from these two premises, it does seem that we have rational support for atheism, that it is reasonable for us to believe that the theistic God does not exist. II

Can theism be rationally defended against the argument for atheism we have just examined? If it can, how might the theist best respond to that argument? Since the argument from (i) and (2) to (3) is valid, and since the theist, no less than the nontheist, is more than likely committed to (2), it's clear that the theist can reject this atheistic argument only by rejecting its first premise, the premise that states that there are instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. How, then, can the theist best respond to this premise and the considerations advanced in its support? There are basically three responses a theist can make. First, he might argue not that (i) is false or

probably false, but only that the reasoning given in support of it is in some way defective. He may do this either by arguing that the reasons given in support of (i) are in themselves insufficient to justify accepting (i), or by arguing that there are other things we know which, when taken in conjunction with these reasons, do not justify us in accepting (i). I suppose some theists would be content with this rather modest response to the basic argument for atheism. But given the validity of the basic argument and the theist's likely acceptance of (2), he is thereby committed to the view that (i) is false, not just that we have no good reasons for accepting (i) as true. The second two responses are aimed at showing that it is reasonable to believe that (i) is false. Since the theist is committed to this view I shall focus the discussion on these two attempts, attempts which we can distinguish as "the direct attack" and "the indirect attack." By a direct attack, I mean an attempt to reject (i) by pointing out goods, for example, to which suffering may well be connected, goods which an omnipotent, omniscient being could not achieve without permitting suffering. It is doubtful, however, that the direct attack can succeed. The theist may point out that some suffering leads to moral and spiritual development impossible without suffering. But it's reasonably clear that suffering often occurs in a degree far beyond what is required for character development. The theist may say that some suffering results from free choices of human beings and might be preventable only by preventing some measure of human freedom. But, again, it's clear that much intense suffering occurs not as a result of human free choices. The general difficulty with this direct attack on premise (i) is twofold. First, it cannot succeed, for the theist does not know what greater goods might be served, or evils prevented, by each instance of intense human or animal suffering. Second, the theist's own religious tradition usually maintains that in this life it is not given to us to know God's

5 One might object that the conclusion of this paragraph is stronger than the reasons given warrant. For it is one thing to argue that it is unreasonable to think that (i) is false and another thing to conclude that we are therefore justified in accepting (i) as true. There are propositions such that believing them is much more reasonable than disbelieving them, and yet are such that withholdingjudgment about them is more reasonable than believing them. To take an example of Chisholm's: it is more reasonable to believe that the Pope will be in Rome (on some arbitrarily picked future date) than to believe that he won't; but it is perhaps more reasonable to suspend judgment on the question of the Pope's whereabouts on that particular date, than to believe that he will be in Rome. Thus, it might be objected, that while we've shown that believing (i) is more reasonable than disbelieving (i), we haven't shown that believing (i) is more reasonable than withholding belief. My answer to this objection is that there are things we know which render (i) probable to the degree that it is more reasonable to believe (i) than to suspend judgment on (i). What are these things we know ? First, I think, is the fact that there is an enormous variety and profusion of intense human and animal suffering in our world. Second, is the fact that much of this suffering seems quite unrelated to any greater goods (or the absence of equal or greater evils) that might justify it. And, finally, there is the fact that such suffering as is related to greater goods (or the absence of equal or greater evils) does not, in many cases, seem so intimately related as to require its permission by an omnipotent being bent on securing those goods (the absence of those evils). These facts, I am claiming, make it more reasonable to accept (i) than to withhold judgment on (i).

65

THE PROBLEM OF EVIL THE P R O B L E M OF EVIL AND SOME VARIETIES OF ATHEISM

purpose in allowing particular instances of suffering. Hence, the direct attack against premise (i) cannot succeed and violates basic beliefs associated with theism. The best procedure for the theist to follow in rejecting premise (i) is the indirect procedure. This procedure I shall call "the G. E. Moore shift/' socalled in honor of the twentieth century philosopher, G. E. Moore, who used it to great effect in dealing with the arguments of the skeptics. Skeptical philosophers such as David Hume have advanced ingenious arguments to prove that no one can know of the existence of any material object. The premises of their arguments employ plausible principles, principles which many philosophers have tried to reject directly, but only with questionable success. Moore's procedure was altogether different. Instead of arguing directly against the premises of the skeptic's arguments, he simply noted that the premises implied, for example, that he (Moore) did not know - of the existence of a pencil. Moore then proceeded indirectly against the skeptic's premises by arguing: I do know that this pencil exists. If the skeptic's principles are correct I cannot know of the existence of this pencil. .'. The skeptic's principles (at least one) must be incorrect.

Moore then noted that his argument is just as valid as the skeptic's, that both of their arguments contain the premise "If the skeptic's principles are correct Moore cannot know of the existence of this pencil," and concluded that the only way to choose between the two arguments (Moore's and the skeptic's) is by deciding which of the first premises it is more rational to believe—Moore's premise "I do know that this pencil exists" or the skeptic's premise asserting that his skeptical principles are correct. Moore concluded that his own first premise was the more rational of the two.6 Before we see how the theist may apply the G. E. Moore shift to the basic argument for atheism, we should note the general strategy of the shift. We're given an argument: />, #, therefore, r. Instead of arguing directly against />, another argument is constructed—not-r, q, therefore, not-/»—which begins with the denial of the conclusion of the first argument, keeps its second premise, and ends with the denial of the first premise as its conclusion. 6

339

Compare, for example, these two: I. p

II. not-r

r

not-p

It is a truth of logic that If I is valid II must be valid as well. Since the arguments are the same so far as the second premise is concerned, any choice between them must concern their respective first premises. To argue against the first premise (p) by constructing the counter argument II is to employ the G. E. Moore shift. Applying the G. E. Moore shift against the first premise of the basic argument for atheism, the theist can argue as follows: not-3. There exists an omnipotent, omniscient, wholly good being. 2. An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse, therefore, not-1. It is not the case that there exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. We now have two arguments: the basic argument for atheism from (i) and (2) to (3), and the theist's best response, the argument from (not-3) and (2) to (not-1). What the theist then says about (i) is that he has rational grounds for believing in the existence of the theistic God (not-3), accepts (2) as true, and sees that (not-i) follows from (not-3) an(* ( 2 )- HC concludes, therefore, that he has rational grounds for rejecting (i). Having rational grounds for rejecting (i), the theist concludes that the basic argument for atheism is mistaken.

Ill We've had a look at a forceful argument for atheism and what seems to be the theist's best response to that argument. If one is persuaded by the argument for atheism, as I find myself to be, how might one best view the position of the theist. Of course, he will view the theist as having a false belief,

See, for example, the two chapters on Hume in G. E. Moore, Some Main Problems of Philosophy (London, 1953).

66

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

340

AMERICAN P H I L O S O P H I C A L QUARTERLY

just as the theist will view the atheist as having a false belief. But what position should the atheist take concerning the rationality of the theist's belief? There are three major positions an atheist might take, positions which we may think of as some varieties of atheism. First, the atheist may believe that no one is rationally justified in believing that the theistic God exists. Let us call this position "unfriendly atheism." Second, the atheist may hold no belief concerning whether any theist is or isn't rationally justified in believing that the theistic God exists. Let us call this view "indifferent atheism." Finally, the atheist may believe that some theists are rationally justified in believing that the theistic God exists. This view we shall call "friendly atheism." In this final part of the paper I propose to discuss and defend the position of friendly atheism. If no one can be rationally justified in believing a false proposition then friendly atheism is a paradoxical, if not incoherent position. But surely the truth of a belief is not a necessary condition of someone's being rationally justified in having that belief. So in holding that someone is rationally justified in believing that the theistic God exists, the friendly atheist is not committed to thinking that the theist has a true belief. What he is committed to is that the theist has rational grounds for his belief, a belief the atheist rejects and is convinced he is rationally justified in rejecting. But is this possible? Can someone, like our friendly atheist, hold a belief, be convinced that he is rationally justified in holding that belief, and yet believe that someone else is equally justified in believing the opposite? Surely this is possible. Suppose your friends see you off on a flight to Hawaii. Hours after take-off they learn that your plane has gone down at sea. After a twenty-four hour search, no survivors have been found. Under these circumstances they are rationally justified in believing that you have perished. But it is hardly rational for you to believe this, as you bob up and down in your life vest, wondering why the search planes have failed to spot you. Indeed, to amuse yourself while awaiting your fate, you might very well reflect on the fact that your friends are rationally justified in believing that you are now dead, a proposition you disbelieve and are rationally justified in disbelieving. So, too, perhaps an atheist may be rationally justified in his atheistic belief and yet hold that some theists are rationally justified in believing just the opposite of what he believes. What sort of grounds might a theist have for believing that God exists. Well, he might endeavor

to justify his belief by appealing to one or more of the traditional arguments: Ontological, Cosmological, Teleological, Moral, etc. Second, he might appeal to certain aspects of religious experience, perhaps even his own religious experience. Third, he might try to justify theism as a plausible theory in terms of which we can account for a variety of phenomena. Although an atheist must hold that the theistic God does not exist, can he not also believe, and be justified in so believing, that some of these "justifications of theism" do actually rationally justify some theists in their belief that there exists a supremely good, omnipotent, omniscient being? It seems to me that he can. If we think of the long history of theistic belief and the special situations in which people are sometimes placed, it is perhaps as absurd to think that no one was ever rationally justified in believing that the theistic God exists as it is to think that no one was ever justified in believing that human being would never walk on the moon. But in suggesting that friendly atheism is preferable to unfriendly atheism, I don't mean to rest the case on what some human beings might reasonably have believed in the eleventh or thirteenth century. The more interesting question is whether some people in modern society, people who are aware of the usual grounds for belief and disbelief and are acquainted to some degree with modern science, are yet rationally justified in accepting theism. Friendly atheism is a significant position only if it answers this question in the affirmative. It is not difficult for an atheist to be friendly when he has reason to believe that the theist could not reasonably be expected to be acquainted with the grounds for disbelief that he (the atheist) possesses. For then the atheist may take the view that some theists are rationally justified in holding to theism, but would not be so were they to be acquainted with the grounds for disbelief—those grounds being sufficient to tip the scale in favor of atheism when balanced against the reasons the theist has in support of his belief. Friendly atheism becomes paradoxical, however, when the atheist contemplates believing that the theist has all the grounds for atheism that he, the atheist, has, and yet is rationally justified in maintaining his theistic belief. But even so excessively friendly a view as this perhaps can be held by the atheist if he also has some reason to think that the grounds for theism are not as telling as the theist is justified in taking them to be.7 7 Suppose that I add a long sum of numbers three times and get result x. I inform you of this so that you have pretty much the same

THE PROBLEM OF EVIL THE P R O B L E M OF E V I L AND SOME V A R I E T I E S OF ATHEISM

In this paper I've presented what I take to be a strong argument for atheism, pointed out what I think is the theist's best response to that argument, distinguished three positions an atheist might take concerning the rationality of theistic belief, and made some remarks in defense of the position called "friendly atheism." I'm aware that the central points of the paper are not likely to be warmly

67 341

received by many philosophers. Philosophers who are atheists tend to be tough minded—holding that there are no good reasons for supposing that theism is true. And theists tend either to reject the view that the existence of evil provides rational grounds for atheism or to hold that religious belief has nothing to do with reason and evidence at all. But such is the way of philosophy.8

evidence I have for the claim that the sum of the numbers is x. You then use your calculator twice over and arrive at resulty. You, then, are justified in believing that the sum of the numbers is not x. However, knowing that your calculator has been damaged and is therefore unreliable, and that you have no reason to think that it is damaged, /may reasonably believe not only that the sum of the numbers is x, but also that you are justified in believing that the sum is not x. Here is a case, then, where you have all of my evidence forjfr, and yet I can reasonably believe that you are justified in believing not-/>—for I have reason to believe that your grounds for not-/> are not as telling as you are justified in taking them to be. 8 I am indebted to my colleagues at Purdue University, particularly to Ted Ulrich and Lilly Russow, and to philosophers at The University of Nebraska, Indiana State University, and The University of Wisconsin at Milwaukee for helpful criticisms of earlier versions of this paper.

This page intentionally left blank

CHAPTER FOUR

The Empirical Argument from Evil

It is one thing to argue that the existence of evil is logically inconsistent with the existence of the theistic God and quite another thing to argue that the world contains evils that render the existence of the theistic God unlikely. The former is the logical argument from evil; the latter is the empirical argument from evil.1 Of these two, I believe the empirical argument is the more serious threat to classical theism. Like the former, however, it too has come under vigorous attack. In what follows I will try to set forth a strong version of the empirical argument, examine two important objections that have been advanced against it, determine whether the argument can survive these objections, and assess the epistemological status of theism in the light of the argument. I Those who believe that the world contains evils that render theism improbable need not hold that just any evil renders theism unlikely.2 An earlier version of this paper was presented at a conference in Philosophy of Religion in Lincoln, Nebraska, April 12-14, 1984. I am grateful to Eleanore Stump, my commentator at the conference, for helpful criticisms and comments. 1. This argument has been called 'evidential', 'probabilistic', and 'inductive'. Although I previously referred to it as the 'evidential' argument, it seems less misleading to call it the 'empirical' argument. The argument reaches the conclusion that the existence of God is unlikely given certain empirical facts about the world. 2. A parallel may be drawn here between the logical argument from evil and the empirical argument. Someone who holds that the existence of evil is logically inconsistent

70

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 228

WILLIAM L. ROWE

But intense human and animal suffering on a rather large scale—for example, the suffering occasioned by the Lisbon earthquake—may be thought to do so. For our purpose it will be sufficient to consider intense human and animal suffering occasioned by natural forces. I shall take it as an obvious truth that such suffering is intrinsically evil. It is also obvious that such suffering occurs daily and in considerable abundance in our world. In setting forth the empirical argument, I will use 'O' to abbreviate the phrase 'an omnipotent, omniscient, wholly good being'. Standard (classical) theism is here understood as any view that holds that 0 exists. (I will later distinguish restricted and expanded versions of standard theism.) Two claims are essential to the strong version of the empirical argument from evil. The first is that there are things we know or have good reason to believe that make it likely that the following proposition is true: (A) There exist evils that O could have prevented, and had O prevented them the world as a whole would have been better. Second, the empirical argument holds that the following proposition is true, if not a conceptual truth. (B) 0 would have prevented the occurrence of any evil it could prevent, such that had O prevented it the world as a whole would have been better.3 Since (A) and (B) entail (C) 0 does not exist, with the existence of God may hold that any evil whatever is inconsistent with standard theism. Another position that can be held, however, is that it is only certain kinds (or a certain kind) that logically cannot exist if the theistic God exists. Similarly, the proponent of the empirical argument might hold that standard theism is rendered improbable by the existence of any evil whatever. On the other hand, he may hold that it is the existence of certain kinds, amounts, and degrees of evil that renders theism unlikely. 3. Most theists subscribe to something like (B). But although it has the appearance of truth, (B) has been brought into question by some, and rejected by others. See, for example, Alvin Plantinga, "The Probabilistic Argument from Evil," Philosophical Studies 35 0979); George Schlesinger, Religion and Scientific Method (Dordrecht: D. Reidel, 1977), chs. 9 and 10; and Michael Peterson, Evil and The Christian God (Grand Rapids: Baker, 1982), chs. 4 and 5. These questions and objections do cast some doubt on (B). A full defense of the empirical argument would need to take account of them. For purposes of this paper, however, I will suppose that (B), or something quite like it, is true.

THE PROBLEM OF EVIL The Empirical Argument from Evil

71 229

if (A) is probable and (B) is true, then it is probable that theism is false. The empirical argument, however, is not to be identified with the argument: (A), (B), therefore (C). Rather, the empirical argument is an argument that seeks to provide a good reason for thinking that (A) is more probable than not, that (B) is true, and, therefore, that (C) is probably true. Our ultimate judgment of the worth of the empirical argument depends, therefore, on our estimate of its success in showing that (B) is true and that (A) is probable in relation to certain things we know or have good reason to believe. Before we examine these issues it will be helpful to see why the empirical argument is so modest with respect to (A). Why does the argument seek to establish only that (A) is probable? Why does the argument not endeavor to show that (A) is something we know with certainty, following perhaps from our knowledge of the existence of evil and certain truths concerning omnipotence? It will help us to answer this question if we consider three ways in which evil may be related to 0. The first way is being needful for O. An evil E is needful for O just in case E is preventable by O, some good state of affairs G can be obtained by O only by permitting E, and the world as a whole would be better with G and E than it would without G and£.

The second way is being recalcitrant for O. An evil E is recalcitrant for 0 in case E is preventable by O, some evil state of affairs E' is preventable by 0 only by permitting £, and the world as a whole would be better with E and not E' than it would with E' and not E.

The third way is being permissible for 0. An evil E is permissible for 0 just in case E is preventable by O and either some good state of affairs G can be obtained by 0 only by permitting E (or some evil just as bad), and the world as a whole would be at least as good with G and E as it would be without G and £, or some evil state of affairs E' is preventable by O only by permitting E (or some evil just as bad), and the world as a whole would be at least as good with E and not E' as it would be with E' and not E.4 4. An evil may be permissible for O without being either needful or recalcitrant for O. If £ and E' are equally bad and O must permit one or the other for the greater good of the world, then each is permissible for O but neither is needful for O. Nor, of course, would either E or E' be recalcitrant for O. However, any evil that is needful or recalcitrant for O is permissible for O.

72

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 230

WILLIAM L. ROWE

The idea of an evil being needful for O is intended to capture what some philosophers have had in mind when they have spoken of an evil being related to a greater or outweighing good in such a manner that O can obtain the greater or outweighing good only by permitting the evil. The idea of an evil being recalcitrant for O is intended to capture what philosophers have had in mind when they have spoken of an evil being related to some greater evil in such a manner that O can prevent the latter only by permitting the former. But it is worth noting that these notions are not the same. Consider, for example, the idea of an evil being necessary for some outweighing good. According to Plantinga, "a good state of affairs G outweighs an evil state of affairs E if and only if the conjunctive state of affairs G and E is a good state of affairs."5 This account of 'outweighing', as plausible as it is, has two consequences we should note. The first, a somewhat peculiar consequence, is that a good whose value is, say, +2 may outweigh an evil E whose disvalue is —5. For suppose we have a good G whose value is +7. The conjunctive state of affairs Gl (G and E) has a value of +2. But Gl9 as well as G, will outweigh E. For the conjunctive state of affairs G2 (Gl and E) is equivalent to Gj, and thus is a good state of affairs. "But if a good state of affairs G includes an evil state of affairs E, then the conjunctive state of affairs G and E is equivalent to G (just as a proposition A is equivalent to A & B if A entails B)."6 The second consequence is that if there is some good state of affairs G that outweighs an evil state of affairs E, then— no matter how unrelated the two may be—there will be a good state of affairs Gl (that is, G and E) that outweighs E and is such that O can obtain it (G,) only by permitting E. We can avoid these two consequences by slightly altering Plantinga's account. (i) An evil is outweighed if and only if there is a good state of affairs G such that the conjunctive state of affairs G and E is a good state of affairs. Given that G has a value of +7, E a value of —5, and the conjunctive state of affairs Gj (G and E) a value of + 2, it follows that E is outweighed. But we need not say that Gl9 whose value is +2, itself outweighs E, whose value is —5. Nor need we say that there is some good state of affairs that outweighs E such that O can obtain it only by permitting E. 5. "The Probabilistic Argument from Evil," p. 7. 6. Ibid.

THE PROBLEM OF EVIL The Empirical Argument from Evil

73 231

There is a good state of affairs that 0 cannot obtain without permitting E, namely Gl9 but we need not say that Gl itself outweighs E. The important point, however, is that on neither account of outweighing do we have anything quite like an evil being needful for 0. On Plantinga's account, a good G may outweigh an evil E and be such that 0 cannot obtain it without permitting E and yet not be such that in relation to it E is needful for 0. Suppose G outweighs E but is obtainable without permitting E. Gl (G and E) will then outweigh E and will be obtainable only by permitting E. Is E then needful for 0? No. For E is needful in relation to Gl only if the world would be better as a whole with G! than it would without E. But since 0 can obtain G without permitting E, the world as a whole may well be better without Gj. For 0 can obtain the good part of Gl (i.e., G) and prevent the bad part of Gl (i.e., E). Suppose that G does not include E but cannot be obtained by 0 without permitting E. Suppose, furthermore, that the conjuctive state of affairs Gl (G and E) is a good state of affairs. On Plantinga's account of outweighing, G will outweigh E, and on the slightly altered account £ will be outweighed. But on neither account must it be true that E is needful for 0. For there may be some good state of affairs G' that is better than G but obtainable only if E is prevented. E then will be outweighed by a good G that O cannot obtain without permitting E, but E will not be needful for O, since the world, we may suppose, will be better if 0 prevents E, thus forsaking G, but obtains G', which is better than G. Returning now to proposition (A), we can perhaps begin to see why the empirical argument does not endeavor to prove (A), to establish (A) as something known with certainty. If (A) is true, then there exist evils that are not permissible for O. And I think if we consider some particular evil and ask ourselves how we might know of it that it is not permissible for 0, we shall be able to see how very difficult it would be for us to be in the position of knowing such a thing. (Of course, we might know that some of the evil that exists is not permissible for O, without being able to know of any particular evil that it is not permissible. But prospects for this general knowledge seem no more promising than are the prospects for particular knowledge.) Consider, for example, the enormous amount of suffering occasioned by the Lisbon earthquake. If that evil is not permissible for 0, then no possible good G is such that it is obtainable by 0 only if 0 permits that evil, and the world as a whole would be at least as good with G as without that evil. Can we know with certainty that this is so? To know that no possible good stands in this relation to

74

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 232

WILLIAM L. ROWE

the suffering occasioned by the Lisbon earthquake it seems that we would have to know all the possible goods there are and be sure of each that in relation to it the evil in question is not permissible by 0. We just do not have such knowledge. For all we know, there are qualities we have never thought of, qualities that, if present in a state of affairs, would render that state of affairs a great intrinsic good. Since these great intrinsic goods would be unknown to us, we would not be able to know that 0 could obtain them without permitting certain evils. At best we know some of the things that are intrinsically good. We know, for example, that happiness, knowledge, love, and the admiring contemplation of beauty are intrinsically good. With respect to such goods we might reasonably judge that they would be obtainable by O without O's having to permit the suffering of the Lisbon earthquake or something just as bad. But we do not know that there will not occur states of affairs hitherto unimaginable by us, states of affairs that are great intrinsic goods and in relation to which certain evils are needful to O. To have such knowledge we should require omniscience. It is tempting to object that, even if there are possible goods of which we presently have no knowledge whatever, nevertheless such goods would be obtainable by 0 without O's having to permit the evils that occur in our world. It is tempting to so object because it is evident that evils that are needful for us in relation to certain goods are not needful for 0 in relation to those goods. Certainly, most of the good states of affairs that are obtainable by us only if we permit some evil would be obtainable by O without O's having to permit that evil. This is true. Evils that are needful for us would not be needful in the same way for O. Leibniz's general may have to permit the suffering and death of some of his troops in order to achieve the good of the safety of the women and children from the attacking forces. But it would be child's play for 0 to achieve the latter without permitting the former. It is also true, however, that an evil may be needful for O and yet not needful for us at all. For if an evil is needful for X, then some intrinsic good is obtainable by X, but only if X permits that evil. Clearly there are ever so many possible goods obtainable by O that are not obtainable by us at all, no matter what we permit. If such goods are obtainable by O only if 0 permits some evil E, E may be needful for O but not needful for us. Furthermore, we must remember that by virtue of his omnipotence 0 may be able to bring good out of evil in such a way that the balance of good over evil in the universe is increased. Not only do we not know all possible goods, we also do not know what lies within the power of omnipotence to bring about. Specifically, we do not know that omnipo-

THE PROBLEM OF EVIL

The Empirical Argument from Evil

75

233

tence cannot achieve great intrinsic goods by virtue of the evils he permits. As Augustine reminds us, "Nor would a Good Being permit evil to be done except that in His Omnipotence He can turn evil into good."7 Augustine's idea that evil may be turned into good can be given two distinct, coherent interpretations. The first way of turning evil into good is by producing a good whole consisting of two parts, the evil part and a good part that is obtainable only by permitting the evil part. Since the whole, composed of the evil part and the good part, is itself good, the good part outweighs the evil part, thus making the whole itself good. Here the evil part is justified, if it is justified, by the good part that cannot otherwise be obtained. When evil is thus outweighed, the other part of the whole will be good and better than the whole that is composed of it and the evil part. A second and perhaps more interesting way in which evil may be "turned into" good is by being defeated rather than outweighed by some good.8 When evil is defeated it is a part of a whole that is good but whose goodness is not exceeded by the other part of the whole. Although plausible examples of the defeat of evil are much more difficult to construct than are examples of evil being outweighed, the following example, drawn from Chisholm, may illustrate the nature of the relation involved in the defeat of evil. The contemplation of some wrongful act on my part may be itself, as opposed to the wrongful act, neutral in value. If I then am filled with remorse and displeasure in contemplating my wrongful act, the remorse and displeasure considered in itself is bad, an evil, but the whole of which it is a part may be something that is good, something that is better than the other part of the whole (the contemplation of my wrong act), a part that has less value than the whole (my being filled with remorse and displeasure in contemplating my wrongful act). Here we may say that the whole is made better by the presence of its bad part, the remorse and displeasure. The crucial difference between outweighing evil and defeating it is that in the former case the other part is better than the whole that consists of it and the evil part; whereas, in the latter case, the defeat of evil, the whole is better than the other part. When evil is thus defeated we may be grateful for its presence, for if the other part of the whole existed alone, the world might well contain less value. 7. Enchiridion, ch. 100. 8. The notion of evil being defeated is analyzed by Roderick Chisholm in "The Defeat of Good and Evil," Proceedings and Addresses of the American Philosophical Association 42 (1968-69): 21-38. The account of defeat I present is due to Chisholm. Also see A. C. Ewing's discussion of the same idea, as applied to the problem of evil, in Value and Reality (London: Allen & Unwin, 1973), pp. 215-25.

76

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 234

WILLIAM L. ROWE

The relevance of these two ways of turning evil into good for our problem is as follows. If O is justified in permitting evil by virtue of the evil being outweighed by some other part of a good whole, we must believe that 0 could not obtain the good part without permitting the evil. But if O is justified in permitting evil by virtue of the evil being defeated, we need not hold that O could not have obtained the other part of the good whole without permitting the evil part. When evil is outweighed, it plays a negative role in the production of the good whole and is tolerated only because the other part of the whole could not otherwise be obtained. When evil is defeated, it plays a positive role in the production of the good whole and is accepted even though the other part might be obtained without it. We noted earlier that even though an evil E is outweighed, it will not follow that E is permissible for O. For, in the first place, the other part of the good whole that includes E will be better than the whole and may be obtainable by 0 without permitting E (or something just as bad). And in the second place, if the other part of the good whole is not obtainable by 0 without permitting E, it may still be true that some good greater than it could have been obtained by 0 if, but only if, 0 had prevented E. In similar fashion, if an evil E is defeated, it may still be that, had O prevented E, he could have achieved a greater good than the good whole that includes the evil E that is defeated. With these thoughts in mind, we can now return to the original premise A, the proposition that there exist evils which are such that 0 could have prevented them, and had 0 prevented them the world as a whole would have been better. Conceding, for the reasons given, that we cannot demonstrate (A) or know with certainty that it is true, we must now consider the weaker claim that (A) is a proposition that we have good reasons to think true. This claim is the crucial point in the version of the empirical argument from evil here being considered. As we have seen, if an evil E is such that had 0 prevented it the world as a whole would not have been better, then either (a) there is some good G that O could obtain only by permitting E (or some evil just as bad) and the world as a whole would not have been better had it lacked G and E or (b) there is some evil state of affairs E' such that O could prevent/?' only by permitting £ (or some evil just as bad) and the world as a whole would not have been any better had it contained Ef and lacked E. Let us concentrate our attention on the first of these: the question of whether the evils that exist are such that had an omnipotent being prevented any of them, without permitting others equally bad or worse, some good would have been lost with the result that the world as

THE PROBLEM OF EVIL The Empirical Argument from Evil

77 235

a whole would have been no better. In an earlier paper,9 I argued that we have rational grounds for thinking that this is not so, for thinking that some instances of suffering, for example, could have been prevented by omnipotence and that their prevention would have resulted in a better world. What grounds? I there claimed that (A) is a rational belief given three things: first, our knowledge of the vast amount of intense human and animal suffering that occurs daily in our world; second, our understanding of the goods that do exist and that we can imagine coming into existence; and third, our reasonable judgments as to what an omnipotent being can do. To these three I now would add a fourth: our reasonable judgments of what an omniscient, wholly good being would endeavor to accomplish with respect to human and animal good and evil in the universe. To illustrate this claim, I set forth an example of intense suffering (the intense suffering of a fawn badly burned in a forest fire occasioned by lightning) and observed that as far as we can determine it serves no greater good at all, let alone one that is otherwise unobtainable by an omnipotent being. Recognizing, however, that, appearances to the contrary, it might nevertheless serve such an outweighing or defeating good, I then claimed that it seems quite incredible that all the instances of suffering that serve no greater good we know or can think of should nevertheless be such that none could have been prevented by an omnipotent being without loss of a greater good.

II These claims about what rationally justifies us in accepting (A) have been challenged by several philosophers. My reasoning, it has been suggested, amounts to nothing more than an argument from ignorance: we do not know of any good that would justify O in permitting the fawn's suffering; therefore, there is none.10 It has been suggested that my reasoning is correct only if we assume that, if there were such goods justifying O in permitting the daily sufferings in our world, we would know what they are—an assumption that at worst is false and at best begs the question against traditional theism.11 It has been sug9. "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16, no. 4 (October 1979): 335-41 (hereafter PEVA). 10. Bruce A. Reichenbach, Evil and A Good God (New York: Fordham University Press, 1982), pp. 25-42. 11. The general point that the assumption in question begs the question is developed by M. B. Ahern in The Problem of Evil (New York: Schocken, 1971). This point is ex-

78

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 236

WILLIAM L. ROWE

gested that what I cite as evidence for (A) cannot possibly be evidence, since if 0 existed the things I point to are just what one would expect to be the case.12 I cannot here deal with all the criticisms that have been advanced. But all of these objections are philosophically important and merit serious discussion. I propose to discuss here two of the important criticisms that have been advanced. The first objection I shall discuss has been set forth at some length and with great care by Stephen Wykstra. We can best appreciate his objection if we begin with a major point in my reasoning; namely, that we are unable to think of any good that exists or might come into existence that both outweighs the fawn's suffering and could not be obtained by an omnipotent, omniscient being without permitting that suffering.13 If we agree with this point, we might then claim that (2) It appears that the fawn's suffering is pointless—that is, it appears that the fawn's suffering does not serve an outweighing good otherwise unobtainable by an omnipotent, omniscient being. We can now see Wykstra's basic objection. Put in its simplest terms it comes to this: if someone claims that it appears that S is not P, that person is entitled to that claim only if he has no reason to think that, if 5 were P, things would strike us pretty much the same. Thus, on detecting no sour odor, the person with a cold is not entitled to the claim that it appears that the milk is not sour, because he presumably has a reason (the cold) to think that if the milk were sour things would appear pretty much the same—he would still detect no odor of sourness. Now let us return to (2). Wykstra believes that we are not entitled to affirm (2) because, in his judgment, we have good reason to think that were the fawn's suffering actually to serve an outweighing good, otherwise unobtainable by 0, things would strike us pretty much the same way—we

panded and defended by F. J. Fitzpatrick, "The Onus of Proof in Arguments about the Problem of Evil," Religious Studies 17 (March 1981): 19-38. Also see Delmas Lewis, "The Problem with the Problem of Evil," Sophia 22, no. i (April 1983): 26-35. 12. Stephen Wykstra, "Difficulties in Rowe's Case for Atheism" (paper presented at the Pacific Division Meeting of the APA, March 1984). Also see Wykstra's "The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of 'Appearance'" and my reply, "Evil and the Theistic Hypothesis: A Response to Wykstra," both in International Journal for Philosophy of Religion 16, no. 2 (1984). 13. In exposition of Wykstra's views I will sometimes omit the possibility of good defeating evil. This is done only for convenience and simplicity. A full account would recognize that evil may be either outweighed or defeated.

THE PROBLEM OF EVIL

The Empirical Argument from Evil

79

237

would still be in the position of not being able to think of any good that exists or might come into existence that both outweighs the fawn's suffering and could not be otherwise obtained by an omnipotent, omniscient being. The reason he gives is that were O to exist, it would be very likely that the outweighing good in relation to which 0 must permit the fawn's suffering would be a good quite beyond our ken. And if such a good were beyond our ken, then we would still be in the position of not being able to see what good is served by the fawn's suffering. Thus Wykstra concludes that were 0 to exist, the fawn's suffering would very likely have just the feature I claim it to have—of serving no outweighing good we know to exist or can think of that is otherwise unobtainable by an omnipotent, omniscient being. It is helpful, I think, to view Wykstra's objection as having two steps. The first step is the claim that in the situation described we are entitled to affirm proposition 2 only if the following proposition is true: (3) We have no reason to think that were O to exist things would strike us in pretty much the same way concerning the fawn's suffering.14 The second step in Wykstra's objection is the claim that (3) is false. For were 0 to exist, Wykstra thinks it likely that the outweighing good in relation to which 0 must permit the fawn's suffering would be a good beyond our ken. My response to Wykstra will focus entirely on the second step in his objection: the step that claims that were O to exist things would strike us in pretty much the way they do as far as instances of human and animal suffering are concerned. Given the history of humans and animals and the sorry tale of their sufferings through the ages and given our inability to discover, among the goods we know or can imagine, goods that both outweigh these sufferings and are unobtainable by omnipotence unless these suffer14. It is important to note that Wykstra might also be holding that in the situation described we are entitled to affirm proposition 2 only if (3') is true. (3') We have reason to think that were O to exist things would strike us differently concerning the fawn's suffering. He might then argue that (3') is false and that, therefore, it is not reasonable to believe that were the fawn's suffering actually to serve an outweighing good, otherwise unobtainable by an omnipotent being, things would likely strike us differently concerning the fawn's suffering. In the discussion that follows I have taken Wykstra to be arguing that (3) is false. It would be another matter to determine whether, in the situation described, (3') is also necessary for being entitled to claim (2), and another matter yet to determine whether (3') is true or false.

80

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 238

WILLIAM L. ROWE

ings are permitted, why should we think, as Wykstra does, that this is just how things would likely be, if O exists? It is true, as Wykstra observes, that O's mind can grasp goods that are beyond our ken. The idea, then, is that since 0 grasps goods beyond our ken, we have reason to think it likely that the goods in relation to which O permits many sufferings would be unknown to us. Let us look at Wykstra's reasoning here. He starts with (4) 0 can grasp goods beyond our ken. moves to (5) It is likely that the goods in relation to which O permits many sufferings are beyond our ken. and concludes with (6) It is likely that many of the sufferings in our world do not appear to have a point; we cannot see what goods justify O in permitting them. The difficulty with this reasoning is that the move from (4) to (5) presupposes that the goods in question have not occurred, or, at the very least, that if they have occurred they, nevertheless, remain quite unknown to us (in themselves or in their connections with the sufferings in our world). And, so far as I can see, the mere assumption that O exists gives us no reason to think that either of these is true. If O exists, it is indeed likely, if not certain, that O's mind grasps many good states of affairs that do not obtain and {hat prior to their obtaining we are simply unable to think of or imagine. That much is reasonably clear. But the mere assumption that 0 exists gives us no reason whatever to suppose either that the greater goods in virtue of which he permits most sufferings are goods that come into existence far in the future of the sufferings we are aware of or that, once they do obtain, we continue to be ignorant of them and their relation to the sufferings. The fact that O's mind can apprehend nonactual good states of affairs that we are not able to think of gives us reason to assent to the following proposition: (7) If O exists, then the outweighing goods in relation to which some sufferings are permitted by O are, antecedent to their obtaining, beyond our ken. But this proposition is insufficient to justify the claim that, if 0 were to exist, the sufferings in our world would appear to us as they in fact do.

THE PROBLEM OF EVIL The Empirical Argument from Evil

81 239

I conclude, therefore, that Wykstra has not provided adequate justification for the second step in his objection. And without that step, the objection fails. Earlier, I characterized standard theism as any view that asserts that 0 exists. Within standard theism, we can distinguish restricted theism and expanded theism. Expanded theism is the view that 0 exists, conjoined with certain other significant religious claims, claims about sin, redemption, a future life, a last judgment, and the like. (Orthodox Christian theism is a version of expanded theism.) Restricted theism is the view that 0 exists, unaccompanied by other, independent religious claims. Standard theism implies that the sufferings that occur are permitted by O by virtue of outweighing or defeating goods otherwise unobtainable by 0. Restricted standard theism gives us no reason to think that these goods, once they occur, remain beyond our ken. Nor does restricted standard theism give us any reason to think that the occurrence of the goods in question lies in the distant future of the occurrence of the sufferings that O must permit to obtain them. So I conclude, contrary to Wykstra, that the mere hypothesis of O's existence gives us no good reason to think that things would appear to us just as they do so far as the sufferings of animals and humans in our world are concerned. I conclude, therefore, that we have been given no adequate reason to reject my view that the items mentioned at the outset give us reason to believe that (A) is true and, therefore, that 0 does not exist. But what about expanded theism? Suppose, for example, that we add to the hypothesis that 0 exists the claim made by Saint Paul in his letter to the church at Rome: "For I reckon that the sufferings we now endure bear no comparison with the splendor, as yet unrevealed, which is in store for us." Playing fast and loose with biblical hermeneutics, suppose we set forth the Pauline doctrine as follows: (8) The goods for the sake of which O must permit vast amounts of human and animal suffering will be realized only at the end of the world.15 15. In criticism of Ahern, Fitzpatrick argues in "The Onus of Proof that the theist needs to give a reason why the evils should appear to be unrelated to greater goods otherwise unobtainable by omnipotence. He argues that such a reason exists in the traditional doctrine that the justifying good is God himself and that God's nature is inscrutable in this life. Since this good is unknowable it should not be surprising that we are unable to understand what necessary connections hold between it and present evils. Although there are several points to be made in response to his view, it is worth noting that this, too, is a version of expanded theism.

82

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 240

WILLIAM L. ROWE

We now have a version of expanded standard theism, a version we may call EST, consisting of the conjunction of the proposition that 0 exists and proposition 8. EST is not rendered unlikely by the items that render restricted standard theism (RST) unlikely, for precisely the reason Wykstra so clearly and carefully sets forth. Given EST we have some reason to think that the fawn's suffering might well appear to us just the way it does appear. So the fact that it appears to us as it does in no way renders EST more unlikely than it otherwise is. We have, then, two versions of standard theism, RST and EST. EST entails RST but is not entailed by it. RST is rendered unlikely by the facts about suffering we have alluded to; EST is not rendered unlikely by those facts. Where does this leave us? Does it mean that the theist who is worried about the empirical argument from evil need worry no more, just do a little expanding and the evidence will no longer render his cherished belief unlikely? Strictly speaking, he can do exactly that and be satisfied that the facts of suffering will not render his expanded theism more unlikely than it otherwise is. But in the world of the intellect, as well as in the world of commerce, it seems that everything has its price. If e disconfirms h (in the sense of making h more unlikely than it otherwise is), we can conjoin h with a proposition that clearly entails or makes e likely (call this conjunction h') and no longer worry about e, for e will not disconfirm h'. But the price is that h' will then be of itself as unlikely as h was, given e. So if the facts about suffering do render RST unlikely, there is not much to be gained by retreating to EST.16 16. Robert M. Adams has suggested an interesting response to this conclusion. Basic to his response, I believe, is the contention that the probability of RST, given the facts about evil I have alluded to, is not something that we can, as it were, read from the propositions in question. It is a matter of judgment on our part. Beginning with this point, Adams suggests the following response. I have claimed that RST is significantly discontinued by E (the facts about evil), that the probability of RST, given E, is a good deal lower than the probability of RST alone. (I have also claimed that the probability of RST, given E, is low. But this second claim is not as relevant to the point Adams is making.) EST, since it implies E (or makes E likely), is not disconfirmed by E. But since EST entails RST, EST is just as improbable, given E, as is RST. The reason this can be so, even though EST's probability is not lowered by E, is that the probability of EST, given E, is a function not only of any tendency of E to disconfirm it, but also of the prior probability of EST, the probability of EST alone. Thus, even though E does not disconfirm EST, since EST commits us to much more than does RST, the probability of EST alone may be much lower than the probability of RST alone. In fact, given that EST accounts for E and entails RST, its prior probability must be much lower than RST's, if the probability of RST on E is a good deal lower than the probability of RST alone. The point just made can be turned around and expressed as follows. Given that EST implies E and is thus not disconfirmed by E, and noting that EST entails RST; if the probability of EST alone is not much less than the probability of RST alone, then it

THE PROBLEM OF EVIL The Empirical Argument from Evil

83 241

III

Having seen the flaw in Wykstra's important objection to the empirical argument from evil, let us look at another criticism. In an interesting article "The Problem with the Problem of Evil," Delmas Lewis proposes two objections to the argument I put forth in PEVA. His first objection concerns what he described as "a logical gap between the acknowledged fact that apparently pointless evils exist and the assertion that genuinely pointless evils exist" (p. 29). Lewis agrees that there are many instances of evil that are apparently pointless. That is, he agrees that we are unable to think of goods that either do exist or might come into existence that both outweigh the instances of suffering (evil) and are such that omnipotence could not have obtained them without permitting the instances of suffering. But the question he asks is how we can get from this fact (apparently pointless suffering) to instances of suffering (evil) which are such that no goods have existed, do exist, or will exist which are such that they both outweigh the suffering and are such that omnipotence could not have obtained them without permitting the cannot be true, as I have claimed, that the probability of RST, given E, is much less than the probability of RST alone, that RST is significantly discontinued by E. We now have two possible judgments before us: (i) the judgment as to how much less probable RST is, given E, than it is apart from E; (2) the judgment as to how much less probable EST is than RST. If we are confident about one of these judgments, we can conclude what our other judgment should be. Adams sees me as starting with a particular judgment of (i) and arriving at the appropriate judgment concerning (2). And his perception is exactly right. But he notes, given that these are matters of judgment, that one might start with a particular judgment of (2) and then arrive at the appropriate judgment concerning (i). And here Adams sees the theist as perhaps supplementing RST with an hypothesis such that the result accounts for E and is not significantly less probable than RST. And if the theist can be confident that the result—EST or some other supplementation of RST that would account for E—is not significantly less probable than RST alone, then the theist can rationally conclude that E does not significantly disconfirm RST. Adams's response, so far as its general strategy is concerned, strikes me as exactly right. A theist who is initially troubled by the empirical argument from evil has four different ways of responding. First, the theist may claim that his own judgment of the probability of RST, given E, is far different from mine, that the facts about evil I have alluded to are perhaps just as likely, given the existence of 0, as they are given the nonexistence of 0. Second, the theist can agree with me that RST is indeed disconfirmed by E but note that other facts F make RST so likely that given E and F it is reasonable to accept RST. Third, the theist may agree that RST is disconfirmed by E but hold some other proposition which, although not itself evidence for RST, is such that the conjunction of it with E does not disconfirm RST. Finally, following Adams's lead, the theist may argue that E does not significantly disconfirm RST by showing that there are not implausible hypotheses that, when added to RST, produce a result that both accounts for E and is not significantly less probable than is RST itself. To pursue this last way would be to endeavor to give some not implausible suggestions concerning O's reasons for permitting E. Whether the theist can succeed in this task remains to be seen.

84

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 242

WILLIAM L. ROWE

instances of suffering in question (genuinely pointless suffering). "If Rowe's project is to succeed, he must show how the former fact provides a good reason to believe that the latter assertion is true" (p. 29, italics added). Here I think Lewis may be demanding too much. If my project is to succeed, then the former fact must be a good reason for the latter assertion. And if the former fact is a good reason for the latter assertion, then my project succeeds. Showing that the former fact is a good reason for the latter assertion is another project, an important and interesting project, but it is not something that must be undertaken, let alone accomplished, in order for the initial project to succeed—the initial project being simply that of giving what is a good reason for thinking that there exist instances of suffering (evil) that are genuinely pointless. It is important to see the difference between Wykstra's objection and the objection advanced by Lewis. Although he would not put it in this way, we can say that Wykstra endeavors to give a reason why apparently pointless suffering is not a good reason for thinking that genuinely pointless suffering exists. In response, I have tried to show that his reason is inadequate. Lewis, however, says in effect that since my reason for thinking that genuinely pointless suffering exists does not entail that genuinely pointless suffering exists (the logical gap), it cannot be a good reason unless I prove that it is a good reason. But this is surely incorrect. I may have succeeded in giving a good reason for P without either proving or being able to prove that my reason for P is a good reason, and this, even though my reason does not entail P. To be fair to Lewis, perhaps we should distinguish two projects: giving a good reason for (A) and showing that the reason given for (A) is a good reason. Although I may accomplish the first project without accomplishing the second, I may utterly fail to convince others that I have accomplished the first unless I do accomplish the second. We may then understand Lewis to be drawing our attention to the importance of this second project. Continuing with his objection, Lewis notes that instances of apparently pointless suffering would entitle us to infer that they are genuinely pointless if we knew the following principle to be true. (E) If there are goods for the sake of which O must permit instances of intense human and animal suffering then we would know or be able to imagine these goods and understand why O must permit the sufferings in order to obtain them.17 17. This is not Lewis's formulation of principle E. It does, however, express the same general idea, and it fits more closely the structure of my argument. Lewis's principle E is:

THE PROBLEM OF EVIL The Empirical Argument from Evil

85 243

Now, of course, if we could establish the truth of this principle, we could establish the truth of (A) and, given (B), establish the nonexistence of 0. But I think it should be clear that our prospects for establishing this principle are not very bright. We could imagine a version of expanded theism that entails this principle. Such a gnostic version of standard theism might hold that O would wish to enlighten us concerning evil, not leaving it as a dark mystery for us to puzzle over. On this view, 0 would have revealed to us just what the goods are in virtue of which he permits intense human and animal sufferings, and he would have enlightened us as to why he can obtain them only by permitting that suffering. Of course, this gnostic version of standard theism would be in serious trouble, and we would have good grounds for rejecting it. Lewis agrees that there is much suffering that is apparently pointless. But because of the logical gap between apparently pointless suffering and genuinely pointless suffering, he thinks we are rationally justified in believing the latter on the basis of the former only if we accept this principle and have good reasons for accepting it (or have recourse to some other way of bridging the gap). He thinks we have no good reasons to accept it. He also thinks that to accept the principle is to beg the question against traditional theism. He remarks: "The insistence on the limitations of human knowledge in regard to God's providence is a characteristic tenet of orthodox Christianity (and, it may be added, of orthodox Judaism and Islam as well). It follows that any attempt to argue from evil to the non-existence of God by assuming (E), or some equivalent proposition, amounts to begging the question in favor of the critic's view" (p. 30). About the charge of begging the question, I have two comments. First, appeal to this principle does not beg the question against restricted standard theism, since it is no part of restricted standard theism to hold views about God's providence that render evil a dark mystery. And if we have good reason to reject restricted standard theism, those reasons will be grounds for rejecting any expanded version of standard theism, including orthodox Christianity, Judaism, and Islam. Second, any person who has grounds for accepting the principle will have grounds for rejecting orthodox Christianity, Judaism, and Islam, and this will be so even if, according to these expanded versions of theism, the principle in question is false. It is a small defect of our principle that these expanded versions of theism imply its falsity. Such question begging shows only that the critic would be ill-advised to try to "If there is a morally sufficient reason which explains why an omniscient, omnipotent being could not prevent some instance of evil without thereby losing some greater good or permitting some evil equally bad or worse, then we would know it" (p. 30).

86

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 244

WILLIAM L. ROWE

convince an adherent of one of these expanded theisms to give up her view by appealing to this principle; it will not show that the critic does not have good and solid grounds for holding these expanded theisms to be false. What of Lewis's implicit claim that the logical gap between apparently pointless suffering and genuinely pointless suffering can be rationally bridged only if we have good reasons to accept this principle or some other principle that bridges the gap? In general, I do not think this is correct. That things appear to us to be a certain way is itself justification for thinking things are this way. Of course, this justification may be defeated. But apart from such defeat, the fact that things appear to us to be a certain way renders us rationally justified in believing that they are that way. If in order to be justified in moving from appearances to reality we had always first to justify a principle linking the two, we would be hard put to avoid skepticism.18 But what of the principle itself? Are there really no considerations that speak in its favor? I am inclined to think that there are. In the first place, unless we are excessively utilitarian, it is reasonable to believe that the goods for the sake of which 0 permits much intense human suffering are goods that either are or include good experiences of the humans that endure the suffering. I say this because we normally would not regard someone as morally justified in permitting intense, involuntary suffering on the part of another, if that other were not to figure significantly in the good for which that suffering was necessary. We have reason to believe, then, that the goods for the sake of which much human suffering is permitted will include conscious experiences of these humans, conscious experiences that are themselves good. Now the conscious experiences of others are among the sorts of things we do know. And we do know the beings who undergo the suffering. So if such goods do occur we are likely to know them. Perhaps the good for which some intense suffering is permitted cannot be realized until the end of the world, but it certainly seems likely that much of this good could be realized in the lifetime of the sufferer. Indeed, when religion was young, expectations of the good to come tended to be imminent. Theodicies that emphasize a heavenly afterlife at the end of the world are forged in the grim recognition that we have no heaven on earth. But the issue here is what is reasonable to believe concerning the good experiences of those who suffer, experiences ob18. For a related view see Richard Swinburne, The Existence of God (Oxford: Oxford University Press, 1979), pp. 254-71. Swinburne, however, would not agree with my use of the "appears-so, is-so" principle. For some critical comments on Swinburne's view see Wykstra's paper "The Humean Obstacle to Evidential Arguments from Suffering."

THE PROBLEM OF EVIL The Empirical Argument from Evil

87 245

tainable by omnipotence only by permitting that suffering. In the absence of any reason to think that 0 would need to postpone these good experiences, we have reason to expect that many of these goods would occur in the world we know. By some such line of reasoning as the above, the principle in question can be shown to be plausible, a reasonable candidate for belief. So my answer to Lewis concerning appealing to Principle E in order to bridge the gap from apparently pointless suffering to genuinely pointless suffering is twofold. First, we do not need to establish the principle. And, second, the principle is plausible, a good candidate for rational belief. When we consider the vast amount of intense human and animal suffering occasioned by natural forces and find ourselves utterly unable to think of outweighing or defeating goods that omnipotence can obtain only by permitting such suffering, we are, I think, rationally justified in concluding that there are instances of suffering that omnipotence could have prevented with the result that the world would have been better. Surely, if O were to exist, the world would likely contain much less human and animal suffering than it does. Against this judgment, the defenders of restricted standard theism can point to two things. First, given his omniscience, O will know of nonactual goods that are not known to us. Second, given his omnipotence, 0 will be able to realize some of these goods in the future. These two points show us that the incredible amounts of seemingly pointless human and animal suffering that form the history of our world might be related to future outweighing or defeating goods in such a manner that O can obtain these goods only by permitting those amounts of suffering. But when we take into account the likelihood that the goods in question would involve good experiences of those who suffered and consider how such goods would have to be connected to the sufferings in question so as to be obtainable by omnipotence only at the cost of the sufferings, our judgment must be that this possibility is very unlikely. Surely it is reasonable to believe otherwise, to believe that at least some of this suffering could have been prevented by omnipotence with the result that the world would have been better. And to believe this is to believe the crucial premise of the empirical argument: that there exist evils that 0 could have prevented and had 0 prevented them the world would have been better.

IV There are two ways of developing the empirical argument from evil. The first way, followed in this paper, is to start with the fact that there exist vast amounts of human and animal suffering that seem to us to

88

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 246

WILLIAM L. ROWE

serve no good whatever, let alone one that is otherwise unobtainable by omnipotence. The second way is to start from the somewhat less complex fact that the world contains vast amounts of intense human and animal suffering. With either of these as the starting point, one then must claim that it gives us grounds for thinking that an omnipotent, omniscient, wholly good being does not exist. In this final section, I propose to discuss briefly the second way, relate it to the first, and use the results to illuminate the epistemological status of theism in the light of the empirical argument from evil. The basic point of the second way is that the existence of vast amounts of human and animal suffering tends to disconfirm the existence of O. Intense human and animal suffering in a world supposedly under the control of O is like the presence of dirt in a room supposedly under the control of a powerful being who aims at making the room clean and attractive—its occurrence tends to disconfirm the existence of such a being. If the only information we possessed relevant to the existence of 0 were that the world contains humans and animals that suffer enormously, we would have good reason to think it more likely than not that 0 does not exist. The tendency of the existence of vast amounts of intense human and animal suffering to disconfirm the existence of 0 can be defeated. Two kinds of defeat are possible. The first kind occurs when we add to the proposition concerning evil a proposition that tends to confirm the existence of 0 to a greater degree than the evil tends to disconfirm it. Thus the theist who possesses strong rational grounds for believing that 0 exists may accept the tendency of the existence of vast amounts of human and animal suffering to disconfirm the existence of 0, holding that this tendency is defeated when his rational grounds for assent are taken into account. The second kind of defeat consists in appealing to a proposition that, if true, would explain why it is that O would permit the existence of vast amounts of human and animal suffering. Unlike the first sort of defeat, the added proposition need provide no confirmation of the existence of 0. Instead, the added proposition will be such that its conjunction with the proposition that O exists will make the existence of vast amounts of human and animal suffering likely. Consider, for example, the following proposition. (9) Were O to exist, 0 would have to permit much human and animal suffering so as to secure goods that render the world as a whole better than it otherwise would be. Proposition 9 does not confirm the existence of O, but its conjunction with the proposition that there exists a vast amount of human and

THE PROBLEM OF EVIL The Empirical Argument from Evil

89 247

animal suffering does not disconfirm the existence of O. Thus, if the theist has rational grounds for (9), he may accept the tendency of vast amounts of human and animal suffering to disconfirm the existence of 0, holding that this tendency is defeated when (9) is taken into account. The first of the two ways of developing the empirical argument is somewhat stronger than the second. (Of course, like the second, it too can be defeated.) Its starting point is not merely the existence of vast amounts of human and animal suffering. To this it adds that much of this suffering is such that we are quite unable to think of goods that would justify O in permitting it. By thus strengthening the starting point we increase the likelihood that there exist instances of suffering that are genuinely pointless (proposition A). And, as we saw at the outset, the justification of the existence of genuinely pointless suffering is the fundamental issue in the empirical argument from evil. Some theists hold that the facts about suffering in our world have no tendency at all to disconfirm the existence of 0. They hold that the existence of vast amounts of apparently pointless suffering provides no reason whatever to think that there are any instances of genuinely pointless suffering. In addition, they hold that the kind of knowledge we would have to possess to be in a position to rationally believe that some suffering is genuinely pointless is a knowledge that theism itself says we do not and cannot possess in this life. In short, they hold with F. J. Fitzpatrick that "the problem of evil cannot legitimately function as an argument for atheism."19 For the argument will work only if we can determine that suffering is indeed pointless, but such determination would "deny a fundamental point of theistic belief and therefore .. . beg the question against the theist."20 In this paper I have been arguing that such a view is mistaken. I have argued that the existence of apparently pointless suffering does tend to disconfirm the existence of O, for it provides rational support for the view that some suffering is in fact pointless. Although the argument may involve claims that beg the question against certain versions of expanded standard theism, it does not involve claims that beg the question against restricted standard theism. Furthermore, I have pointed out that begging the question against some forms of expanded theism does not render an argument for atheism useless. The argument may be a perfectly good argument and one for whose premises its proponents have good reasons. So I would say that the problem of evil can legitimately function as an argument for atheism.

19. "The Onus of Proof," p. 28. 20. Ibid.

This page intentionally left blank

CHAPTER FIVE

Evil and Theodicy

Many people feel that some of the human and animal suffering going on in our world makes it difficult to believe in the existence of an omnipotent, omniscient, wholly good being (hereafter referred to as 'O'). Why, for example, would such a being permit the awful suffering and near extermination of the Jews in Europe? Clearly, it is very difficult to understand why an omnipotent, omniscient, perfectly good being would permit this evil. But if our awareness of such evils does make it difficult to believe that O exists, what sort of difficulty is it? It might be a psychological difficulty. We might be so disposed that when we view films of the victims being herded into box cars, being forced into gas chambers, etc., we simply find ourselves inclined to doubt the existence of O, or inclined to abandon our belief in O. On the other hand, the difficulty may be epistemological. We may think that disbelief is somehow rationally justified by our awareness of these terrible evils. It is this latter topic that I want to discuss.

I. The Structure of the Problem I will begin with two instances of evil that have been mentioned in the literature. The first is a case of animal suffering due to natural forces, what would be called a natural evil. The second is a case of human suffering and death due to the intentional action of a human agent, what would be called a moral evil. The first case, although I made it up, is surely a familiar sort of tragedy, played not infrequently on the stage of nature. In some distant forest lightning strikes a dead tree, resulting in a forest fire. In the fire a fawn is trapped, horribly burned, and lies in terrible agony for several days before death relieves its suffering.

92

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION I turn now to a second case, an actual case reported in the Detroit Free Press of January 3, 1986. The case involves a little girl in Flint, Michigan who was severely beaten, raped and then strangled early on New Year's Day of 1986. Here is Bruce Russell's account of the case, which I take from his paper, "The Persistent Problem of Evil."1 The girl's mother was living with her boyfriend, another man who was unemployed, her two children, and her 9-month-old infant fathered by the boyfriend. On New Year's Eve all three adults were drinking at a bar near the woman's home. The boyfriend had been taking drugs and drinking heavily. He was asked to leave the bar at 8:00 p.m. After several reappearances he finally stayed away for good at about 9:30 p.m. The woman and the unemployed man remained at the bar until 2:00 a.m. at which time the woman went home and the man to a party at a neighbor's home. Perhaps out of jealousy, the boyfriend attacked the woman when she walked into the house. Her brother was there and broke up the fight by hitting the boyfriend who was passed out and slumped over a table when the brother left. Later the boyfriend attacked the woman again, this time she knocked him unconscious. After checking the children, she went to bed. Later the woman's 5-year-old girl went downstairs to go to the bathroom. The unemployed man returned from the party at 3:45 a.m. and found the 5-year-old dead. She had been raped, severely beaten over most of her body and strangled to death by the boyfriend. Let's refer to these two cases as El (for the fawn's case) and E2 (for the little girl's case). Now about El and E2,1 want to make the following initial judgment. P. No good state of affairs we know of is such that an omnipotent, omniscient being's obtaining it would morally justify that being's permitting El or E2. What am I implying in making this assertion? I am implying that we have good reason to believe that no good state of affairs we know of would justify an omnipotent, omniscient being in permitting either El or E2. I don't mean simply that we can't see how some good we know about (say, my enjoyment on smelling a good cigar) would justify an omnipotent being's permitting El or E2. I mean that we can see how such a good would not justify an omnipotent being's permitting El or E2. For we can see that an omnipotent being wouldn't have to permit El or E2 in order 120

THE PROBLEM OF EVIL

93

to obtain the good of my enjoyment on smelling a good cigar. And we can see that even were that not so, obtaining such a good wouldn't justify any being in permitting El or E2. Is there some other good state of affairs we know of that would justify an omnipotent being in permitting El or E2? I don't believe there is. The good states of affairs I know of, when I reflect on them, meet one or both of the following conditions: either an omnipotent being could obtain them without having to permit El or E2, or obtaining them wouldn't morally justify that being in permitting El or E2. And if this is so, I have reason to conclude that: Q. No good state of affairs is such that an omnipotent, omniscient being's obtaining it would morally justify that being in permitting El or E2. II. Three Responses I propose now to describe three general ways of responding to the line of reasoning that moves from consideration of El and E2 to the conclusion

Q.

A. One response would be to argue that the reasoning is in some way faulty. Either it makes some untenable or unsupported claim or engages in some faulty inference. For example, one might challenge the claim that among the goods we know about none is such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2. Alternatively, one might argue that the reasoning is fallacious, engaging in an inference from 'we don't know of any such good' to the conclusion that there aren't any, or in an inference from 'no goods we know about would justify such a being in permitting El or E2' to the conclusion that no goods we don't know about would justify such a being in permitting El or E2. B. A second response might acknowledge that the considerations I've mentioned do tend to support Q. But it might be argued that we have good reasons to believe that O exists. And surely, if O exists then there is some good state that O brings about that justifies O in permitting El, and some good state O brings about that justifies O in permitting E2.2 So, since if O exists Q is false, it might be contended that our reasons to believe that O exists outweigh the tendency of the considerations mentioned to support Q. In brief, this response contends that although the considerations mentioned do support Q, when we take into account our reasons for thinking that O exists we see that on balance we have more reason to think Q false than to think it true. 121

94

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION C. A third response, like the second, may acknowledge that the considerations mentioned support Q. Again, like the second this response contends that there is something we know or have good reason to believe that, when conjoined with the considerations mentioned, gives us something that on balance does not support Q. Unlike the second response, however, the thing we know or have good reason to believe is neutral with respect to Q. For example, one might have good reason to believe that if there are good states of affairs the obtaining of which would justify an omnipotent, omniscient being in permitting El or E2, all would be goods we know nothing about. Now this information by itself does not make Q unlikely, but when we conjoin this information with our initial judgment P (no good we know of is such that obtaining it would justify an omnipotent, omniscient being's permitting El or E2), the conjunction does not support Q. So here, instead of saying that the tendency of these considerations to support Q is outweighed, we shall say that it is defeated. My chief interest here is to discuss the first and third of these responses. To evaluate the strength of the second response would require an evaluation of our reasons for believing that O exists, a topic far too large to take up here. According to the first response, the attempt we have made to reason from the existence of certain evils, such as El and E2, to the nonexistence of O is flawed in one of two ways, if not both. First, it may be flawed in its implicit claim to have good reason to make the initial judgment P (no good we know of is such that obtaining it would justify an omnipotent, omniscient being in permitting Ei or E2.) Second, it may be flawed in concluding from this initial judgment that Q (no good is such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2.) Let's now consider whether this first sort of response is successful. To simplify our discussion, I will use the letter T to stand for the property a good has just in case obtaining that good would justify an omnipotent, omniscient being in permitting El or E2. If a good is such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2, then that good has J; and if not, the good lacks J. The first question, then, that this response raises is whether we are really able to judge with any assurance that P: all the goods we know of lack J. And the second question this response raises is whether from our initial judgment, P, we can reasonably infer that Q: all goods (including those we don't know of) lack J? Before turning to these questions, however, it's worth noting that both 122

THE PROBLEM OF EVIL

95

questions imply that the class of goods can be divided into those we know of and those we don't know of. What is it to know of an intrinsically good state of affairs?3 Well, as a start, I'm sure we would all be able to list some states that are intrinsically good, some that are intrinsically bad, and some that are neutral, neither intrinsically good nor bad. On our good list we would have: pleasure, happiness, love, the exercise of virtue, good intentions, etc. On our bad list we would have pain, unhappiness, hatred, the exercise of vice, bad intentions, etc. Other states, the existence of a stone, say, we would judge to be neither intrinsically good nor intrinsically bad. Now if we could imagine the list of intrinsic goods as completed, I think the completed list would include some states of affairs we would recognize as intrinsically good, even though they did not come to mind when we first started composing the list. We might, for example, find knowledge on the list, and respond with the recognition that it too is intrinsically good. But the complete list might well include states that are enormously complex, so complex as to tax our powers of comprehension. Determining that such a state is intrinsically good may be as difficult for us as it would be for a child of three to determine whether Bayes' theorem is true. Goods of this sort are what I have in mind when I speak of goods we don't know of.4 Among the goods we know of, is any such that it has J? It is obvious that many of these goods lack J. For when we contemplate them we see that their value is not high enough (e.g., my enjoyment on smelling a good cigar) to offset the evils in question. Other goods we know of may have a great deal of intrinsic value, perhaps even more value than El or E2 have disvalue. But here we readily see that an omnipotent, omniscient being could obtain them without permitting El or E2.5 Reflections of this sort justify us, I believe, in making our initial judgment, P, that the goods we know about lack J. Suppose we accept P. What about the inference to Q? Are we justified in believing that no goods (including those we don't know of) have J on the basis of our justified premise that no goods we know of have J? Perhaps any such inference commits some fallacy. After all, if we don't know of these goods, how can we be justified in concluding either that they do or do not have J?.My answer is that we are justified in making this inference in the same way we are justified in making the many inferences we constantly make from the known to the unknown. All of us are constantly inferring from the A's we know of to the A's we don't know of. If we observe many A's and all of them are B's we are justified in believing 123

96

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION that the A's we haven't observed are also B's. If I encounter a fair number of pit bulls and all of them are vicious, I have reason to believe that all pit bulls are vicious. Of course, there are all sorts of considerations that may defeat this inference. I may discover that all the pit bulls I've encountered have been trained for dog-fighting, a training that engenders viciousness. I may also come to know that there are many pit bulls that are not so trained. If so, then this additional information, along with my initial information, may not justify me in believing that the pit bulls I haven't encountered are also vicious. So, too, with my inference from P to Q. But considerations of this sort (defeaters) properly belong to the third response, not the first. The first response, I believe, is not successful.6 The third response holds greater hope for discrediting the argument from evil here being considered. Before seeing how it proceeds, we should look a bit more carefully at the notion of defeat. Let's do this with an example. Suppose it is true that 1. The wall appears red. From (1) we infer that 2. The wall is not white. The inference is undoubtedly justified. If (1) were all we knew relative to the truth of (2), we would be justified in believing (2). How might the tendency of (1) to support (2) be defeated? Well, suppose we knew the following proposition: 3. If the wall were white, the wall would appear red anyway. Clearly, although (1) tends to support (2), this tendency would be defeated if (3) were added to our knowledge. That is, the conjunction of (1) and (3) does not support (2). [Note that (3) by itself does not support the denial of (2). So this is not a case in which our new knowledge outweighs the support we have for (2).] How might we come to know or rationally believe (3)? Well, suppose we learn that 4. There are red lights shining on the wall and when red lights shine on a white wall the wall will appear red. (4) gives us a good reason for accepting (3), and, as we've seen, when we conjoin (3) with (1) we no longer are justified in believing (2)—the tendency of (1) to support (2) is defeated. Although it is risky to generalize from a single example, let's note the following pattern. From A we infer B. Instead of having evidence against B itself, as in attempting to outweigh A's tendency to support B, we have 124

THE PROBLEM OF EVIL

97

reason to believe that ifnot-B, A would be (or would likely be) true anyway. When we conjoin this hypothetical with A, we have a conjunction that does not support B. With this pattern in mind, let's return to our original inference to Q (no goods have J) from P (no goods we know of have J). To defeat P's tendency to support Q we need some good reason to believe that if not-Q (some good has J), P would be (or likely be) true anyway. Do we have any good reason to believe that if not-Q, P would be true anyway? In considering our question it is important to note that our hypothetical refers to omniscience and omnipotence, as opposed to something akin to human knowledge and power. If a human being permits some great evil to occur, we would have some reason to think that if the obtaining of some good does justify the permission, we would likely know of the good; for another human being's knowledge and power resembles quite closely our own. For this reason we generally would have reason to reject the hypothetical that if the obtaining of some good would justify this person in permitting the evil, it would still be true that it is a good we do not know. But omniscience would view all goods, including the ones we don't know of. So, unlike the hypothetical referring to human knowledge and power, we have no similar reason to think that if some good is such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2, it would likely be one we know of. But all that we have seen, thus far, is that when we replace omniscience by human knowledge and power we have good reason to think that the hypothetical that would correspond to If not-Q, P anyway (i.e., the hypothetical concerning human knowledge and power), \^ false. The question before us, however, is whether we have some reason to think the original hypothetical [if not-Q, P anyway] is true. And I must confess that I cannot think of any special reason to think that this hypothetical is true. Given the high degree of intrinsic badness of El and E2, we do know that the justifying goods would be very significant goods. But this, by itself, is no good reason to think that the goods would be ones we don't know of. For we do know of very significant goods. Consider, for example, the state consisting of a vast number of conscious beings deeply enjoying the admirable qualities exhibited by one another. This is a very significant good we know of. So the mere fact that the goods must be significant is not, of itself, good reason to think that the goods that might justify an omnipotent, omniscient being in permitting El or E2 would be goods we do not know. Perhaps there is some fact such that if we knew it we would 125

98

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION have a good reason to think that the hypothetical is true. We certainly cannot rule out there being such a fact. But until we learn of it we have no good reason to think that if Q were false, P would be true anyway. And lacking such a reason, we are unable to defeat P's tendency to support

Q 7Let's look back, briefly, at the path along which we have come. Our original question was whether certain facts about evil provide rational support for the view that O does not exist. In considering this question, we noted two particular instances of evil, El and E2, and noted that P [no good we know of would justify an omnipotent, omniscient being in permitting El or E2.] From P we inferred Q [no good at all would justify an omnipotent, omniscient being in permitting El or E2]. Suppose we are justified in believing Q on the basis of our belief that P. If so, then, since we see that Q would be false if O existed, we are justified in believing that O does not exist. In response to this line of argument, I suggested that there are three responses worthy of consideration, only two of which (the first and third responses) I would take up in this paper. Against the first response, I argued that we are justified in believing P and in inferring Q from P. Against the third response, I argued that although having a good reason for believing the hypothetical [If not-Q, P anyway] would defeat our acceptance of Q on the basis of P, we do not in fact have a good reason to accept this hypothetical. In short, unless we have good reasons to think that O exists (or some other reason to reject Q), we are justified in believing that O does not exist. III. Theodicy and The Problem of Evil Thus far we've taken no account of the various theodicies that have been mounted to reconcile our belief in theism with the difficulty raised by some of the human and animal suffering that goes on in our world. What I propose to do in this section is to look briefly at one of the more promising theodicies to determine just where it impinges on the difficulty, as we have developed it. In addition, we will need to consider the extent to which the theodicy reduces the difficulty. The theodicy I have in mind is the one developed and defended by John Hick, and referred to as a theodicy of "soul-making."8 Before giving a synopsis of this theodicy, it will be helpful to reflect on the general bearing of theodicies on the difficulty we have developed. Just what does a theodicy endeavor to do? Does it propose to tell us in some detail just what good 126

THE PROBLEM OF EVIL

99

state it is that justifies O, if O exists, in permitting El and E2? No. Such an account would presume a knowledge of O's specific purposes, a knowledge that it would be unreasonable to expect we would have without some detailed revelation to us from O. What a theodicy does endeavor to do is to fasten on some good state of affairs and argue that it would justify O in permitting evils like El and E2. Whether obtaining the good in question is O's actual reason for permitting evils like El and E2 is not really part of what a theodicy tries to establish. It only hopes to show that if obtaining the good in question were O's aim in permitting evils like El and E2, then (given what we know) O would be justified in permitting such evils. In addition, it is important to the success of the theodicy that the good in question not be a good we have reason to believe does not obtain. It would not be helpful, for example, to argue that the good in question is every human being turning to Christ before that person experiences bodily death. For we have very good reason to think that such a good does not obtain. The theodicist doesn't have to establish that the good in question does obtain, but she must at least argue that we do not have any really strong reasons to think that it doesn't obtain. In terms of our presentation of the difficulty, then, we should see the theodicist as denying P. What the theodicist claims, so it seems to me, is that some good we know of is such that obtaining it would justify O in permitting El and E2. Need a theist be a theodicist? I don't think so. A theist must reject Q. But she need not reject P. Someone could hold that her reasons for thinking O exists outweigh P's support of Q, so that it is rational on balance for the theist to believe that Q is false. Such a theist might hold that the good that would justify O in permitting El and E2 is a good that is totally beyond our ken in this life. Such a position is not inconsistent with theism, although it does abandon the project of providing a theodicy.9 Some theodicies depend on having very strong reasons for believing that O exists. Leibniz' theodicy, as I read it, is a case in point. Having established to his satisfaction that O exists, Leibniz claimed to deduce that the good in question is this being the best of all possible worlds. Now if this is the best of all possible worlds then clearly such a good would justify an omnipotent, omniscient being in permitting El, E2 and all the other evils our world contains.10 Other theodicies need not depend for their success on proving the existence of O. Hick's theodicy, as I read it, does not depend on such a proof. With these preliminary remarks out of the way, we can now consider the basic structure of Hick's "soul-making" theodicy. 127

100

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION There are two good states that figure in Hick's theodicy. The first is the state in which all human beings develop themselves through their free choices into moral and spiritual beings. The second good state is that in which all such beings enter into an eternal life of bliss and joy in fellowship with O. The second of these good states (and probably the first) obtains only if O exists. But, as I've suggested, in order to be successful a theodicy need not establish that the good in question obtains. Let's begin our synopsis by considering the first of these states, the state in which all human beings develop themselves through their free choices into moral and spiritual beings. How might the obtaining of such a good justify an omnipotent, omniscient being in permitting evils like El or E2? Since El and E2 are instances of natural and moral evil, different answers may be required. Let's begin with horrendous moral evils like E2. Hick's first step is to argue that if moral and spiritual development through free choices is the good in question then an environment in which there is no significant suffering, no occasion for significant moral choices, would not be one in which moral and spiritual growth would be possible. In particular, a world in which no one can harm another, in which no pain or suffering results from any action, would not be a world in which such moral and spiritual growth could occur. I think we can concede to Hick that a pain-free paradise, a world in which no one could be injured and no one could do harm, would be a world devoid of significant moral and spiritual development. But what are we to make of the fact that the world we live in is so often inimical to such moral and spiritual development? For clearly, as Hick is careful to note, much of the pain and suffering in our world frustrates such development. The overall situation is thus that, so far as we can tell, suffering occurs haphazardly, uselessly, and therefore unjustly. It appears to be only randomly related either to past desert or to future soulmaking. Instead of serving a constructive purpose, pain and misery seem to fall upon men patternlessly and meaninglessly, with the result that suffering is often undeserved and often occurs in amounts exceeding anything that could have been morally planned.11 Hick's response to this point is to ask us what would happen were our world one in which suffering occurred ". . .not haphazardly and therefore unjustly, but on the contrary justly and therefore non-haphazardly."12 In such a world, Hick, following Kant, reasons that people would avoid 128

THE PROBLEM OF EVIL

101

wrongdoing out of fear rather than from a sense of duty. Moreover, once we saw that suffering was always for the good of the sufferer, human misery would no longer ". . . evoke deep personal sympathy or call forth organized relief and sacrificial help and service. For it is presupposed in those compassionate reactions both that the suffering is not deserved and that it is bad for the sufferer."13 Hick then concludes: It seems, then, that in a world that is to be the scene of compassionate love and self-giving for others, suffering must fall upon mankind with something of the haphazardness and inequity that we now experience. It must be apparently unmerited, pointless, and incapable of being morally rationalized. For it is precisely this feature of our common human lot that creates sympathy between man and man and evokes the unselfishness, kindness and goodwill which are among the highest values of personal life.14 Let's assume with Hick that an environment fit for human beings to develop the highest qualities of moral and spiritual life must be one that includes real suffering, hardships, disappointments, failure and defeat. For moral and spiritual growth presuppose these. Let's also assume that such an environment must operate, at least for the most part, according to general and dependable laws, for only on the basis of such general laws can a person engage in the purposeful decision-making essential to rational and moral life.15 And given these two assumptions it is, I think, understandable how an omniscient, omnipotent being may be morally justified in permitting the occurrence of evils, both moral and natural. Our excursion into John Hick's theodicy has shown us, perhaps, how a theodicy may succeed in justifying O's permission of both natural and moral evil. But so far we haven't been given any justification for O's permission of El or E2. In the case of El we can say that given the existence of the animals in our world and the operation of the world according to natural laws, it is unavoidable that instances of intense and prolonged animal suffering would occur. In the case of E2 we can say that on their way toward moral and spiritual development it is perhaps unavoidable that human beings will sometimes seriously harm others through a bad use of freedom. But neither of these points will morally justify an omnipotent, omniscient being in permitting El and E2. In the case of E2, it is simply unreasonable to believe that if the boyfriend acted freely in brutally beating and raping the little girl, his moral and spiritual development would have been permanently frustrated had he been prevented 129

102

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION from doing what he did. And it is also unreasonable to believe that permitting such an act is morally justified even if preventing it would somehow diminish the boyfriend's moral and spiritual odyssey.16 And in the case of El, it is simply unreasonable to believe that preventing the fawn's being severely burned, or mercifully ending its life so that it does not suffer intensely for several days, would so shake our confidence in the orderliness of nature that we would forsake our moral and spiritual development. I think Hick is not unaware of this limitation to his theodicy, at least with respect to natural evils. With respect to human pain due to sources independent of the human will, he remarks: In response to it, theodicy, if it is wisely conducted, follows a negative path. It is not possible to show positively that each item of human pain serves God's purpose of good; on the other hand, it does seem possible to show that the divine purpose, [. . .] could not be forwarded in a world that was designed as a permanent hedonistic paradise.17 What Hick says here is, I believe, right, with one minor modification. The minor modification is this. In Hick's theodicy some evils themselves are necessary to the attainment of the good of moral and spiritual development. But many evils, particularly those resulting from the operation of the laws of nature, will not themselves serve moral and spiritual development. They may even hinder it. Nevertheless, as the by-product of the general operation of the laws of nature, something that is essential to rational decision-making and moral living, an omnipotent, omniscient being would have reason to permit even these evils, at least to the extent that further intervention on its part would diminish our expectation that the world operates on general laws. What we've seen is that Hick's theodicy fails if it is intended to provide a good that would justify an omnipotent, omniscient being in permitting El or E2. The best that Hick can do is to argue that a world utterly devoid of natural or moral evil would preclude the realization of the goods he postulates as justifying an omnipotent, omniscient being in permitting evil. However, since the prevention of El or E2 would not leave our world utterly devoid of natural or moral evil, his all-or-nothing argument provides no answer to our question. Nor will it do to say that if an omnipotent, omniscient being were to be morally obligated to prevent El or E2 it would thereby be obligated to prevent all such evils. For were it to do so it may well be that we would cease to engage in very significant soul-making. 130

THE PROBLEM OF EVIL

103

The problem Hick's theodicy leaves us is that it is altogether reasonable to believe that some of the evils that occur (El and E2, for example) could have been prevented without either diminishing our moral and spiritual development or undermining our confidence that the world operates according to natural laws. Hick's theodicy, therefore, does not succeed in showing that we have no good reason to accept P. Nor does his theodicy show that we aren't justified in inferring Q from P. In short, Hick's theodicy does little to diminish the claim that the existence of certain evils provides rational grounds for atheism. Some may feel that I have criticized Hick's theodicy for failing to do what it was never meant to do: to show us the good the obtaining of which would justify an omnipotent, omniscient being in permitting El or E2. Theodicies, one might argue, are not intended to provide a justification for particular evils. I acknowledge the merit of this criticism. There are, I think, four different things a theodicy might aim at doing, each more difficult than its predecessor. First, a theodicy might seek to explain why O would permit any evil at all. Second, a theodicy might endeavor to explain why there are instances of the various kinds of evil we find in our world—animal pain, human suffering, wickedness, etc. Third, a theodicy might endeavor to explain why there is the amount of evil (of these kinds) that we find in our world. And, finally, a theodicy might endeavor to explain certain particular evils that obtain. I think Hick's theodicy may be successful on the first level, and perhaps the second. In so far, therefore, as we argue against the existence of O solely by appealing to the fact that our world contains evil, or to the fact that our world contains certain kinds of evil, Hick's theodicy may show that the relevant P-like premise can be reasonably rejected. But arguments based on the third and fourth levels, so far as I can see, remain undiminished by Hick's theodicy. If his theodicy is not intended to diminish the strength of these arguments—by giving grounds for rejecting their P-like premises—then it is a mistake to charge that the theodicy fails to accomplish its end. But the point will remain that the theodicy does little or nothing to diminish the force of the strongest arguments from evil. NOTES 1. Bruce Russell's paper is forthcoming in Faith and Philosophy. 2. For sake of simplicity, I shall ignore the possibility that O permits El or E2 in order to prevent some equal or worse state from obtaining. I shall also ignore the possibility that there are two evils equally bad such that either could be prevented without loss of the 131

104

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION justifying good, but one or the other must be permitted to obtain the good. 3. Roughly, to know of a good state of affairs is to (a) conceive of that state of affairs, and (b) recognize that it is intrinsically good. 4. In addition, there may be simple properties we have never thought of, properties whose presence in a state of affairs might render that state a great intrinsic good. 5. Of course, the conjunction of one of these great goods, G, with El, say, will be unattainable by omnipotence without permitting El. But since we have reason to think that G can be obtained by omnipotence without permitting El, obtaining the conjunction of G and El won't justify this being in permitting El. The conjunction of G and El, like G itself, will lack J. 6. The first response in effect claims that if P were all we knew relative to the truth or falsity of Q, we would not be rationally justified in believing Q. It is this claim that I am rejecting as false. 7. Stephen Wykstra has pointed out to me that it may suffice to defeat P's support of Q to show that if not-Q, P would be no less likely than not-P. We should, therefore, distinguish a stronger and more modest defeater. To give a reason for believing // notQ P would be true anyway is to provide a strong defeater. To give a reason for believing if not-Q P would be no less likely than not-P is to provide a modest defeater. In the text I discuss only the strong defeater. A more comprehensive discussion would need to consider the more modest defeater as well. 8. See Evil and the God of Love (New York: Harper and Row, 1966), particularly Ch. XVII of the revised edition, published in 1978, God and the Universe of Faiths (New York: St. Martin's Press, 1973), and Chapter Four of Philosophy of Religion, 3rd edition (Englewood Cliffs, New Jersey: Prentice-Hall, 1983). 9. The account I've given of the aim of a theodicy takes it to be providing a plausible justification for O's permitting particular evils. Perhaps this is too ambitious. Perhaps the aim of a theodicy is only to provide a plausible justification for O's permitting some instances of various types of evil. One might succeed at the latter and yet fail at the former. If we do take the second account, we need to revise my judgment of Hick's theodicy. One might then conclude that although Hick's theodicy is successful, it is largely irrelevant to the question of whether the existence of particular evils provides rational grounds for atheism. (See the final paragraph of this essay.) 10. Is ttus being the best of all possible worlds a good we know of? I'm not sure of the proper answer to this question. If it isn't, then my suggestion that theodicies are committed to denying P needs to be qualified. But apart from good reasons to believe that O exists, I would argue that we have reason to think that this is not the best of all possible worlds. 11. God and the Universe of Faiths, p. 58. 12. Ibid. 13. Ibid., p. 60. 14. Ibid. 15. This point is forcefully argued by William Hasker in "Suffering, Soul-Making and Salvation," International Philosophical Quarterly, Vol. XXVIII, No. 1 (March 1988), pp. 3-19. 16. See Bruce Russell's paper for further argument on this point. 17. Philosophy of Religion, p. 46.

132

CHAPTER SIX

Paradox and Promise: Hick's Solution to the Problem of Evil

John Hick has written extensively concerning the problem of evil.1 His writings on the problem are set within two general restraints. First, he holds constant the conception of God as omnipotent and perfectly good. Hick does not pursue the question of whether the facts about evil necessitate some drastic revision of our conception of God. In my discussion of his work, I shall accept this restraint. Secondly, Hick distinguishes the question of whether the existence of God is logically consistent with the facts about evil from the question of whether the facts about evil render belief in God unreasonable or irrational. It is the second of these questions that is the focus of Hick's attention. Some philosophers hold that the facts about evil do not provide any rational grounds or evidence for disbelief in God. Thus, for them, the only serious problem is the question of logical consistency. I side with Hick on this issue. The proposition that a given man is seventy-five years old and has an arthritic knee is logically consistent with his winning the next Boston marathon. But surely the fact that he is seventy-five and has an arthritic knee gives us good rational grounds for believing that he won't win the next Boston marathon. So too, although the free-will defence may establish that the facts about evil are logically consistent with the existence of God, there remains the serious problem of the extent to which the facts about evil render it unreasonable or irrational to believe in God. Before we begin our examination of Hick's proposed solution to the problem of evil, I want to make explicit an assumption of mine, which I believe Hick shares, and note several different forms the problem of evil may take for us. The assumption is that, although we may have difficulty in defining "intrinsic good" and "intrinsic

106

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 112

Paradox and Promise

evil", this does not preclude us from knowing that certain things are intrinsically good and other things are intrinsically evil. We know, for example, that pleasure, happiness, love, the exercise of virtue, good intentions, and so on, are intrinsically good. We also know that pain, unhappiness, hatred, the exercise of vice, bad intentions, and so forth, are intrinsically evil. The problem of evil, then, is the problem of explaining why an omnipotent, perfectly good being would permit intrinsic evils to exist, or to exist in such abundance.2 But just how are we to conceive of our problem? Are we asking why an omnipotent, perfectly good being would permit any evil at all in our world? If this is our problem, then we can solve it by explaining how such a being would be justified in permitting just one instance of the evils that exists in our world. Are we asking why an omnipotent, perfectly good being would permit the various kinds of evil we find in our world? If this is our problem, then we can solve it by explaining how such a being would be justified in permitting an instance of each of these kinds of evil to exist in our world. Are we asking why an omnipotent, perfectly good being would permit the amount of evil (of these kinds) that we find in our world? If this is our problem, then we can solve it only by explaning how such a being would be justified in permitting all the evil (in its various kinds) that exists in our world. Or, finally, are we focusing on some particular evils and asking why an omnipotent, perfectly good being would permit these particular evils to occur? If this is our problem, then we can solve it only by explaining how such a being would be justified in permitting those particular evils to occur. So, once we think about the problem of evil, we come to see that there are four distinct problems of evil. And, as we shall see when we look at Hick's soul-making theodicy, what will solve one of these problems may be wholly inadequate to solve the rest of them. To any of the four problems of evil that I have just distinguished, a theist may make one or more of four different responses. First, as we have already noted, the theist may insist that the facts about evil that are noted provide no rational grounds at all for rejecting belief in God. The argument from the facts of evil to the non-existence of God may be stated as follows. No good we know of is such that obtaining it would justify God in permitting evil (these kinds of evil, the amount of evil, these particular evils). Therefore, it is reasonable to believe that no goods at all would

THE PROBLEM OF EVIL

William Rowe

107

113

justify God in permitting the evil in question. But, if no good at all would justify an omnipotent, perfectly good being in permitting the evil in question, then God does not exist. To this argument, theists who make the first response typically reject the reasoning as fallacious, involving an illicit inference from "we don't know of any such good" to the conclusion that there aren't any, or from "no goods we know about would justify such a being in permitting the evil in question" to the conclusion that no goods we don't know about would justify such a being in permitting the evil in question.3 A second response would be to acknowledge that the argument does provide rational support for disbelief in God, but to go on and argue that our rational grounds for believing in God are stronger and, therefore, outweigh the grounds for disbelief provided by the facts of evil.4 A third response, like the second, may acknowledge that the argument, as stated, does provide rational grounds for disbelief. But this response goes on to suggest that there are other things we know or have reason to believe which, when conjoined with the premise of the argument ("No good we know of is such that obtaining it would justify God in permitting the evil in question"), gives us something that does not provide rational grounds for disbelief. Unlike the second response, however, the additional thing appealed to is not itself a reason for belief in God. For example, suppose one had good reason to believe that, if there are any goods that would justify an omnipotent, perfectly good being in permitting the evils in question, they would be goods we do not know. Now this information by itself is not a reason for belief in God. But, when we conjoin this information with the premise of the argument, the conjunction does not provide rational support for the conclusion that no good is such that obtaining it would justify an omnipotent, perfectly good being in permitting the evil in question. So here, instead of saying that our reasons for disbelief are outweighed, we can say that they are defeated. Finally, one may directly challenge the premise of the argument by arguing that there are goods we know (or can imagine) the obtaining of which would justify God in permitting the evil in question. It is this final response that Hick gives to the problem of evil; in common with other theodicists, he endeavours to single out some good or goods which can be plausibly believed to justify an omnipotent, perfectly good being in permitting evil. In evaluating Hick's proposals about the justifying goods, it

108

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 114

Paradox and Promise

would be incorrect to require that he must show that the goods in question will obtain. A theodicist does not have to establish that, but it must be argued that we do not have any good reasons to think that the goods will not obtain. It would not be helpful, for example, to propose that the good in question is every human being achieving moral perfection, or even moral competence, before bodily death. For we have very good reason to think that such a good does not obtain. To the extent that the good in question is one that we have reason to think will not obtain, we have reason to think that the good proposed does not justify any omnipotent, perfect being in permitting the evil in question. There are two good states that figure in Hick's theodicy. The first is the state in which all human beings freely develop themselves into moral and spiritual beings.5 The second good state is that in which all such beings freely enter into an eternal life of love and fellowship with God.6 In order for these goods to be an omnipotent, perfectly good being's reasons for permitting moral and natural evil, something like the following two principles must be true: a person's freely developing into a morally better being is intrinsically more valuable that a person's being made to be a morally better being; a person's freely coming to love another being is intrinsically more valuable than a person's being made to love another being. I believe both principles are true. But we should be wary of questionable inferences that may be drawn from such principles. First, it would be a neglect of Moore's principle that the value of the whole need not be the sum of the values of the parts (the principle of organic unities) to conclude that freedom itself is of great intrinsic value.7 Secondly, to admit that freely developed goodness is intrinsically better than readymade goodness does not imply that it would always be permissible to obtain the former rather than the latter. This would be so only if the degree to which the former is intrinsically better than the latter is infinite. By appealing to the two good states just mentioned, Hick proposes to explain why, if there is an omnipotent, perfectly good being, it would be true that (1) human beings are not immediately aware of his existence, (2) human beings exist as morally immature beings, (3) the world is a place where real harm can be inflicted by one person on another, and (4) the world operates with a fixed structure in which real pain and suffering occur. For only in an environment in which (l)-(4) obtain can human persons freely come to know and love God, and freely develop themselves into mature moral and spiritual beings.8

THE PROBLEM OF EVIL

William Rowe

109

115

Is Hick right about all of this? So far as (2), (3) and (4) are concerned, I think reason is on his side. My only real doubts concern (1) and Hick's claim that unless human beings existed at an "epistemic distance" from God they could not freely come to know and love God. And, even on this point, I am not clear as to the extent of my disagreement with Hick. If all Hick means to say is that, in order to come freely to know God (or to believe in God), one must exist at an epistemic distance from God, then I grant him his point. If God is directly present to me in all his power, glory and love, my intellect compels my assent to the proposition that he exists; there is no room for free assent. The problem is that Hick does not stop with this point. He sometimes says that, in order to be a a person, in order to be morally free, in order to be free with respect to loving God, we must exist at an epistemic distance from God. My difficulty here is twofold: I cannot find compelling reasons in Hick's writings for these further claims, and these claims do not seem to me to be true. Does Hick make such claims? Consider the following passages: In such a situation [existing directly in God's presence] the disproportion between Creator and creatures would be so great that the latter would have no freedom in relation to God; they would indeed not exist as independent autonomous persons . . . . In order to be a person, exercising some measure of genuine freedom, the creature must be brought into existence, not in the immediate divine presence, but at a "distance" from God.9 Within such a situation [epistemic distance from God] there is the possibility of the human being coming freely to know and love one's Maker. Indeed, if the end-state which God is seeking to bring about is one in which finite persons have come in their own freedom to know and love him, this requires creating them initially in a state which is not that of their already knowing and loving him.10 . . . if men and women had been initially created in the direct presence of God, who is infinite in life and power, goodness and knowledge, they would have had no genuine freedom in relation to their Maker. In order to be fully personal and therefore morally free beings, they have accordingly (it is suggested) been created at ... an epistemic distance [from God].11

110

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 116

Paradox and Promise

I take it as evident that a person could exist in a state of epistemic immediacy with God. Could a person be genuinely free and exist in a state of epistemic immediacy with God? Clearly such a person could not be free with respect to coming to know or believe in God's existence. But could such a person be free with respect to developing morally in relation to other human beings? I don't see why not. Of course, this person would know that God exists, loves all his creatures, and so on. But knowledge is one thing; developing moral character and acting out of a sense of duty toward your neighbours is something else. It is very hard to see how the knowledge of God would destroy one's freedom with respect to moral development in relation to one's neighbours. But what about loving God? Would one still be free with respect to loving God? In the second of the three passages quoted above, Hick, after noting, correctly, that epistemic distance would leave the person free to come to love God, points out that, were God to create the person in a state where he or she (from the moment of creation) already loves God, the person would not be free to come to love God. This is certainly true, but somewhat irrelevant to the point at issue. The question at issue is whether a person can exist in epistemic immediacy with respect to God and still be free to come to love God. Such a person would know God's power and greatness, be aware of God's immense love for all his creatures, and so on. But isn't all this quite compatible with the person not responding to God's love with love of his or her own? To say that a person's freedom to come to worship and love God would be destroyed in such a situation is, I think, to confuse coming to have a very good reason to worship and love God with being compelled to worship and love God. It is, alas, part of the very nature of freedom to have the power not to do what one has a very good reason to do. So, while I fully agree with Hick that epistemic distance from God is necessary to cognitive freedom in relation to God, I can find no good reasons in Hick's writings to support his further claims that epistemic distance from God is necessary for the very existence of human persons, for their being free to develop morally, and for their being free with respect to coming to love God. And, if I am right about this, one must wonder about what good is served by our state of epistemic distance from God. Of course, without epistemic distance we would come to know God of necessity, not freely. And perhaps Hick would add a third to the two principles concerning intrinsic value noted earlier:

THE PROBLEM OF EVIL

William Rowe

111

117

freely coming to know of a being's existence is intrinsically better than coming of necessity to know of the being's existence. But it is far from clear that this third principle is true. Let us, however, put aside these qualms about what epistemic distance is required for, and return to the thread of the argument. So far, I think, Hick has built a reasonable case for the hypothesis that (1) if we believe there is an omnipotent, perfectly good being, it would not be unreasonable for us to believe that this being might have the realization of the two goods noted above as his goal, and (2) if such a being did have the realization of these two goods as its goal, it would permit the existence of moral and natural evil in the world. For, unless there are real obstacles in nature to overcome, and unless human beings are capable of doing real harm to one another, freely attained moral and spiritual growth would be practically, if not theoretically, impossible. Has Hick, then, provided a reasonable solution to the problem of evil? As we have already seen, this is not just one question but four, for there are (at least) four distinct problems of evil (or four distinct levels of the problem of evil). The answer to our question depends, therefore, on which problem of evil we have in mind. My own view is that Hick has provided a reasonable answer to our first question: why an omnipotent, perfectly good being would permit any evil in the world. Moreoever, if we limit kinds of evil to natural evil, moral evil, human pain and suffering, and so on, I also think that Hick has provided a reasonable answer to our second question: why an omnipotent, perfectly good being would permit instances of the various kinds of evil to occur. But thus far nothing has been said that would explain why such a being would permit the amount of evil (of these kinds), or certain particular evils that exist in our world. The plain fact, as Hick recognizes, is that evil occurs in such massive amounts in our world that it often as not defeats the development of moral and spiritual growth. Why then does God permit evil in such massive amounts? If the excessive amounts of evil were to fall on those humans who are particularly recalcitrant to moral and spiritual development, we might reason that God allows it to fall on them to enable them to become more sensitive to their need for such development. But incredible amounts of pain and suffering fall equally on the innocent and the guilty. Moreover, evil falls so unrelentingly on some people (whether saints or sinners) that it can only be seen by us as destructive of

112

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 118

Paradox and Promise

the soul-making enterprise. And, when we look at particular evils, it seems ludicrous to suppose that, had God prevented any one of these evils, someone's moral and spiritual development would thereby have been prevented or in some way frustrated. That Hick recognizes the need for theodicy to address the amount and intensity of evil in the world is clear from the following remark: It [theodicy] attempts to explain how it is that the universe, assumed to be created and ultimately ruled by a limitlessly good and limitlessly powerful Being, is as it is, including all the pain and suffering and all the wickedness and folly that we find around us and within us. (Emphasis added)12 What then is Hick's response? So far as I can determine, Hick takes the remaining difficulties to be principally three: (1) the sheer amount and intensity of evil in the world; (2) the fact that this evil falls randomly and haphazardly on both the just and unjust; and (3) the question of whether the postulated goods for the sake of which God permits such evils can be worth the price (Dostoevsky's question). My interest here is Hick's attempt to deal with the first two of these difficulties. That he takes these to constitute major difficulties for his theodicy is evident from the following: the problem does not consist in the occurrence of pain and suffering as such; for we can see that a world in which these exist in at least a moderate degree may well be a better environment for the development of moral personalities than would be a sphere that was sterilized of all challenges. The problem consists rather in the fact that instead of serving a constructive purpose pain and misery seem to be distributed in random and meaningless ways, with the result that suffering is often undeserved and often falls upon men in amounts exceeding anything that could be rationally intended.13 What seems obvious to Hick and to us is (1) that the amount and intensity of evil in our world far exceeds what is needed for soul-making, and (2) that the evils in our world are distributed in a haphazard fashion, apparently unrelated to anyone's stage of development in soul-making. In the light of this, how can anyone seriously propose the good of soul-making as the reason for God's permission of all the pain and suffering in our world?

THE PROBLEM OF EVIL William Rowe

113 119

Hick's persistent answer is to employ what he calls the method of "counterfactual hypothesis" and to emphasize the importance of mystery in soul-making. Let us see how the argument goes with respect to the fact that the amount and intensity of evil in our world appears far to exceed what could be rationally intended for soul-making. In response, Hick asks us to consider a world in which no evil occurs in an amount beyond what is needed to play a role in significant soul-making. Moreoever, he asks us to suppose that we all know that this is so. He then argues that the result would be that we would make no significant efforts to overcome evil. But it is precisely such efforts (or the need for them) that lead to significant moral growth and development.14 A similar line of argument is developed for the haphazard distribution of evil among the just and unjust. In a world in which suffering by a person is permitted only if it is merited or needed for soulmaking, then, if we further suppose that we all know this to be so, no one would make efforts to relieve the suffering of others.15 Paradoxically, then, soul-making would be considerably limited in a world in which we all knew or rationally believed that suffering is permitted only as it is required for soul-making. There is, I believe, real merit to this line of argument. To help us understand and assess Hick's argument, I shall use the expression "an excess evil", where "e is an excess evil" means that e is an evil that an omnipotent being could have prevented without loss of significant soul-making. The point of Hick's argument seems to be this. Significant soul-making requires not only the existence of evils, but also that it be rational for us to believe that excess evils exist; it must be rational for us to believe that evils occur that omnipotence could have prevented without loss of significant soul-making.17 For, if we were to believe that each evil that occurs is one that even an omnipotent being could not prevent without loss of soul-making, we would make no significant efforts to overcome such evils. And, as we have noted, it is precisely such efforts that are crucial to significant moral growth and development. Significant soulmaking, then, has two requirements. First, it has a factual requirement: there must be real evils to be overcome. Secondly, it has an epistemic requirement: it must be rational for us to believe that excess evils occur in our world. This second requirement has the air of paradox. It seems to say that evils not needed for soulmaking are, after all, needed for soul-making. But it does not say this. What Hick's paradox says (roughly) is that rationally believ-

114

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 120

Paradox and Promise

ing that there are evils not needed for soul-making is, after all, needed for soul-making. And, although paradoxical, such a claim is not incoherent.18 How does Hick's argument strengthen his theodicy? Well, having noted that soul-making requires real evils to be overcome, the problem was that it seems obvious to us that the amount and intensity of evil is far in excess of what an omnipotent being would have to permit for significant soul-making to occur. Hick's ingenious response is that, if it were not rational for us to believe that excess evil occurs, soul-making would be significantly diminished. Some might reject his claim. I am inclined to accept it. And what this implies is that the amount, intensity and distribution of evil in our world must be such as to create and sustain our belief that evils occur in excess of what an omnipotent being would need to permit for our moral and spiritual growth. Suppose we grant the force of Hick's argument. My objection to it is that it does not really solve the problem of the amount and intensity of evil in our world. For it not only seems obvious to us that evil occurs far in excess of what an omnipotent being would have to permit for soul-making; it also seems obvious to us that evil occurs far in excess of what an omnipotent being would have to permit for us to be rational in believing that excess evil occurs. Clearly, if there is an omnipotent being, such a being could have prevented a good deal of evil in our world without in the least altering the fact that the amount and intensity of evil makes it rational for us to believe that evils occur in excess of what an omnipotent being would need to permit for our moral and spiritual growth. Hick's argument does show that our world must have enough evil to support the belief that there are excess evils. But, since it is clear that evil occurs far in excess of what is necessary to support such a belief, Hick's argument does not solve the problem of the amount, intensity and distribution of evil in our world. Hick's argument has the defect of being an "all-or-nothing argument" or a "where will it all end argument". Let me illustrate this defect by an example. When I first taught at Purdue many, many years ago, the teaching-load was twelve hours a week (four courses a semester). I recall a discussion with a Purdue administrator in which I suggested that the quality and quantity of research might be improved by reducing the teaching-load to nine hours a week. He looked at me with a shocked expression, and

THE PROBLEM OF EVII William Rowe

115 121

then proceeded to point out that, if we were to reduce the teaching load to nine, the same reasons would apply and we would then have to reduce the load to six hours a week. But, again, the same reasons would apply and we would eventually have to reduce the teaching load to no hours, with the result that the university would close. "And so, Professor Rowe," he triumphantly concluded, "you see that the teaching load must be kept at twelve hours a week." This argument is a "where will it all end" argument, or an "all-or-nothing argument". If it were a good argument, which it is not, it would have justified a teaching-load of fifteen, eighteen, twenty-one, twenty-four or even more hours a week. (I refrained from pointing out to the administrator this possible extension of the argument.) And I cannot help but believe that Hick's argument suffers from the same defect. We look around our world and observe massive amounts of evil, enormously beyond what anyone would expect in a world governed by an omnipotent, perfectly good being bent on providing its creatures with an environment in which they could grow morally and spiritually. We say to Hick, "Surely this is not a world governed by an omnipotent, perfectly good being bent on providing its creatures with an environment in which they can grow, morally and spiritually." In reply Hick says, "But don't you see: if the apparently excess evil were cut in half, the same objection would still apply. So the only real alternative to the actual situation is one in which the amount of apparently excess evil is reduced to zero, one in which it is obvious to all of us that no evil or degree of evil is permitted unless its permission is absolutely essential to moral and spiritual growth. But, if that were the way of it, none of us would actively seek to prevent or overcome evil, with the result that significant moral and spiritual growth would be thwarted."19 The mistake in this argument is the more-or-less implicit assumption that the same objection would be applicable until apparently excess evils were reduced to the level of zero. If Hick is right in holding that some measure of apparently excess evil is required for significant soul-making, then it must be acknowledged that a world fit for moral and spiritual growth would be one in which some evil appears to us as excessive and unrelated to soul-making. There would be a threshold below which the deity could prevent such evils only at the cost of limiting moral and spiritual development. But isn't it abundantly clear that we are far above that threshold? For Hick's argument to work he needs to

116

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 122

Paradox and Promise

contend that, were the amount and degree of evil in our world that we take to be beyond what could be "rationally planned" any less than it in fact is, there would probably be some loss in soulmaking. What he in fact shows is that, were the amount and degree of such evil reduced to zero, there would be some loss in soul-making. In like manner, the Purdue administrator does show that, were the teaching-level reduced to zero, some unfortunate consequence would ensue. What he needs to show is that, were the teaching-load reduced to something significantly below twelve but significantly above zero, that same, or some other, unfortunate consequence would ensue. Earlier I quoted Hick's statement that what a theodicy aims to do is explain, on the assumption that there is an omnipotent, perfectly good being, how it is that the universe "is as it is, including all the pain and suffering and all the wickedness and folly that we find around us and within us". I have argued that, if this is the aim of theodicy, we must judge Hick's own theodicy as falling substantially short of its goal. In terms of the four problems of evil I noted earlier, I think we can say that Hick's theodicy does seem to solve the first two problems: Hick provides a reasonable answer to the question of why an omnipotent, perfectly good being would permit any evil at all in the world, and he goes some way toward providing a reasonable answer to the question of why an omnipotent, perfectly good being would permit instances of the different kinds of evil to occur in the world. Moreover, I think his final argument enables us to see how an omnipotent, perfectly good being would be justified in permitting instances of pain and suffering that appear to us to be excessive and unrelated to the soul-making process. But, when we turn to our final two problems - the explanation of the amount and degree of intrinsic evil, and the explanation of particular evils - I believe Hick's theodicy, like all other theodicies, fails to offer any believable solution. And, since it is particular evils, and the amount and degree of evils, that are generally regarded as providing the strongest basis for rational disbelief in an omnipotent, perfectly good being, I conclude that Hick's theodicy, although perhaps the best we have, does not succeed in turning aside the strongest arguments for disbelief based on the facts of evil.

THE PROBLEM OF EVIL

William Rowe

117

123

NOTES 1.

2.

3. 4. 5.

6.

7. 8.

9. 10. 11. 12.

See, for example, Evil and the God of Love, 2nd edn; God and the Universe of Faiths; Philosophy of Religion, 3rd edn, ch. 4; and Stephen T. Davis (ed.), Encountering Evil: Live Options in Theodicy (Atlanta: John Knox Press, 1981). This statement is quite close to Hick's general formulation of the problem: "Can the presence of evil in the world be reconciled with the existence of a God who is unlimited both in goodness and in power?" (Evil and the God of Love, p. 3). In "Evil and Theodicy", Philosophical Topics, 16 (1988) 119-32, I have argued that this response is inadequate. In "The Problem of Evil and Some Varieties of Atheism", American Philosophical Quarterly, 16 (1979) 335-41,1 labelled this response by the theist "the G. E. Moore Shift". I hold that freely performing act A is logically inconsistent with being causally determined to do A. Hick holds a different concept of freedom. But he emphasizes authentic freedom, which in fact has implications similar to the implications of my concept of freedom. The second of these good states (and probably the first) obtains only if God exists. But, as I have suggested, in order to be successful a theodicy need not establish that the good in question obtains. See G. E. Moore, Principia Ethica (Cambridge: Cambridge University Press, 1903) pp. 27-31. Several criticisms have been advanced against Hick's efforts to explain why God would permit (l)-(4) to obtain. Hick replies to some of these criticisms in chapter 17 of Evil and the God of Love, 2nd edn. For responses to further criticisms and a strong defence of soul-making theodicy see William Hasker, "Suffering, SoulMaking and Salvation", International Philosophical Quarterly, 28 (1988) 3-19. Encountering Evil, pp. 42-3. Ibid., p. 43. Philosophy of Religion, 3rd edn, pp. 45-6. Encountering Evil, p. 39. It would be a mistake, I think, to understand Hick to be here charging theodicy with the task of explaining why God would permit this or that particular evil - if to provide such an explanation is to show how God's permission of that evil is necessary for some particular exercise of human freedom, or some particular bit of moral and/or spiritual development. It would also be a mistake to take the passage quoted as implying that Hick thinks that a theodicy must explain the precise amount and precise degree of intensity of evil in the world - if to provide such an explanation is to show why God's permission of just that amount or that degree of intensity was necessary for optimizing moral and spiritual growth through free human choices and actions. Instead, I think Hick's view is that a theodicy need only explain why our world contains a considerable amount

118

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION

124

13. 14. 15.

16.

17. 18.

19.

Paradox and Promise of evil and why it sometimes appears to us that some evils could have been prevented by God without any loss of soul-making. In addition, I think he holds that it must not be unreasonable to believe that the goods proposed may well be the reason why an omnipotent, perfectly good being would permit the world to be as it is. (I take it, though, that Hick would hold that if God exists there is an explanation of why he permits particular evils, and why he permits the precise amount and intensity of evil in the world.) Evil and the God of Love, 2nd edn, p. 333. Ibid., pp. 334r-5. Hick also notes that, if punishment for evil doing were immediate and apparent to us, morality would be replaced by prudential self-interest. See God and the Universe of Faiths, p. 59; Evil and the God of Love, 2nd edn, pp. 333-6. Perhaps there are two equally severe evils such that either could be prevented without loss of soul-making, but one must be permitted to prevent such a loss. I shall ignore such complications in order to keep the notion of an excess evil as simple as possible. A complete account of the notion would need to take account of such possibilities. Actually, all the argument technically shows is that it must not be true that it is rational for us to believe that no excess evils exist. An analogy to Hick's paradox might be the following. Suppose a marathon runner is such that, if he believes that he will win, he won't train and, therefore, won't win. But, if he has grounds for believing that he will lose, he will train to the utmost so as to come as close to winning as he can. Of such a person it might be correct, although paradoxical, to say, "rationally believing that he won't win is, after all, required if he is to win". For examples of "all-or-nothing arguments", see Encountering Evil, pp. 49-50; and Evil and the God of Love, pp. 327-8, 333-6.

CHAPTER SEVEN

Ruminations about Evil

Some years ago, in an effort to set forth an argument for atheism based on certain cases of intense human and animal suffering, I claimed that we have reason to believe that 1. There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. Now, if we have reason to believe (1), then since even many theists will accept 2. An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse. we have reason to conclude that the God of traditional theism (an omnipotent, omniscient, wholly good being) does not exist.1 Confronted with this argument, theists, I suggested, would be wise to acknowledge that it does provide some rational grounds for atheism. Whether atheism is the position that from the rational point of view we all ought to adopt, however, will depend on what else we know or have reason to believe that is relevant to whether theism or atheism is true. While some philosophers may have been persuaded both of the strength of the argument and of the wisdom of the advice, others

120

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 70 / William L. Rowe have taken a more critical position; they have argued either that the considerations given in support of premise (1) are wholly inadequate to provide it with rational support or that premise (2), despite its appearance of being a necessary truth, is a proposition of dubious intellectual merit. In this paper I want to examine two important attacks on the claim that we are rationally justified in accepting premise (1), and one important attack on premise (2). In the course of discussing these attacks, I hope to cast the original argument into a form that more clearly reveals why particular instances of evil count against the existence of the God of traditional theism.

I Alvin Plantinga raises the question of what makes (1) probable.2 He then notes my answer. First, I think, is the fact that there is an enormous variety and profusion of intense human and animal suffering in our world. Second, is the fact that much of this suffering seems quite unrelated to any greater goods (or the absence of equal or greater evils) that might justify it. And, finally, there is the fact that such suffering as is related to greater good (or the absence of equal or greater evils) does not, in many cases, seem so intimately related as to require its permission by an omnipotent being bent on securing those goods (the absence of those evils) (561). Putting these points of mine together into a single claim, Plantinga suggests that I take (1) to be probable with respect to 3. Many cases of evil are apparently pointless. Suppose we agree that it is something like (3) that is being claimed to provide support for (1). Plantinga thinks there is a problem with (3), a problem of understanding what one is claiming in asserting (3). He suggests that on one reading (1) is probable with respect to (3), but on this reading (3) is not something we could know to be true. On a second reading, Plantinga notes that (3) turns out to be something of which we can be certain, but (1) is not in the least probable with respect to this reading of (3). Here is his discussion. But how shall we understand Rowe here? Shall we see him as holding that in fact there are many cases of evil such that it is apparent that an omnipotent and omniscient God, if he existed, would not have a reason for permitting them? But this is much too

THE PROBLEM OF EVIL Ruminations about Evil / 71 strong; ...we could sensibly claim something like this only if we had reason to think that if such a God did have a reason for permitting such evils, we would be likely to have some insight into what it is. But if theism is true, then this is false; from the theistic perspective there is little or no reason to think that God would have a reason for a particular evil state of affairs only if we had a pretty good idea of what [that] reason might be. On the theistic conception our cognitive powers, as opposed to God's, are a bit slim for that. God might have reasons we cannot so much as understand; he might have reasons involving other free creatures—angels, devils, the principalities and powers of which St. Paul speaks—of which we have no knowledge. Shall we take (3) as pointing out, then, just that there are many evils such that we have no idea what God's reason, if any, is for permitting them? That seems right; but why suppose (1) is probable with respect to it? We could sensibly claim that (1) is probable with respect to (3) (taken thus) only if we had good reason to think we would be privy to God's reasons for permitting evil. (561) The problem, as Plantinga sees it, is to understand (3). He poses a dilemma. Either (3) is to be understood as 3a. It is apparent to us that certain evils are pointless.3

or 3b. It is not apparent to us that certain evils have a point. I agree with Plantinga that (3a) is too strong and (3b) too weak to provide proper support for (1). (3a) implies that we know that certain evils are pointless. And clearly this is something we do not know. (3b), as Plantinga rightly notes, would support (1) only if it is conjoined with some claim to the effect that if there were a God it is likely we would know what his reason is for permitting these evils. But has Plantinga got it right? When we consider (3) are we driven to understand it as either (3a) or (3b)? I don't believe so. When I say: His remark was apparently misleading. must I be saying either that it is apparent to me that his remark was misleading or that it is not apparent to me that his remark wasn't misleading. No. For the first is stronger and the second weaker than what I said. What I said implies that there was something about the situation that makes it reasonable to believe that his remark was misleading. And what I'm saying about certain evils is that there is something about them and the situation in which they occur that

121

122

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 72 / William LRowe makes it reasonable for us to believe that they are pointless. What is it about certain cases of evil that makes it reasonable for us to believe that they are pointless? I might here simply reiterate the three points Plantinga quotes at the outset of his discussion. But it will be helpful, I think, to try to answer this question without using expressions like 'appears' or 'seems'.4 Consider, for example, our hypothetical case of the fawn's intense suffering over several days as a result of being severely burned in a forest fire started by lightning. Or consider the actual case of a 5-year old girl who was raped, severely beaten over most of her body and strangled by her mother's boyfriend.5 Let's refer to the first case as El and the second case as E2. Now about El and E2 I make the following judgment. P. No good state of affairs we know of is such that an omnipotent, omniscient being's obtaining it would morally justify that being's permitting El or E2.6 What am I implying in making this assertion? I am implying that we have good reason to believe that no good state of affairs we know of would justify an omnipotent, omniscient being in permitting either El or E2.1 don't mean simply that we can't see how some good we know about (say, my enjoyment on smelling a fine cigar) would justify an omnipotent being's permitting El or E2.1 mean that we can see how such a good would not justify an omnipotent being's permitting El or E2. For we can see that an omnipotent being wouldn't have to permit El or E2 in order to obtain the good of my enjoyment on smelling a fine cigar. And we can see that even were that not so, obtaining such a good wouldn't justify any being in permitting El or E2. Is there some other good state of affairs we know of that would justify an omnipotent being in permitting El or E2? I don't believe there is. The good states of affairs I know of, when I reflect on them, meet one or both of the following conditions: either an omnipotent being could obtain them without having to permit El or E2, or obtaining them wouldn't morally justify that being in permitting El or E2. And if this is so, I have reason to conclude that: Q. No good state of affairs is such that an omnipotent, omniscient being's obtaining it would morally justify that being in permitting El or E2.7 Since Q, slightly qualified, is tantamount to (l)'in the earlier argument for atheism, if we are justified in accepting P and justified in inferring

THE PROBLEM OF EVIL Ruminations about Evil / 73 Q from P, we are justified in accepting premise (1). To simplify our discussion, I will use the letter 'J' to stand for the property a good has just in case obtaining that good would justify an omnipotent, omniscient being in permitting El or E2. If a good is such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2, then that good has J; and if not, the good lacks J. In briefer terms, then, I have claimed P (no good we know of has J), inferred Q (No good has J), and concluded that since Q is tantamount to (1) we are justified in accepting (1). The main point in Plantinga's earlier criticism is that we can get to (1) only if we have a good reason to think that if there is a good that has J, we would likely know of that good and likely see that it has J. Now since Q is tantamount to (1), we can take Plantinga to be claiming that we are justified in inferring Q (no good has J) from P (No good we know of has J) only if we have a good reason to think that if there were a good that has J it would be a good we are acquainted with and could see to have J. For the question can be raised: How can we have confidence in this inference unless we have a good reason to think that were a good to have J it would likely be a good within our ken? My answer is that we are justified in making this inference in the same way we are justified in making the many inferences we constantly make from the known to the unknown. All of us are constantly inferring from the A's we know of to the A's we don't know of. If we observe many A's and note that all of them are B's we are justified in believing that the A's we haven't observed are also B's. Of course, these inferences may be defeated. We may find some independent reason to think that if an A were not a B it would likely not be among the A's we have observed. But to claim that we cannot be justified in making such inferences unless we already know, or have good reason to believe, that were an A not to be a B it would likely be among the A's we've observed is simply to encourage radical skepticism concerning inductive reasoning in general. In considering the inference from P to Q it is very important to distinguish two criticisms: A. One is entitled to infer Q from P only if she has a good reason to think that if some good had J it would be a good that she knows of. B. One is entitled to infer Q from P only if she has no

123

124

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 74 / William L. Rowe reason to think that if some good had J it would likely not be a good that she knows of. Plantinga's criticism is of type A. For the reason given, it is not a cogent criticism. But a criticism of type B is entirely proper to advance against any inductive inference of the sort we are considering. For if one does have some independent, positive reason to think that were an A not a B it would likely not be among the A's we've observed, then one's confidence in the inference must be diminished. Thus one important route for the theist to explore is whether there is some reason to think that were a good to have J it either would not be a good within our ken or would be such that although we apprehend this good we are incapable of determining that it has J. Before we explore a criticism of type B, however, we need to consider another point Plantinga makes against the view that it is likely that (1) is true. It is part of the Free Will Defense to claim that it is logically possible that an omnipotent, omniscient being could not have created a world with less evil, but as much good, as our world contains. So long as we accept (3) that a world with as much good as our world would have to contain significantly free beings, and (4) that freely performing an action is not compatible with being causally necessitated to perform that action, the argument Plantinga gives for this important claim in the Free Will Defense is, I think, compelling.8 Now if this logical possibility were actually true then an omnipotent, omniscient being would not have been able to improve on our world; our world would be on balance as good as any that it was in his power to actualize. All this, I think, is right. But how does it bear on our argument that it is in fact likely that an omnipotent, omniscient being could have done a better job of things? For reasons that pass me by, Plantinga thinks that this claim at the heart of the Free Will Defense should lead us to reject the view that we have reason to think it likely that an omnipotent, omniscient being could have done better. I'm sure it strikes most of us as just plain reasonable to think that if there were an omnipotent, omniscient being, then a bit more activity on his part would have made the world somewhat better. Suppose Hitler had been made to die in his sleep prior to carrying out his plan to exterminate the Jews. Wouldn't this have made things better? Probably, I would say. Perhaps, but perhaps not, Plantinga

THE PROBLEM OF EVIL

Ruminations about Evil / 75 would say. Now I grant that we don't know that things would have been better. It is possible that had Hitler died in his sleep some other person would have taken his place and brought about more evil than even Hitler did. But surely it is reasonable to believe that the number of German leaders who would have outdone Hitler in wickedness is relatively small. Or are all such conjectures just an epistemic tossup? No one, I believe, really thinks this way. The mere fact that it is logically possible that an omnipotent, omniscient being couldn't have made things better is no real reason at all to doubt that such a being in fact could have made things better. Consider the proposition: If an omnipotent being had caused Hitler to die in his sleep, G. E. Moore would have been so distressed at the news of Hitler's death that he would have ended up promoting more evil than Hitler did. This proposition, no less than the proposition at the heart of the Free Will Defense, is logically possible. But not much can be gained from that fact to make us doubtful of the future course of Moore's moral life, even had an omnipotent being caused Hitler to die in his sleep. Perhaps, however, we are assuming the world's free creatures to consist of just human beings like us, with the possible exception of an omnipotent, omniscient being. And perhaps it is just this assumption that Plantinga is unwilling to make. For all we know, there is a great variety of free creatures involved in our history and in the history of our world—creatures of whose nature and activity we are at best but dimly aware. Perhaps there are angels and devils (Satan and his cohorts), perhaps there are the principalities, powers and dominions of whom St. Paul speaks. We know far too little about the world that is in fact actual, far too little about the sorts of creatures it contains and the counterfactuals of freedom that characterize them, to be justly confident of opinions about God's alternatives. (565)

The idea here, I take it, is that if it were only human beings and their freedom that we had to worry about, some reasonable judgments perhaps could be made to the effect that an omnipotent, omniscient being could have done better.9 But who knows how Neptune or one of Satan's cohorts would have reacted if Hitler had been made to die in his sleep? And who among us is in a position to say what would have gone on in the principalities, powers and

125

126

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 767 William L Rowe dominions, if such there be? The result, then, is that once we recognize the possibility that such beings as these are studying the daily twists and turns of our world, we really cannot make any rational judgments at all about what would have happened had an omnipotent being endeavored to make things better. But again, why should the mere logical possibility of the existence of Neptune, Satan's cohorts, etc., tell against our reasonable judgments of what would have transpired had an omnipotent, omniscient being caused Hitler to die in his sleep, prevented the fawn's agony, or prevented the mother's boyfriend from raping and savagely murdering the 5 yr. old girl? II

The second criticism I want to discuss is due to Stephen Wykstra.10 This criticism consists of two related points. The first point may be viewed as an attack on our assertion that no good we know of has J. How do we tell whether a good has J? Well, Wykstra might suggest, we contemplate that good and find ourselves unable to see how it could morally justify an omniscient, omnipotent being in permitting El or E2. Having looked hard at that good and seeing no way in which it would justify such a being in permitting El or E2, we conclude that that good doesn't have J. But, Wykstra objects, this procedure is reliable only if having J has "seeability."11 Our failure to see an elephant in a room is a good reason to conclude that one isn't there. But our failure, after careful inspection, to see a sand flea in a room is not a good reason to think that one isn't there. The reason for this is that the sand flea has what Wykstra calls "low seeability."12 Wykstra's second point is that there is good reason to think that were a good to have J it would likely be among the goods with which we are not acquainted. This is the point we need to consider carefully, for it is a type-B criticism of the inference from P to Q. If Wykstra is correct on this point, therefore, we have an effective attack against the argument for atheism based on the occurrence of certain types of evils. Before examining this point, however, it will be helpful to make a brief remark concerning his first point. When we contemplate the good of my sitielling a fine cigar, is it true that we see that this good lacks J, or do we simply fail to see

THE PROBLEM OF EVIL Ruminations about Evil / 77 that it has J? The answer, I think, is clearly the former. For we see clearly that this good is not good enough to justify any being in permitting El or E2. We also see clearly that obtaining this good would not require such a being to permit El or E2. So the absence of J is something we can determine, often with certainty, when we consider various goods with which we are acquainted. What is much more difficult, I believe, is to determine of some good that it has J. For suppose we do know of some good that far outweighs in value either El or E2. And suppose we see by the light of nature that even an omnipotent, omniscient being could not obtain that good without permitting El or E2. Would we be in a position to affirm for sure that the good in question has J? No. For it may be, for all we know, that there is some good state of affairs G* that significantly exceeds G in value and is such that an omnipotent, omniscient being can obtain it only by preventing El and E2. If this were so, then, even though an omnipotent being could obtain G only by permitting El or E2, G might very well lack J.13 I said earlier that Wykstra's main point is that there is reason to think that were a good to have J it would likely be among the goods with which we are not acquainted. In fact, however, this does not quite reflect his position. His view is that were a good not only such that obtaining it would justify an omnipotent, omniscient being in permitting El or E2, but also such that it is purposed by the omnipotent, omniscient, creator of all things as His reason for permitting El or E2, then that good would very likely be a good that is beyond our ken. Now we can ask what Wykstra's purposed by the omnipotent, omniscient, creator of all things adds to a good that has J. It adds, I think, two points. First, a good can have J even if there is no omnipotent, omniscient being. For to have J a good need only be such that obtaining it would justify an omnipotent, omniscient being in permitting El or E3. And that may be true of a good even though there is in fact no omnipotent, omniscient being. But, of course, if a good is purposed by an omnipotent, omniscient being as his reason for permitting El or E2, then there must be such a being. Second, a good that is purposed by the creator of all things would presumably be part of a large-scale plan that the creator has for the world as a whole. It would fit into some coherent plan that the creator has for his entire creation. What kind of overall moral plan might such a creator have for his creation? Wykstra distinguishes between a "morally shallow" uni-

127

128

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 78 / William L. Rowe verse and a universe with "moral depth/' arguing that such a being would likely plan the latter sort of universe. If God exists, then ...he would allow an evil only if doing so served some outweighing good. Such a good would be the "moral (or axiological) cause" for his allowing the evil. If God exists, there is some such cause for each of his allowings of evil. But here there are two options. We might think that if our universe is the creation of God, these moral causes would likely be "near the observable surface" of their effects. ...In contrast, we might judge that if our universe is the creation of God, it would likely have great "moral depth." By this I mean that many of the goods below its puzzling observable surface, many of the moral causes of God's current allowings and intervenings, would be "deep" moral goods. The question is this: if our universe is the creation of God, is it more likely to be morally shallow or morally deep? ...if our universe is the creation of God, a God with the sort of wisdom and vision entailed by theism, then it is eminently likely that it is morally deep rather than morally shallow.14

We might quarrel with Wykstra's way of framing the issue. What is "morally deep" sounds better than what is "morally shallow". But these are here only epistemic expressions. The difference is between a world in which it is understandable to us what the goods are (or might plausibly be) that justify God in permitting El or E2, and a world in which we are left without a clue. And what Wykstra is saying is that it is more likely that an omnipotent, omniscient, wholly good being who is the creator of everything would bring about the second kind of world, rather than the first. Unfortunately, Wykstra doesn't tell us why this is so. Wykstra suggests that God may have had a choice about El and E2. On the one hand, he could have produced a world in which the goods for the sake of which he permits these evils—and others like them—are within our ken (the "morally shallow" world). On the other hand, he could have produced a world in which the goods for the sake of which he permits these evils are beyond our ken (the "morally deep" world). Faced with these alternatives, it is highly likely, so Wykstra thinks, that he would choose the latter. Why? Is it that the latter world is somehow better? If it is a better world, it must be that the goods that justify El and E2 in the deep world are much better, other things being equal, than the goods that justify El and E2 in the shallow world. For clearly, a good deal of suffering in this world is occasioned by the simple fact that so far as we can see there is

THE PROBLEM OF EVIL Ruminations about Evil / 79 no good that justifies an omnipotent, omniscient being in permitting El or E2. If we could see these goods occurring in our lives and could have some glimmer as to how they are obtainable by such a being only through permitting El and E2, it is quite likely that the grief and suffering occasioned by El and E2 would be less than it is. Until we are provided with some reasons for thinking that if there were an omnipotent, omniscient, wholly good being who created our world, then the goods in virtue of which he permits El and E2 would be undetectable by us, we certainly are within our rights to infer Q from P and conclude that it is likely that there is no such being.15 If we take standard theism to be the view that there exists an omnipotent, omniscient, wholly good being who is the creator of the world, the question before us is whether the fact (or rational belief) that P (No good we know of has J) is a reason for thinking that standard theism is false. If we have some good reason to think that such a being would likely permit El and E2 only by virtue of goods that are beyond our ken, then P will not be a reason to think that no such being exists. For the reason for thinking that such a being would likely permit El and E2 only by virtue of goods that are beyond our ken will diminish the strength of our inference form P to Q. So the task for Wykstra is to show how standard theism implies (or makes it quite likely) that our world is morally deep (obscure). If we simply attach to standard theism the view that The goods for the sake of which an omnipotent, wholly good being must permit vast amounts of human and animal suffering are beyond our ken and will be realized only at the end of the world, we can create an expanded version of theism that will not be rendered unlikely by P. For this expanded theism will make it likely that if Q were false, P would be true anyway. But as Wykstra and Russell point out, the danger here is that one may purchase immunity at the price of enlarging the original hypothesis, thus increasing (decreasing) its original implausibility (plausibility).16

Ill The final objection to be discussed is radical, if not revolutionary.

129

130

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 80 / William L. Rowe This objection is due to William Masker.17 Instead of attacking the reasoning in support of premise 1 (in the original argument) and the inference from P to Q (in the revised argument), Masker attacks premise 2 (in the original argument) and the inference from Q to atheism (in the revised argument). For Masker thinks that if theism is true then, given our moral obligations to refrain from harming others, there would have to be gratuitous evils. [A gratuitous evil is an instance of evil such that an omnipotent being could have prevented it without thereby preventing the occurrence of some greater good. (p. 2)] Masker reasons as follows. If God prevents any evil that does not lead to a greater good, morality will be undermined —for no sense could then be made of principles prohibiting wrongdoing or the infliction of harm. Given, then, that morality is not to be undermined, God must permit evils that do not lead to any greater good. But for God to permit such evils is nothing more than for gratuitous evils to exist. Although Masker elaborates his view in terms of both consequentialist and deontological moral theories, his central point can best be grasped in terms of a consequentialist moral theory. If God prevents any evil that does not lead to a greater good then it will be impossible for anyone to act wrongly. For a wrong act would be among those that do not lead to a greater good. But what then becomes of morality? The whole fabric of moral principles prohibiting us from harming others and from wrongdoing would simply collapse. For moral principles prohibiting such actions are intelligible only if it is possible for us to perform them. My answer to Masker is twofold. First, if there is a God and he prevents all gratuitous evil, it hardly seems that we are deprived of the possibility of significantly harming others. We can still torture and murder the innocent. These are paradigm cases of harming, of wrongdoing. Of course, God would prevent these deeds unless He could not do so without thereby foregoing obtaining some good that justifies Him in permitting them. Do these actions then become morally right? It hardly seems so. The actions themselves do not produce the greater good. It is God in his freedom (according to theism) who is responsible for obtaining the justifying goods. But let's put this consideration aside, and pursue a second answer to Masker's objection. In commenting on an earlier version of Masker's important paper,18 I argued that if an evil is such that its prevention by an

THE PROBLEM OF EVIL Ruminations about Evil / 81 omnipotent, omniscient being would undermine morality, then if morality is a good to be preserved, God may permit that evil in order to achieve the good of sustaining morality. So evils whose permission is not justified by any other goods may be justified by the good of sustaining the application to us of fundamental moral principles. But if this is so, these evils are not after all gratuitous. I then concluded: And what this means, I'm afraid, is that Masker's argument contains the seeds of its own destruction. A world suitable for significant morality would not require an omnipotent, omniscient, perfectly good being to permit any gratuitous evil at all.19 In the version of his paper now being considered, Masker argues that my answer is inadequate because a person could still reason that any evil action he performs that serves no other good will be permitted by God only if it serves the greater good of maintaining significant morality. But now, surely, we have reached a situation exactly parallel to the one which led to our previous conclusion about the undermining of morality. The agent will be able to say to himself that if he performs the action in question, the resulting evil is permitted by God as the most economical means to an end of overriding importance—namely, the maintenance of significant morality. So once again, morality is undermined, ...(p. 28) What Masker is arguing here, I think, is this. Suppose God permits evils (otherwise gratuitous) so as to maintain significant morality. Presumably, there would be a threshold beyond which his permission of evils for this purpose (the maintenance of significant morality) would no longer be required. Beyond that threshold, such evils would be genuinely gratuitous and, therefore, God would not permit them to occur.20 My point was that the evils permitted up to the threshold would not be gratuitous. Masker acknowledges this, but argues that morality, nevertheless, would still be undermined. For a person contemplating performing some evil deed will know that God's permitting the deed will mean that some justifying good is obtained, either (a) a good unrelated to the maintenance of morality, or (b) the good of sustaining morality. If the person knows that (a) is the case, morality is undermined. The prohibition against doing the evil deed will make no sense. But the person doesn't know this. For all he knows, it is (b) that is the case. But here, Masker argues, our agent can reason further. First, he can infer that if (b) is so then God will

131

132

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 82 / William L. Rowe permit the evil so as to maintain the threshold of evils required for morality. And second, he can infer that if he were to refrain from his evil deed, God would have to permit some other evil no less severe in order to maintain the threshold at the level required to sustain morality. The result, in either case, is that the agent is powerless to make the world better by refraining from performing any putative evil deed. The prohibition against wrongdoing will make no sense. Morality is again undermined. I'm inclined to think that this argument has some force if we are operating within a purely consequentialist ethical theory. Suppose, however, that we operate within a deontological moral theory. We may suppose that there are moral principles forbidding certain acts of harm to others. Consider an agent who contemplates an evil deed of seriously harming some other human being. Our agent may reason that if he refrains God will have to permit some other evil just as severe in order to sustain significant morality. So he may conclude that his evil action won't have consequences that make the world any worse than it will be if he refrains. But, of course, on a deontological theory it won't follow that he has no duty to refrain. His duty to refrain does not depend merely on the consequences of his action. It depends on the fact that his proposed action is itself something that is at least prima facie wrong for him to do. Thus, it is intelligible that there would be moral principles prohibiting human actions that on occasion are permitted by God for the purpose of making such moral prohibitions a significant fact in human conduct. Given such a deontological moral theory, God could prevent all gratuitous evil without morality thereby being undermined. Toward the end of his paper Masker acknowledges that there may still be a problem with the existence of natural evils (like the fawn's suffering) that we have reason to believe are gratuitous. So, he endeavors to prove that the existence of gratuitous natural evils is fully consistent with the existence of an omnipotent, omniscient, perfectly good being. Why would such a being permit natural evils that he could prevent without thereby losing some outweighing good? Masker points out that there are goods that come about as a result of our having to deal with natural forces that bring about harm of various kinds. These threatening forces motivate us to develop knowledge of nature's workings, courage, cooperation, compassion for others, etc. Masker claims that if God were to prevent all such evils, and we knew it,

THE PROBLEM OF EVIL Ruminations about Evil / 83 we would not be motivated to develop these goods. Surely the motivation to acquire and/or respond in accordance with any or all of these goods would be greatly reduced, if not eliminated entirely, if we really believed that God would prevent any natural evils which were not essential to the realization of still greater goods, (p. 34) So, in order for humans to develop in these desirable ways, God must either massively deceive us about what is going on in nature (He makes it seem as though a good deal of gratuitous, natural evil occurs) or he must permit natural evils that serve no balancing or outweighing good at all. My answer strikes a now familiar refrain. If God must permit natural evils that serve no other goods in order to obtain the goods of our developing ourselves into courageous, cooperative, compassionate, knowledgeable beings, then those natural evils are not gratuitous after all. What Hasker needs to establish is that God would be justified in permitting natural evils that he could prevent without losing any outweighing goods, including the goods of human knowledge, courage, compassion, etc. Hasker fails to do this. From our earlier discussion, we can construct a Haskerian response to this criticism. Suppose a certain level of actual, otherwise gratuitous, natural evil must occur in order to provide rational motivation for humans to develop certain good qualities and dispositions in themselves. If so, God will then permit this level of otherwise gratuitous, natural evil in order to provide us with the requisite rational motivation to develop in these ways. But now consider some agent who is in a position to prevent some serious animal suffering that is about to be occasioned by natural forces. We will suppose our agent is in a position to rescue the fawn from the fire (occasioned by lightning) before it is severely burned. Our agent can reason that if she prevents this otherwise gratuitous, natural evil, God will have to permit the occurrence of some other equally severe, otherwise gratuitous, natural evil in order to maintain the level necessary to provide humans with rational motivation to develop in virtuous ways. Knowing this, our agent may well be deprived of rational grounds for her motivation to prevent this otherwise gratuitous, natural evil. Paradoxically, if God prevents all natural evils unless He cannot do so without losing some outweighing good'(including the good of providing rational motivation for humans to develop knowledge, courage, compassion in relation to harmful effects of natural forces),

133

134

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 84 / William L. Rowe and we know this to be so, then our motivation to prevent what appear to be gratuitous, natural evils would be diminished. Thus, we would be less likely to develop knowledge, courage, compassion, in relation to harmful effects of natural forces.21 Suppose there is a good demon with power and inclination to prevent any toothaches from occurring. The good demon, however, recognizes that it is good for human beings to develop and exercise their capacities to relieve their fellows of toothaches. The fostering of human motivation to achieve this good requires the demon to permit toothaches to occur every now and then. Moreover, achieving the good in question justifies the demon in permitting these toothaches. Potential toothaches not required for this end are prevented by the good demon. Now let's also suppose that, knowing this, I come upon you at the moment you are about to have a toothache. I might well reason that were I to struggle to find a way to prevent your toothache, the good demon will have to permit some other toothache to occur so that toothaches occur at the required rate to provide humans with the motivation to develop and exercise their capacities to relieve them. The result is that I am not rationally motivated to struggle to find a way to prevent your toothache. But it is precisely our efforts to find a way to prevent or relieve instances of toothache suffering that enable us to develop the knowledge we need to prevent or relieve them. Paradoxically, the policy of the good demon, once known by us, appears to render it somewhat unreasonable for us to develop the capacities for the sake of whose development and exercise the demon permits certain toothaches to occur. Even more paradoxically, it would seem that if the good demon is to accomplish the good of our developing and exercising our capacities to relieve others of toothaches, he must permit toothaches to occur that are not required for us to develop and exercise these capacities. But if they are not required for us to develop and exercise these capacities, why does he have to permit them? What is our good demon to do? His goodness inclines him to prevent any gratuitous toothaches. But it now seems that the only way he can achieve the end of humans developing the capacity to relieve others of toothaches is by allowing gratuitous toothaches. The good demon's problem is like God's problem, as Hasker sees it. God's goodness, I would think, inclines him to prevent hideous moral evils and natural evils that involve extraordinary human and/or animal suffering. But some such evils must be permitted to

THE PROBLEM OF EVIL Ruminations about Evil / 85 achieve certain goods: a world with significant morality, a world in which human beings are motivated to develop courage, compassion, knowledge of the workings of nature, etc. These evils, therefore, are not genuinely gratuitous. But it seems that if we all know what policy God and the good demon are following, the achievement of the very goods for the sake of which these evils are permitted is undermined, if not defeated. Let's suppose that the arguments for these conclusions are both rigorous and compelling. What are God and the good demon to do? Will they let gratuitous evils and gratuitous toothaches abound, so that (paradoxically) the various goods noted above can be achieved? Or is there a better way? Consider again the good demon. The problem contains an essential element that is epistemic in nature: the knowledge human beings have that the good demon exists and is operating according to the policy described above. It is that knowledge that undermines my motivation to struggle to find a way to prevent your toothache. If the good demon is powerful enough, he then has two ways out of his plight. He can allow genuinely gratuitous toothaches to abound or he can refrain from making his presence and policy so decisively known that human beings become unmotivated to develop the capacity to relieve people of their toothaches. Is there any real doubt as to how a good demon would solve the problem? If my son will be undergoing delicate and dangerous surgery in the morning and I have been advised that a sound night's rest is quite important to a successful outcome, what will I say if he asks me before sleep if the operation will be dangerous? Suppose I know that by telling him it is he will be so frightened that he will hardly sleep at all. Clearly, I need to assure him that all will be well. Similarly, for the good demon. Similarly, also for God. If God really is in the paradoxical corner that Hasker thinks He is in, then clearly the best course is not to make his presence and policy so decisively known that his very purposes for human life are undermined, if not defeated.22 Hasker thinks this engages God in a "massive disinformation campaign." God will need to allow it to appear that gratuitous evils occur, even though they in fact do not. For so long as human beings believe that gratuitous evils occur, neither morality nor our motivation to develop courage, compassion, etc., will be diminished or undermined. Surely, other things being equal, God will judge that our lack of decisive knowledge of His presence and policy is better

135

136

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 86 / William L. Rowe than letting horrendous, gratuitous evils abound in the world. As many theists believe, God will wait for the next life to provide us with a decisive knowledge of his nature and actions. This being so, Masker's thesis fails. In a universe with significant morality, human courage, compassion, and knowledge of the workings of nature, there is no need whatever for God to permit evils that He could prevent without loss of any balancing or outweighing good. Hasker might (but only for purpose of argument) be willing to concede all of the above. But he would then go on to argue that the concession is rather detrimental to the argument for atheism based on evil. For that argument presumed that if it is rational to believe that genuinely gratuitous evil exists then it is rational to believe that an omnipotent, omniscient perfectly good being does not exist. But this is no longer so. For if the reasoning above is correct such a being would have to make it (or allow it to) seem to us that gratuitous evils occur. Even though the being prevents all gratuitous evil, the realization of some of His good purposes for us (significant morality, knowledge of the workings of nature, compassion, etc.) depends on it being rational for us to believe that gratuitous evils exist. So we can't appeal to this point as our grounds for concluding that it is rational to believe that such a being does not exist. My answer to this is that it is not only likely that evils occur that an omnipotent, omniscient being could have prevented without loss of any balancing or greater good, it also is likely that such evils occur far in excess of what such a being would have to permit in order for us to be rational in believing that such evils occur. Clearly, if there is an omnipotent being, such a being could have prevented a good deal of evil in our world without in the least altering the fact that the amount and intensity of evil makes it rational for us to believe that evils occur in excess of what an omnipotent being would need to permit to achieve a balancing or greater good (including the goods of significant morality, knowledge, compassion, etc.) Who would say that if only five million had been permitted by omnipotence to perish in the holocaust it would not have been rational to believe that evils occur that omnipotence could have prevented without loss of any greater good? So, even if the reasoning above is accepted, there remains plenty of space for a strong argument for atheism to be constructed.

THE PROBLEM OF EVIL

Ruminations about Evil / 87 Notes 1. "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16 (1979): 335-341. 2. "Epistemic Probability and Evil," Archivio Difilosofia LVI (1988): 557-584. 3. An evil is pointless provided there is no good the obtaining of which would justify an omnipotent, omniscient being in permitting that evil. 4. The use of these expressions has led to some misunderstanding of the argument. For a good discussion of these problems see Stephen J. Wykstra, "The Humean Obstacle to Evidential Arguments From Suffering: On Avoiding the Evils of "Appearance," International Journal for the Philosophy of Religion 16 (1984): 73-93. Also see my reply, "Evil and the Theistic Hypothesis: A Response to Wykstra," International Journal for the Philosophy of Religion 16 (1984): 95-100. 5. For the details of this case and an excellent discussion of the problem of evil see Bruce Russell, "The Persistent Problem of Evil," Faith and Philosophy 6 (1989): 121-139. 6. For sake of simplicity, I shall ignore the possibility that such a being might permit El or E2 in order to prevent some equal or worse state from obtaining. 7. For a defense of P and the inference of Q from P see my "Evil and Theodicy," Philosophical Topics 16 (1988): 119-132. 8. Plantinga's lucid presentation of the argument for this claim may be found in God, Freedom, and Evil, (New York: Harper Torchbook, 1974). 9. This is what I am reading into Plantinga's discussion. He nowhere explicitly concedes this. 10. See his "The Humean Obstacle to Evidential Arguments From Suffering: On Avoiding the Evils of 'Appearance'." Also see his co-authored piece (with Bruce Russell), "The Inductive' Argument from Evil: A Dialogue," Philosophical Topics 16 (1988): 133-160. 11. I discuss Wykstra's first point here only in order not to have been thought to have intentionally refused to address it. Actually, his first point is not directed at the question of justifying the claim that a good lacks J. It is directed at the question of justifying the claim than an evil serves no justifying good. 12. "The "Inductive" Argument From Evil: A Dialogue," 143. 13. It is important to note that there are two ways in which the permission of an evil may be justified. First, it may be part of a good whole that has another part that is better than the whole. Second, it may be part of a good whole in which the other part is not better than the whole. In the first case, the evil is outweighed; in the second case, it is defeated. Here I only discuss the first way. For some remarks about the second way see 233-234 of my essay, "The Empirical Argument From Evil," Rationality, Religious Belief, and Moral Commitment, ed. by R. Audi and W. Wainwright, Cornell University Press, 1986. Also see Roderick Chisholm's analysis of the idea of evil being defeated in "The Defeat of Good and Evil," Proceedings and Addresses of the American Philosophical Association 42 (1968-69): 21-38.

137

138

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 88 / William L. Rowe 14. "The "Inductive" Argument From Evil: A Dialogue," 146-147. 15. What I say here isn't quite right. It would be sufficient to weaken the inference from P to Q to show that if there were an omnipotent, wholly good being who created our world, then the goods in virtue of which he permits El and E2 would be as likely to be undetectable by us as to be detectable by us. (I owe this correction to Stephen Wykstra.) 16. "The "Inductive" Argument From Evil: A Dialogue," 157-160. Also see my discussion of this point in "The Empirical Argument from Evil," 239-241. 17. "The Necessity of Gratuitous Evil," forthcoming in Faith and Philosophy. References to the pages of this manuscript will be included in the text. 18. The comments were presented at the meeting of the Central Division of the APA, April, 1988. 19. Quoted in Masker's paper, p. 22 20. I would argue that we have reason to believe that a good deal of evil could have been prevented by an omnipotent being without loss of any justifying good, including the good of sustaining morality. It is reasonable to believe, for example, that an omnipotent being could have prevented much of the horror of the holocaust without rendering senseless our prohibitions against harming others. Therefore, it is likely that genuinely gratuitous evil exists and, therefore, likely that an omnipotent, morally perfect being does not exist. 21. I've compressed this argument considerably. As it stands, I doubt that Masker would endorse it. It is intended only to suggest a similar kind of response to my objection to his argument concerning gratuitous moral evil. 22. A position quite like this is held by John Mick, the foremost theodicist of our time. See the revised edition of Evil and the God of Love, (New York: Harper and Row, 1978).

CHAPTER EIGHT

William Alston on the Problem of Evil

W

illiam Alston is in the forefront of prominent philosophers who support and defend the rationality of belief in the God of traditional theism: an omnipotent, omniscient, perfectly good being who is the creator of the world. His project includes positive attempts to show that belief in theism can be rationally grounded in ordinary religious experiences, as well as negative attempts to show that there are no good reasons for rejecting theism. In 1991, Alston published a major essay in defense of theism against the charge that the facts about evil in our world constitute a strong prima facie case for atheism.1 I propose here to evaluate the strengths and weaknesses of this essay. Noting that attempts to argue for the logical inconsistency of God and evil have largely been abandoned, Alston focuses on the inductive (probabilistic, evidential) argument from evil, adopting the following formulation of that argument: 1. There exist instances of intense suffering that an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. 2. An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse. 1 William P. Alston, "The Inductive Argument from Evil," Philosophical Perspectives, 5, Philosophy of Religion (Atascadero, Calif.: Ridgeview, 1991), pp. 29—67. Page references appear in parentheses in the text.

140

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 72 William Rowe

3. There does not exist an omnipotent, omniscient, wholly good being.2 Alston uses the term 'gratuitous suffering" for any case of intense suffering E that satisfies premise i or—to avoid some inadequacies in i—that an omnipotent, omniscient, wholly good being would have no sufficient reason to permit. Having thus characterized gratuitous suffering, Alston writes: "E's being gratuitous, then, is the contradictory of the possibility of God's having a sufficient reason to permit it, and equivalent to the impossibility of God's having a sufficient reason for permitting it" (p. 30; italics mine). Although the quoted remark seems to play no vital role in the argument he develops, to forestall misunderstanding it is important to note that if the modal term 'possible* is taken in either its logical or epistemic sense, the remark is false. For E to be gratuitous, all that is required is that there be no sufficient reason for God's permitting E. It is not required that there be no sufficient reason for God's permitting E in every possible world in which E occurs. For in some possible world, E occurs in circumstances such that had God prevented E a worse world would have obtained. The point is that some evil in the actual world may be such that God would have no sufficient reason for permitting it without that evil being necessarily such that God would have no sufficient reason for permitting it. And, as for the epistemic use of 'possible', it's clear that the question whether a certain evil is gratuitous has nothing to do with what any human being knows or is (isn't) justified in believing. To justify accepting premise i (that gratuitous evil exists), all one need do is justify the belief that some particular evil that occurs is gratuitous. Alston notes two particular examples of evils that I've held are reasonably believed to be gratuitous. In (1979) and subsequent papers Rowe focuses on the case of a fawn trapped in a forest fire and undergoing several days of terrible agony before dying (hereinafter 'Bambi'). In (1988) he adds to this a (real life) case introduced by Bruce Russell (1989), a case of the rape, beating, and murder by strangulation of a five-year old girl ('Sue') by her mother's boyfriend. Since I am specifically interested in criticizing Rowe's argument I will argue that we are not justified, and cannot be justified, in judging these evils to be gratuitous, (p. 32) 2 This is how I put the argument in "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16 (1979), 335—41.

THE PROBLEM OF EVIL 73 William Alston on the Problem of Evil

Alston accepts the task of establishing that no one is or can be justified in believing that the sufferings of Bambi and Sue are such that an omnipotent, omniscient being could have prevented them without thereby losing some greater good or permitting some evil equally bad or worse.3 Indeed, Alston undertakes to show we cannot be justified in asserting any instantiations of "E is such that an omnipotent, omniscient being could have prevented it without thereby losing some greater good or permitting some evil equally bad or worse (pp. 30-31)." It is, of course, one thing to argue that no one can be rationally justified in accepting premise i and another thing to argue that we are justified in accepting the denial of premise i. It seems clear that apart from appealing to a rationally justified belief in the existence of God and inferring the denial of i from that belief, it would be quite incredible to suppose that any human being is justified in accepting the denial of i.4 Alston seems to agree with this. "To sum up: I think that examining the interconnections of good and evil in the world by our natural powers cannot suffice to establish either i. or its negation. For particular cases of suffering we might conceivably be able to establish nongratuitousness in this way, but what I shall argue in this paper is that no one can justifiably assert gratuitousness for any case" (p. 31). Alston notes the growing body of literature devoted to answering the 3

Alston later shows that it is not quite right to imply (as I have) that God could have no other reason for permitting a certain evil except that preventing it would involve losing a greater good or permitting some equally bad or worse evil (pp. 33-35). The corrections needed in the formulation of premise i are not directly relevant to the central argument of his essay. So, for the purpose of evaluating Alston's argument, we can continue to think of God's reason for permitting an instance of intense suffering as the securing of a greater good or the prevention of some evil that is equally bad or worse. 4 In an endnote Alston says that "Rowe proceeds as if he supposed that the only alternatives are (a) its being reasonable to believe i and (b) its being reasonable to believe not-i" (p. 62). He then quotes a passage in which I claim that not-i is unlikely and conclude that "although we cannot prove that (i) is true, it is, nevertheless, altogether reasonable to believe that (i) is true, that (i) is a rational belief." I am a bit puzzled by the charge that I have overlooked the possibility that it may be rational to believe neither i nor its denial. For in a footnote to the very passage Alston quotes, I am careful to remind the reader that "it is one thing to argue that it is unreasonable to think that (i) is false and another thing to conclude that we are therefore justified in accepting (i) as true." I then point out that "there are propositions such that believing them is much more reasonable than disbelieving them, and yet are such that withholding judgment about them is more reasonable than believing them." After citing an example from Chisholm, I note that "it might be objected that while we've shown that believing (i) is more reasonable than disbelieving (i), we haven't shown that believing (i) is more reasonable than withholding belief" ("The Problem of Evil," p. 338). These do not strike me as the remarks of someone who supposes that the only alternatives are rationally believing i or rationally believing not-i. The lesson, I suppose, is that anything really important to the text should be in the text and not relegated to a footnote.

141

142

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 74 William Rowe

inductive argument from evil. Much of this literature focuses on the cognitive disparity between the human mind and an omniscient mind. Being omniscient, the theistic God would know many propositions we don't know. Presumably, some of these propositions would be about various goods and their interconnections with horrendous evils like Bambi and Sue.5 Starting from the disparity between God's knowledge and ours, recent critics of the problem of evil have argued that we cannot have a good reason to believe that Bambi and Sue serve no good that would justify an omnipotent, omniscient being in permitting their occurrence. For, so they claim, given all that God would know that we don't know, it is reasonable to believe that the goods, if such there be, that would justify God in permitting Bambi and Sue would be beyond our ken. While indicating some sympathy with these critics,6 Alston proposes to focus less on the cognitive distance between the human mind and the divine and more on the numerous considerations one would have to be justified in excluding as God's reason(s) for permitting Bambi and Sue, if one is to be justified in making the "universal negative" claim that there is no sufficient reason a divine being would have for permitting these evils. Starting from the fact that this universal negative claim excludes any sufficient divine reason for permitting Bambi and Sue, Alston argues that given our limited perspective, we cannot be justified in ruling out all such potential divine reasons and, therefore, cannot be justified in believing that an omnipotent, omniscient, morally perfect being would not permit Bambi and Sue. What potential divine reasons for permitting Bambi and Sue are we unjustified in ruling out? A major part of Alston's project is to consider familiar Christian theodicies and to explore their possible application to Bambi and Sue. These theodicies divide into those proposing divine reasons for suffering that are concerned with possible goods to those who endure the suffering and those proposing divine reasons for suffering concerned with possible goods not restricted to the sufferers. His examples of the first sort are punishment for sin, soul making, and having a vision of the inner life of God. His examples of the second sort are the value of free will, benefits to those who cause or witness the suffering, and the value generated by a lawlike, natural order. 5 For brevity, we can think of God's permitting Bambi and Sue as God's permitting the particular sufferings of Bambi and Sue that were earlier described. Sometimes I use 'Bambi' and 'Sue' to refer to the individuals Bambi and Sue. At other times, I use the names to refer to their particular sufferings. 6 In correspondence Alston has indicated that he doesn't agree with these critics in holding that "given all that God would know that we don't know, it is reasonable to believe that the goods, if such there be, that would justify God in permitting Bambi and Sue would be beyond our ken."

THE PROBLEM OF EVIL 75 William Alston on the Problem of Evil

The goods delineated in Christian theodicies are among the class of goods we know about. Within this class there also are goods that do not figure prominently in Christian theodicies. In addition, we may presume that there is a class of goods we do not know about, goods beyond our ken. In what follows, we look first at Alston's efforts to find a partial divine reason for Bambi and Sue among the goods we know about, particularly those goods that figure in certain Christian theodicies. Later we critically evaluate his claim that we cannot possibly be justified in ruling out potential divine reasons for Bambi and Sue that concern goods beyond our ken. Alston emphasizes that he is considering Christian theodicies only as live possibilities for divine reasons for permitting evil. He is not undertaking to show that any of these theodicies is correct (p. 36). Since he does not spell out what he means by a live possibility in this context, I shall take a live possibility here to be something that we have no good reason to believe would not wholly or partially justify an omniscient, omnipotent, wholly good being in permitting some evil.7 The question then becomes whether any or all of the theodicies are live possibilities for Bambi and Sue. If they are not live possibilities, if they are implausible suggestions as to what might justify an omnipotent, omniscient being in permitting Bambi and Sue, then, insofar as his use of Christian theodicies is concerned, Alston will have failed to make his case, that is, he will have failed to show that no one is justified in believing premise i of the inductive argument from evil. Do Christian theodicies provide live possibilities for divine reason to permit Bambi and Sue? Alston admits that theodicies that focus on some good for the sufferer do not seem of much help here. I am, naturally, most interested in suggestions that constitute live possibilities for divine reasons for permitting Bambi's and Sue's suffering. And many familiar theodicies do not pass this test.. .. Bambi's suffering, and presumably Sue's as well, could hardly be put down to punishment for sin, and neither case could seriously be supposed to be allowed by God for the sake of character building, (p. 37) Nor does Marilyn Adams's suggestion that "temporal suffering itself is a vision into the inner life of God" look promising for either Bambi or Sue. But Alston thinks it has a some chance of success. 7 In correspondence Alston has indicated that he would explain a live possibility in terms of what we have no sufficient reason to believe would not wholly or partially justify an omniscient, omnipotent, wholly good being in permitting some evil.

143

144

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 76 William Rowe

The only real chance for an exception is Adams's suggestion that the experience of suffering constitutes a vision of the inner life of God. Since this is not confined to those who identify it as such, it could apply to Sue, and perhaps to Bambi as well, though presumably only Sue would have a chance to recognize it and rejoice in it, retrospectively, in the light of the beatific vision. However, I don't want to insist on this exception. Let us say that a consideration of the theodicies thus far canvassed does nothing to show that we can't be justified in affirming an instantiation of i. for Bambi and Sue. (p. 44) Expressions such as "I don't want to insist on this exception" and "Let us say" sound a bit like concessions to the proponent of the argument from evil. It's as though, if all else were to fail, one could make one's stand on the possibility that the suffering of Bambi and Sue are permitted because they are visions of God. But since there are livelier possibilities down the road, we can afford not to insist on this possibility. Nevertheless, consider how utterly implausible this "exception" is. Can anyone seriously entertain the idea that Bambi's excruciating suffering on being burned in the forest fire constitutes a vision fry Bambi of the inner life of God? And even if one could bring oneself seriously to entertain this idea, could one believe that such a vision would justify God in permitting Bambi's excruciating suffering? After all, Bambi is acutely aware of the suffering but totally unaware that the suffering is a vision of God, as Alston is careful to note. If Alston is willing to countenance this theodical suggestion as a potential divine reason for Bambi's suffering, we can, I think, rightfully complain that he is not sticking to his intended plan of presenting us with real live possibilities for God's reasons for permitting Bambi and Sue. For this suggestion, whatever merit it may have when applied to the suffering of the Christian martyrs, is surely a dead option for a divine reason for permitting Bambi. One might overlook this particular deviation from Alston's avowed intention to present only live possibilities for divine reasons to permit Bambi and Sue if it were the only such lapse in his essay. But after agreeing that the theodicies of the first sort (those concerned with possible goods to those who endure the suffering) do "nothing to show that we can't be justified in affirming an instantiation of i. for Bambi or Sue," Alston argues that for all we know, the sufferings of Bambi and Sue are necessary, "in ways we cannot grasp, for some outweighing good of a sort with which we are familiar, e.g., supreme fulfillment of one's deepest nature" (p. 44). It is, I think, difficult to square this judgment with Alston's earlier rejection of the idea that the soul-making

THE PROBLEM OF EVIL 77 William Alston on the Problem of Evil

theodicy of John Hick provides a live possibility for a divine reason to permit Sue's suffering. For Alston notes that it "strains credulity" to suggest that Sue's moral and spiritual development in an afterlife would justify God in permitting Sue's suffering. I agree that Hick's suggestion when applied to Sue's suffering "strains credulity," and I also agree with Alston's judgment that the proponent of the argument from evil is justified in rejecting the soul-making theodicy as a live possibility for a divine reason to permit Sue's suffering. But isn't it just as much a strain on our credulity to suggest that the torture, rape, and strangulation of a five-year-old child (or God's permission of it) is necessary in order for her to attain the supreme fulfillment of her nature? (What, after all, is the supreme fulfillment of one's nature from the Christian perspective if not, as Hick's theodicy supposes, a life of eternal fellowship with God?) And what does Bambi's suffering and death have to do with the supreme fulfillment of Bambi's nature? Alston suggests that the suffering and death may be "necessary in ways we cannot understand" for the supreme fulfillment of Bambi's nature. And, for my part, I have to admit that I have no proof that this isn't so. But since Alston is here considering a good (supreme fulfillment of one's nature) with which we are familiar, I think we do know enough to be entitled to view with considerable skepticism the suggestion that Bambi's excruciating suffering and death are so related to the supreme fulfillment of Bambi's nature that even an omnipotent being could not enable Bambi to achieve the latter without permitting the former. At this point I might as well lay my cards on the table concerning the two examples I've given that strike me as falling short of Alston's goal of providing live possibilities for divine reasons for Bambi and Sue. A live possibility for a divine reason for Bambi and Sue is something that we have no good reason to believe would not justify an omniscient, omnipotent being in permitting Bambi and Sue. In rejecting the suggestions Alston makes of Bambi's suffering as a vision of God (which he doesn't insist on) or as necessary (in ways we don't or can't comprehend) for the supreme fulfillment of Bambi's nature, I am of course contending that they are not live possibilities for divine reasons to permit Bambi's excruciating suffering and death as a result of being severely burned in the forest fire. That is, I am contending that it is rational to believe that if the goods in question are obtainable at all, an omnipotent, omniscient being could obtain them without having to permit Bambi's excruciating suffering and death (or something equally bad or worse). Now we've already seen that Alston needs to establish that there are live possibilities for Bambi and Sue if he is to show that no one is justified in believing premise i of the inductive argument by

145

146

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 78 William Rowe

appealing to the cases of Bambi and Sue. And here I want to expose what I believe are two fallacious lines of argument that someone might employ in an effort to establish that either of the goods we've been discussing is a live possibility for a divine reason to permit Bambi and Sue. First, one may point out that p is logically possible—or that we don't know that it isn't logically possible—and conclude that p is, therefore, a live possibility. Second, one may note that we don't know with certainty that not-p—or that no one can prove that not-p—and conclude that p is, therefore, a live possibility. It is obvious that both lines of argument are fallacious. Many things we know to be false are, nevertheless, logically possible and known to be logically possible. But if p is known to be false, p is not a live possibility. And although one may not know that not-p or be able to prove that not-p, one may have good reasons to believe that not-p. So the fact that we don't know or can't prove not-p doesn't entitle us to conclude that p is a live possibility. It would be a mistake to think that Alston is explicitly or intentionally engaging in either form of fallacious inference. For he is at times quite explicit in rejecting as live possibilities considerations that he and I agree are logical possibilities that no one can claim to know for certain to be false. My point in mentioning these two fallacious inferences is both to note their seductive charm and to emphasize the necessity of moving beyond mere logical possibility and inability to disprove as grounds for taking various goods as potential live possibilities for Bambi and Sue. It is important to bear in mind that something more than this needs to be done if we are to be persuaded that something is a genuine live possibility for a divine reason to permit Bambi and Sue. So far as goods for the sufferer are concerned, what we've seen is that Alston believes that the prominent Christian theodiciesym/ to provide live possibilities for the cases of Bambi and Sue. But, as we've also seen, he thinks that he can establish that some goods we are familiar with (for example, the supreme fulfillment of one's nature) are such that we have no good reason to think that an omnipotent, omniscient being can obtain them without having to permit Bambi's and Sue's sufferings. And he also holds that it is in principle impossible for us to be justified in precluding patient-centered goods we do not know about as providing an omnipotent, omniscient being with a morally sufficient reason to permit those sufferings. As for my part, I've indicated that aside from logical possibility an inability to disprove Alston does little or nothing by way of "establishing" that we are not justified in precluding familiar goods like the supreme fulfillment of Bambi's nature or Bambi's having a vision of God from providing an omnipotent, omniscient being with a morally

THE PROBLEM OF EVIL 79 William Alston on the Problem of Evil

sufficient reason to permit Bambi's excruciating torment and death. As for goods beyond our ken, we treat this matter later, once we have considered Alston's discussion of theodical suggestions for divine reasons that extend beyond benefit to the sufferer. The other theodicies Alston explores concern goods that are not restricted to the sufferer. An initial problem with such theodicies is that they seem to suggest that it is permissible for God to allow a creature to undergo severe suffering so that He may bring a benefit to others—either other creatures or Himself. But this strikes many as unjust.8 Alston partially agrees with their judgment. He thinks it would be unjust for God to permit someone to suffer for the welfare of others if the suffering resulted in the person's life as a whole being less than good. But if the person's life as a whole is on balance good and the person so sees it, then, in Alston's judgment, it may be permissible for God to allow the person to suffer so as to bring a benefit to others. A perfectly good God would not wholly sacrifice the welfare of one of His intelligent creatures simply in order to achieve a good for others, or for Himself. This would be incompatible with His concern for the welfare of each of His creatures. Any plan that God would implement will include provision for each of us having a life that is, on balance, a good thing, and one in which the person reaches the point of being able to see that his life as a whole is a good for him... . Nevertheless, this is compatible with God having as part of his reason for permitting a given case of suffering that it contributes to results that extend beyond the sufferer. So long as the sufferer is amply taken care of, I can't see that this violates any demands of divine justice, compassion, or love. (p. 48) It's difficult to determine whether Alston intends to include Bambi in the position he here takes. In the passage quoted it first looks as though he does so intend. For he talks of intelligent creatures and God's concerns for the welfare of each of His creatures. But in the next sentence he talks only of persons and makes remarks that could scarcely apply to Bambi. In any case, if he does intend to be making a point about any plan that God would implement for all his intelligent creatures, including Bambi, then he must be supposing that there is an afterlife for Bambi as well as Sue. For it is evident that some of God's creatures do not have an earthly life that is as a whole good for them. 8

Alston quotes Eleonore Stump on this point and observes that many theists and atheists concur with her judgment.

147

148

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

80 William Rowe It also seems clear that Alston is opposing the dominant Christian conviction that there is an eternal hell as well as an eternal heaven. For on the doctrine of an eternal hell, it is exceedingly unlikely that its inhabitants will all come to see their lives as "good on balance" and "joyfully endorse" what they have endured.9 Alston begins his discussion of particular goods that extend beyond the sufferer with an examination of the free will theodicy. According to this theodicy, God is justified in permitting evil actions and their consequences because He has bestowed on some of His creatures genuine freedom in a range of actions, and it is a conceptual impossibility for God to crate a free agent with respect to some action and also determine the agent to choose (not choose) to perform the action. So, according to this theodicy, God permits certain horrors to occur because to prevent them would be to prevent a certain degree of freedom in His creatures. Against this theodicy, Alston notes that it has been argued (i) that God could have created free creatures who always choose to do what is right, and (2) that permitting free will with respect to certain actions at certain times isn't worth the horrendous evils (for example, Sue's suffering) that result from the use of that freedom. Concerning the first objection, Alston points out that if we Set aside middle knowledge, as he does in the essay, God would not be able to create beings with genuine freedom and "guarantee" that they will always choose to do what is right. And he further notes that even if we grant middle knowledge, Plantinga "has established the possibility that God could not actualize a world containing free creatures that always do the right thing" (p. 49). Before considering his response to the second objection, we should note two points. First, to conduct his case on the assumption that there is no middle knowledge considerably weakens Alston's argument. For there is no consensus on whether middle knowledge is possible. And if middle knowledge is possible, then we have reason to think that an omnipotent, omniscient being could have created a world with less evil, but as much good, as our world contains.10 Nor will it do to emphasize that Plantinga has established the possibility that even with middle 9

Alston avoids rejecting the Christian doctrine of an eternal hell by supposing that God's plan might simply ensure that each person has ample opportunity to choose freely so as to have a life that is on balance a good to them. But since, presumably, Bambi lacks free will, it would seem (i) that an afterlife for such sentient creatures must be postulated in order to ensure that their lives are on balance a good for them, and (2) that no such creatures can be assigned to an eternal hell 10 For a brief discussion of this point, see my essay, "Ruminations about Evil," in Tomberlin, Philosophical Perspectives, 5, Philosophy of Religion, pp. 74-76.

THE PROBLEM OF EVIL 81 William Alston on the Problem of Evil

knowledge God could not have created a world with free creatures that always do the right thing. For Alston needs to show that even with middle knowledge it remains a live possibility that God could not have created a world with free creatures who always do what is right. But all that Plantinga has shown is that it is a logical possibility that such would be the case. Here again, I'm afraid, we have a slide from what is a logical possibility to what is a live possibility, from what is broadly logically possible to what we have no good reason for thinking isn't so. Indeed, that such a slide has occurred is indicated by the sentence with which Alston begins his very next paragraph. "Thus we may take it to be a live possibility that the maintenance of creaturely free will is at least part of God's reason for permitting wrongdoing and its consequences" (p. 50; italics mine). Suppose we endeavor to apply the free will theodicy to the case of Sue's suffering on being beaten, raped, and strangled by her mother's boyfriend. The first question we need to ask is whether the possession of free will is something that is in itself of such great value as to merit God's permission of the horrendous moral evils in the world. I think the answer must be no. We should distinguish the intrinsic value of possessing free will from its extrinsic value. The mere possession of free will does not strike me as itself having much in the way of intrinsic value.11 But the possession of free will does seem necessary to attaining states that are of great intrinsic value. Thus, if we can agree that free will is necessary for the existence of things of great intrinsic value, we can agree that an omnipotent, omniscient, perfectly good being would likely endow His creatures (or some of them) with free will, providing that it does result to a sufficient degree in the things of great intrinsic value for which its possession is necessary. But, of course, our free will is limited; there are many things to which it does not extend. Moreover, it is sometimes right to curtail a particular exercise of free will when one foresees or predicts that its exercise is evil and or will result in considerable suffering. Since curtailing a particular exercise of free will does not significantly diminish a person's overall degree of freedom, the question at hand is whether it is rational to believe that an omnipotent, omniscient, perfectly good being would have prevented the particular exercise of free will (if that is what it was) the mother's boyfriend engaged in when he beat, raped, and strangled the five-yearold child. As Alston put the issue: 11

Thus, it seems to me that the free will theodicy needs to be included within something like Hick's soul-making theodicy, a theodicy that stresses some intrinsic goods for which free will is a necessary condition.

149

150

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 82 William Rowe

Presumably a tiny additional constriction such as would be involved in God's preventing Sue's attacker from committing that atrocity would not render things radically different, free-will-wise, from what they would have been without that. So God could have prevented this without losing the good emphasized by this theodicy. Hence we can be sure that this does not constitute a sufficient reason for His not preventing it. (p. 50) Alston thinks, however, that the preservation of Sue's attacker's free will is a live possibility for at least part of God's reason for permitting the suffering Sue undergoes at the hands of her attacker. For he reasons that the value of free will is such that God can intervene in only a small proportion of cases. Rowe's claim would then have to be that Sue's murder was so horrible that it would qualify for the class of exceptions. But that is precisely where the critic's claims far outrun his justification. How can we tell that Sue falls within the most damaging n% of what would be cases of human wrongdoing apart from divine intervention. To be in a position to make such a judgment we would have to survey the full range of such cases and make reliable assessments of the deleterious consequences of each. Both tasks are far beyond our powers, (p. 51) Alston's point seems to be this. God cannot intervene in all cases of the use of free will in doing evil that results in suffering, for this would severely limit human capacity to choose between good and evil. God, then, would intervene only in a certain percentage (n) of such cases. The cases in which God intervenes would be selected so as to minimize human suffering overall or to maximize human welfare. And we simply don't know enough about what would be cases of human wickedness (past, present, and future) apart from divine intervention to determine whether or not Sue's case would be included among the percentage of cases God would select to eliminate. Hence, by the nature of the case, we are simply not in a position to make a warranted judgment that Sue's case is among the n% worst cases of wrongdoing in the history of the universe. No doubt, it strikes us as incomparably horrible on hearing about it, but so would innumerable others. Therefore, the critic is not in a position to set aside the value of free will as at least part of God's reason for permitting Sue's murder, (p. 51)

THE PROBLEM OF EVIL 83 William Alston on the Problem of Evil

Does this argument establish that we aren't justified in excluding the value of human free will as part of God's reason for permitting Sue's suffering? Well, if the basic premise of Alston's argument is correct, the argument does establish just that. For to be so justified, Alston requires that we compare the episode involving Sue with the whole range of cases of human wrongdoing in the universe (past, present, and future) that would occur apart from divine intervention in order to determine whether Sue's case is sufficiently bad to warrant God in setting aside the value of free will in that case. And of course, no human being knows enough to engage in such a comparison. But I do not think this knowledge is required. The free will theodicy is built around the idea that the possession of freedom to do good and evil is a good in each person that God creates. What is important, therefore, is that each person have some measure of freedom to do good and evil. Now, as Alston will readily admit, (i) we don't possess unlimited freedom, and (2) it is sometimes right to curtail someone's freedom in order to prevent some horrendous evil act that results in considerable suffering to an innocent person. But if God were to select some person and effectively prevent that person from ever choosing to do an evil deed that results in suffering to an innocent person, we can agree that this might severely diminish the value of that person's freedom to do good and evil. So God has a reason to permit each person effectively to engage in doing good and evil acts. Of course, God would be able to intervene in some of those acts without significantly diminishing the person's general freedom to do good and evil. What, then, do we have to make a judgment about in order to determine whether the prevention of Sue's attacker's freedom to beat, rape, and strangle Sue would have diminished entirely the value of that person's freedom to do good and evil? Do we have to compare it with all the cases of human wrongdoing, past, present, and future, as Alston claims? Clearly not. What is at issue here is not some sort of amount of freedom to do good and evil in the entire universe, past, present, and future. What is at issue is the degree and value of Sue's attacker's freedom to do good and evil. Would this particular intervention severely diminish the value of that individual's overall degree of freedom to do good and evil? And if it would, is the value of that individual's overall degree of freedom to do good and evil worth the price of permitting the act in question? Perhaps some rational judgments are required about these questions in order for us to be justified in taking Sue's suffering as an instance of gratuitous suffering. But this is a far cry from having to do a survey

151

152

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

84 William Rowe of all human acts of wickedness in the universe, past, present, and future.12 What strikes me as the most promising suggestion of a live possibility for a divine reason to permit natural evils like Bambi is the last theodicy Alston considers: "the appeal to the benefits of a lawlike natural order, and the claim that suffering will be an inevitable byproduct of any such order" (p. 52). Clearly, if human beings are to make meaningful free choices concerning good and evil, they must be able to anticipate the consequences of their choices, and this is possible only if things happen (in relation to their choices) in a lawlike manner. Instances of suffering like Bambi are the inevitable by-product of the system of laws that enables us to anticipate the consequences of our choices. The standard objections to this theodicy are two. First, God may intervene to prevent various cases of natural evils without significantly weakening our confidence in what will happen as a result of our choices. Second, it has seemed to some critics that God could have instituted a set of laws of nature that would result in far less disastrous natural evils than actually occur in our world.13 Alston develops formidable replies to each of these objections. I consider only the first standard objection and his response to it. Alston agrees that a natural order can be regular enough to enable us to anticipate consequences of our choices, even if there are exceptions to the regularities. So long as God doesn't do this too often, we will still have ample basis for suppositions as to what we can reasonably expect to follow what. But note that by the same line of reasoning God cannot do this too often, or the desired predictability will not be forthcoming. Hence, though any one naturally caused suffering could have been miraculously prevented, God certainly has a strong prima facie reason in each case to refrain from doing this; for if He didn't He would have no reason for letting nature usually take its course. And so He has a possible reason for allowing nature to take its course in the Bambi case, a reason that would have to be overridden by stronger contrary considerations, (p. 53) 12 One might argue that were God to intervene in Sue's case, He would have to intervene in every similar case, with the result that no human being would be free to do evil acts of the kind represented by Sue. I respond to an argument of this sort below in connection with the theodicy that rests on the need for the world to operate in accordance with relatively stable laws of nature. 13 See, for example, Quentin Smith, "An Atheogical Argument from Evil Natural Laws," International Journal for Philosophy of Religion 29 (1991), 154-74.

THE PROBLEM OF EVIL 85 William Alston on the Problem of Evil

Alston concludes that to be justified in believing God has no sufficient reason for not intervening to prevent Bambi, we would have to be justified in supposing God has a sufficient reason to make Bambi an exception to His general policy of letting nature take its course. And how could we be justified in supposing that? We would need an adequate grasp of the full range of cases from which God would have to choose whatever exceptions He is going to make, if any, to the general policy of letting nature take its course. Without that we would not be in a position to judge that Bambi is among the n% of the cases most worthy of being miraculously prevented. And it is abundantly clear that we have and can have no such grasp of this territory as a whole. We are quite unable, by our natural powers, of determining just what cases, or even what kinds of cases, of suffering there would be throughout the history of the universe if nature took its course. We just don't know enough about the constituents of the universe even at present, much less throughout the past and future, to make any such catalogue. . . . Hence we are in no position to judge that God does not have sufficient reason for refraining from interfering in the Bambi case. (p. 53) Alston claims that in order to be justified in believing God would have no sufficient reason for permitting Bambi, we would first have to compare Bambi with the entire range of cases of suffering there would be in the universe, past, present, and future, if nature took its course. For since only n percent exceptions could be made to the general policy of letting nature take its course, we would need to make such a comparison in order to be justified in judging that Bambi falls within the percentage of allowable exceptions. Since we cannot make any such comparison, we cannot be justified in believing that God would not have a sufficient reason to refrain from interfering in the Bambi case. Again, I don't think Alston is right in what he insists we must know if we are to make such a judgment in the Bambi case. In the first place, he gives a poor argument for the conclusion that God has a strong prima facie reason in each case to let natural regularities obtain. The only reason given in the theodicy under examination for letting natural regularities obtain is that human beings need to be able to anticipate the consequences of their choices. Consider, then, far-off regions of the universe where events would take millions of years to register any impact upon human beings. So far as this theodicy is concerned, God may have no reason at all to impose regularities on those regions. Moreover, if He were to impose them, He would have no reason at all (so

153

154

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

86 William Rowe far as this theodicy is concerned) to interfere in those regions in only a small percentage of cases. For all Alston has said, God might be interfering quite frequently in those regions. And, coming closer to home, it is clear that the cases in which God's interference must be measured and kept to a certain percentage distributed over time are not global in the way Alston suggests. They are in fact limited to cases in which divine intervention could interfere with our human ability to anticipate the consequences of our choices. Thus, if we were supposing that God would change bullets into butter on certain occasions when they are fired at innocent persons, we might well run afoul of the theodicy we are examining. But the Bambi case could easily be resolved without any effect at all on our ability to anticipate the consequences of our choices. God might have arranged for the lightning to strike harmlessly some high ground, rather than a tree; thus forestalling the forest fire. Or, lightning having struck the tree, God might have prevented the tree from bursting into flames. Or, the tree having burst into flames, God might have put out the flames by causing a quick rainfall, and so on. An important feature of the Bambi case is that the sorts of interference required to prevent Bambi's excruciating pain and slow death over several days are those that likely would have no influence at all on our ability to anticipate the consequences of our choices. Alston's mistake here is to neglect to distinguish cases where interference with the usual course of nature would weaken our ability to anticipate the consequences of our choices from cases where interference with the usual course of nature would quite likely not weaken our ability to anticipate the consequences of our choices. His argument does have some validity for cases of the first sort: occasionally turning bullets into butter when they are fired at an innocent person, bringing some wellknown person back to life after it is established that the person has been dead for four days, and the like. It is with respect to cases of this sort that God might have a strong prima facie reason not to interfere. But no such prima facie reason has been established for cases of the latter sort, cases that include Bambi. It might be argued that if God had interfered in Bambi, then He would have had to interfere in every case where, unnoticed by humans, some sentient animal was about to be injured and subjected to excruciating pain while slowly dying. And, so the argument might go, if this were so, it is rather likely that human beings eventually would become aware of this interference and cease to be morally concerned for the supposed suffering of animals due to natural forces. The answer to this is that there would be a threshold beyond which further massive interference by God would perhaps have such an untoward effect. But

THE PROBLEM OF EVIL 87 William Alston on the Problem of Evil

it seems eminently clear that we are far away from any such supposed threshold. I conclude, therefore, that Alston's appeal to the theodicy based on the human need for a natural order so as to be able to anticipate the consequences of human choices does not succeed in establishing that we cannot be justified in thinking that an omnipotent, omniscient being would have no morally sufficient reason for permitting Bambi. I've now completed my examination of Alston's efforts to identify among the goods we know a live possibility for a divine reason to permit Bambi and Sue. As we've seen, his strategy was to search for such a reason among various Christian theodicies, both those that emphasize a good primarily to the sufferer and those that do not. My judgment has been that his efforts in both sorts of theodicies have failed. Oddly enough, this judgment is only slightly different from Alston's own conclusion. Alston allows that none of the reasons offered by the sufferercentered theodicies "could be any part of God's reasons for permitting the Bambi and Sue cases" (p. 58). (We should remember here, however, that he concedes this only "for the sake of argument".) He does claim that reasons offered by the non-sufferer-centered theodicies could be a part, but only a part, of God's reasons for allowing any case of suffering (p. 58). So Alston agrees with me that the reasons presented in the theodicies he has examined do not constitute a live possibility for a sufficient reason for God to permit Bambi and Sue.14 Our only main difference thus far is over the question of whether the reasons presented in the non-sufferer-centered theodicies could constitute a part of God's sufficient reason for permitting Bambi and Sue. Alston sums up his situation with respect to finding a reason for Bambi and Sue among Christian theodicies. This left me without any specific suggestions as to what might be a fully sufficient reason for God to permit those cases. And hence showing that no one can be justified in supposing that reasons of the sort considered are not at least part of God's reason for one or another case of suffering does not suffice to show that no one can be justified in supposing that God could have no sufficient reason for permitting the Bambi and Sue cases. And hence it does not suffice to show that Rowe cannot be justified in asserting i. (p. 58) How then does Alston propose to make up the difference, to establish that we cannot be justified in believing that there is no sufficient 14

1 should emphasize here another area of agreement with Alston. Although I have rejected Christian theodicies as live possibilities for Bambi and Sue, I agree with him that these theodicies may well provide live possibilities for other cases of suffering that

155

156

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

88 William Rowe divine reason for permitting Bambi and Sue? His reply is to claim that we are not justified in ruling out goods we are not aware of as constituting a sufficient divine reason for Bambi and Sue. Even if we were fully entitled to dismiss all the alleged reasons for permitting suffering that have been suggested, we would still have to consider whether there are further possibilities that are undreamt of in our theodicies.... Since it is in principle impossible for us to be justified in supposing that God does not have sufficient reasons for permitting E that are unknown to us, and perhaps unknowable by us, no one can be justified in holding that God could have no reasons for permitting the Bambi and Sue cases, or any other particular cases of suffering, (pp. 58-59) Indeed, Alston claims that even if everything else he has argued were in vain and his opponent could rule out all the theodical suggestions he has put forward, "she would still face the insurmountable task of showing herself to be justified in supposing that there are no further possibilities for sufficient divine reasons. That point by itself would be decisive" (p. 59; italics mine). Before turning to consider Alston's appeal to goods beyond our ken as the real source for God's sufficient reason for Bambi and Sue, I want to enter a slight protest in the "who has to show what" category. The critic claims to be justified in believing that there is no sufficient divine reason to permit Bambi and Sue. Alston explicitly accepts the task of establishing that the critic is not justified in believing that there is no sufficient divine reason to permit Bambi and Sue. Fair enough. But in the remark quoted just above, we are told that the critic is now faced with the task of proving (showing) that she is justified in believing that there are no sufficient divine reasons for Bambi and Sue. As things started out, Alston was going to prove that the critic is not and cannot be justified in believing that there are no sufficient divine reasons for Bambi and Sue. Now we are told that the burden is on the critic to prove that she is justified in so believing. Isn't it enough for it to be true that the critic is justified in this belief? Why is the burden now placed on the critic to provide justification for being justified in believing that there are no sufficient divine reasons for Bambi and Sue? Surely Alston doesn't think that in order to be justified in believing p, occur in our world. Much of what Alston argues on this score strikes me as important and correct.

THE PROBLEM OF EVIL

89 William Alston on the Problem of Evil

one must prove that she is justified in believing p.15 So I think we can safely put to one side the point that the critic must show that she is justified in believing that Bambi and Sue are gratuitous. Instead, let us concentrate on Alston's argument to establish that the critic is not justified in believing that Bambi and Sue are gratuitous because she cannot be justified in ruling out the securing of certain goods beyond our ken as constituting God's sufficient reason for permitting Bambi and Sue. Alston gives both a positive and a negative argument on this matter. The positive argument is intended to establish that our present knowledge about values is insufficient to merit any judgment about what future goods might justify God in permitting Bambi and Sue. The negative argument is intended to refute an argument I have given for the conclusion that it is likely that no good state of affairs is such that an omnipotent, omniscient being's obtaining it would morally justify that being in permitting Bambi or Sue. Let's look first at Alston's attack on this argument. I have endeavored to support the proposition (Q) No good state of affairs is such that an omnipotent, omniscient •being's obtaining it would morally justify that being in permitting Bambi and Sue. by giving as a reason for it the proposition (P)

No good state of affairs we know of is such that an omnipotent, omniscient being's obtaining it would morally justify that being's permitting Bambi or Sue.16

Of course, as we've seen, Alston does not think I am justified in holding (P) to be true. But his point here is that even if we are fully justified in accepting (P), we are not justified in accepting (Q) on the basis of (P). In support of the inference, I have said that it is simply one more instance of our common practice of making inferences from the known to the unknown, that if we've observed many As and found them all to be B's, we have some reason to believe that the As we haven't observed will likely be B's as well. I went on to note that this reason is subject to defeat by having some reason to think that were some As not B's, they would likely not be found among the As we encounter or 15 For proof of this, see his elegant essay, "Level Confusions in Epistemology," in his Epistemic Justification: Essays in the Theory of Knowledge (Ithaca: Cornell University Press, ^SgX PP- 164-65. 16 See "Evil and Theodicy," Philosophical Topics 12 (1988), 119-32.

157

158

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 90 William Rowe

know about.17 But barring having some such defeater, I claimed that (P) gives us a reason to believe (Q). Alston raises an important objection to taking (P) as a reason in support of (Q). His point is that the epistemic worth of any such inference depends on our having some background knowledge that the characteristic in question (being such that obtaining it would not justify an omnipotent, omniscient being in permitting Bambi or Sue) "can reasonably be expected to be present in all or most goods if it is found in a considerable sample" (p. 46). Sometimes we do have such background knowledge. We know, for example, that being a good watchdog is a characteristic that can be expected to be present in all or most dogs of a given breed if it is found in a considerable sample of that breed. Sometimes we have the opposite background knowledge. If we observe a considerable sample of a given breed of dog and note that each dog in the sample has the characteristic of being female, we know that this characteristic can reasonably be expected not to be found in all or a very large percentage of the breed. The inference of (Q) from (P) is, Alston contends, like neither of these. We simply have no background knowledge to go on here. So he concludes that we cannot be justified in holding (Q) by virtue of inferring it from our justified belief in (P). This is, as I've said, an important objection to the argument I've given in support of the inference from (P) to (Q). Although I continue to hold that (P) is a good reason for (Q) both in the sense of (a) making (Q) more likely than it otherwise would be and in the sense of (b) making (Q) more probable than not, I now recognize that the argument I gave, and which Alston criticizes, is sufficient to establish only that (P) is a good reason for (Q) in the sense of (a) making (Q) more likely than it would otherwise be. The fact that one has observed many As and have found all of them to be B's does make the proposition that all As are B's more likely than it would otherwise be. (Alston would not deny this.) But one proposition may make another proposition more likely than it would otherwise be without making it more probable than not. And it is the latter, and not just the former, that must be true if one is to be justified in believing (Q) by virtue of inferring it from (P). So the argument I have used to support the inference from (P) to (Q) is inadequate for something like the reason Alston gives. To shore it up we would need some reason to believe that the goods we know of (the As we've observed) are representative of the goods there are (the As there are). Noting the variety of goods we know of would be relevant to the task of providing such a reason. Having a good argument to 17

See ibid., p. 124.

THE PROBLEM OF EVIL 91 William Alston on the Problem of Evil

think that most goods are known to us would also be relevant.18 But quite apart from trying to shore up the argument I gave in support of the inference from (P) to (Q), I believe a different, more effective argument can be given to show that (P) is a good reason for (Q) in sense (b), as well as sense (a). But I don't have space here to spell out that argument.19 Alston's positive argument for why we aren't justified in believing that no good beyond our ken is such that obtaining it would morally justify an omnipotent, omniscient being in permitting Bambi or Sue rests on the idea that our knowledge of values and the conditions of their realization is analogous to our knowledge of physical science. The development of physical science has made us aware of a myriad of things hitherto undreamed of, and developed the concepts with which to grasp them—gravitation, electricity, electro-magnetic fields, space-time curvature, irrational numbers, and so on. It is an irresistible induction from this that we have not reached the final term of this process, and that more realities, aspects, properties, structures remain to be discerned and conceptualized. And why should values, and the conditions of their realization, be any exception to this generalization? A history of the apprehension of values could undoubtedly be written, parallel to the history just adumbrated, though the archeology would be a more difficult and delicate task. (p. 45) Later in his essay, Alston again stresses what he takes to be an analogy between scientific knowledge and our knowledge of values. Why should we suppose that the theodicies thus far excogitated, however brilliant and learned their authors, exhaust the field. The points made in the earlier discussion about the impossibility of anticipating future developments in human thought can be applied here. Just as we can never repose confidence in any alleged limits of future human theoretical and conceptual developments in science, so it is here, even more so if possible. It is surely reasonable to suppose that God, if such there be, has more tricks up His sleeve than we can envisage. (PP- 58-59) 18 Michael Tooley has advanced such an argument in "The Argument from Evil," in Tomberlin, Philosophical Perspectives, 5, Philosophy of Religion, pp. 114-15. 19 1 present this argument in "The Evidential Argument from Evil: A Second Look," forthcoming in Daniel Howard-Snyder, ed., The Evidential Argument from Evil (Bloomington: Indiana University Press, 1996).

159

160

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION

92 William Rowe What are we to make of this argument? Is there a history of values to be written that would parallel the history of physical science? I really don't think so. The reason is that the development of physical science concerns empirical matters the determination of which depends greatly on technological advances. In the area of values, however, the question of whether a given state of affairs is intrinsically good is not an empirical matter at all. Indeed, that something is intrinsically good, evil, or neutral is a matter of necessity and to be discerned by the comprehending intellect. And so it is that many of the things that are intrinsically good were as clearly known by the ancients as by us. Of course, the knowledge of how to bring it about that certain intrinsic goods obtain is akin to scientific knowledge. But, strictly speaking, this is not a matter of knowledge of values at all. So as long as we are discussing intrinsic values and our knowledge of them, Alston's claim that the development of our knowledge of values parallels the development of our knowledge of physical science is implausible. Of course, we do have to allow that there are possible states of affairs that have great intrinsic value and are unknown to us. But, as Michael Tooley has pointed out, basic moral truths that we do know place limits on what intrinsic values could justify a perfectly good being to permit innocents to suffer. For example, according to Tooley, our moral truths about justice imply "that an omnipotent and omniscient person would be justified in allowing an innocent individual to suffer only if the individual in question benefitted from that suffering."20 So, no matter how grand and glorious an unknown good might be, we know that God would not be justified in permitting Sue's suffering solely in order to obtain that good, if Sue were not to share in that good (or have the freedom to share in it). Moreover, when we take fully into account what an omnipotent, omniscient being would be able to bring into existence without permitting Bambi and Sue, it becomes doubtful, I believe, that intrinsic goods we do not know are any more likely to justify such a being in permitting Bambi or Sue than are the goods within our ken. Thus, the mere fact that God would know a lot more than we know doesn't seem a sufficient basis for holding, as Alston does, that it is in principle impossible for any human being to be justified in believing that there are any gratuitous evils. Although I do not think Alston has succeeded in his principal aim of establishing that the inductive argument from evil is a failure, I 20

Tooley, "Argument from Evil," p. 127. We should note that Tooley thinks this moral truth need not apply in the case of a being that is not omnipotent and omniscient (see pp. 112-13). Even with this emendation, some would object to his claim. Perhaps it may be strengthened by restricting it to cases of involuntary suffering by innocent persons.

THE PROBLEM OF EVIL

93 William Alston on the Problem of Evil

cannot deny that he has raised significant doubts about the argument that I have not laid to rest.21 I believe that the points he has brought up in this important, thoughtful essay will be discussed for some time to come.22 21 In particular, as I noted earlier, he has shown that the reason I gave in support of the inference from (P) to (Q) is inadequate. 22 1 am grateful to Martin Curd, Paul Draper, William Hasker, David O'Connor, Daniel Howard-Snyder, Jeff Jordan, Bruce Russell, and Eleonore Stump for helpful comments on an earlier draft of this essay. I am also indebted to William Alston for his incisive comments on the penultimate draft.

161

This page intentionally left blank

CHAPTER NINE

The Evidential Argument from Evil: A Second Look

It is as misleading to speak of the evidential argument from evil as it is to speak of the cosmological argument. Just as there are distinct arguments that qualify as cosmological arguments, there are distinct arguments that qualify as evidential arguments from evil.1 My purpose here is to look again at an evidential argument from evil that I first presented in 1979.2 Since that time I have made several changes in that argument in an effort to make it clearer and to patch up weaknesses in earlier statements of it. Starting with the latest published account of the argument, I will discuss some important criticisms of it and will continue my efforts to clarify, simplify, and strengthen the argument. The latest formulation I have given of the evidential problem of evil goes something like this.3 (El is the case of a fawn trapped in a forest fire and undergoing several days of terrible agony before dying. E2 is the case of the rape, beating, and murder by strangulation of a five-year-old girl.)4

164

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

263

P: No good we know of justifies an omnipotent, omniscient, perfectly good being in permitting El and E2; therefore, Q: no good at all justifies an omnipotent, omniscient, perfectly good being in permitting El and E2; therefore, not-G: there is no omnipotent, omniscient, perfectly good being. The first inference, from P to Q, is, of course, an inductive inference. My claim was that P makes Q probable. The second inference, from Q to not-G, is deductive. Against this argument from evil a variety of criticisms are possible. One might claim (1) that none of us is in a position to be justified in believing P. One might claim (2) that the inference from P to Q is not a good inductive inference, that P does not make Q more probable than not. So, one cannot be justified in • believing Q on the basis of P. One might claim (3) that even though the inference from P to Q is a good inductive inference, we have reasons (defeaters) on the basis of which it is rational to refrain from accepting Q on the basis of P. Finally, one might claim (4) that not-G does not deductively follow from Q because two possibilities are not excluded. First, (4a) it could be that the prevention of some worse evils is what justifies God in permitting El and E2. 5 Second, (4b) it could be that in a world with free, morally responsible creatures God needs to permit the occurrence of unjustified evil (gratuitous evil).6 Criticisms that have been advanced against this version of the evidential argument from evil have focused mainly on (1) and (2). Although several important papers have developed such criticisms, special mention should be made of William Alston 's essay "The Inductive Problem of Evil and the Human Cognitive Condition" (chapter 6 in this volume) and several papers by Stephen Wykstra.7 Obviously, the whole issue of whether what we know about evil in our world makes it likely that the theistic God does not exist is quite complex and cannot be satisfactorily addressed in any single essay. But I will undertake here to answer some serious objections concerning the inference from P to Q (raised by Alston, Wykstra, and others) and some serious objections concerning P itself (mostly raised by Alston). After discussing two preliminary matters, I will take up the objections to the inference from P to Q, later turning to some issues concerning P itself.

// Initially, we need to do two things. First, we need to specify just what P and Q assert. Although I have endeavored to do this in earlier writings on this topic, I don't think I have been as clear as is necessary. So I want to begin with this task, departing somewhat from earlier formulations. Second, we need to settle on our background information, connecting it to the question of God's existence in a way that will make our discussion relevant to the dispute between theists and nontheists over the problem of evil. As we've already noted, at least initially our discussion revolves around

THE PROBLEM OF EVIL 264

/

W I L L I A M L. R O W E

two particular evils, El (Bambi) and E2 (Sue). What is it that P affirms about El and E2? P says no good we know of justifies an omnipotent, omniscient, perfectly good being in permitting El and E2. What then does P entail? P entails that among the good states of affairs that we know of (however dimly or through a glass darkly) none is such that it justifies an omnipotent, omniscient, perfectly good being in permitting El and E2. So long as we keep in mind the features of the being in question, we can abbreviate our formulation of P as follows: P: No good we know of justifies God in permitting El and E2.

Since we are talking about a good that justifies God in permitting El and E2, we should allow, if not expect, that the good in question would be a conjunctive good. Perhaps there is a good we know of that justifies God in permitting El. Perhaps there is some other good we know of that justifies God in permitting E2. If so, then we will allow that it is true that some good we know of (a conjunction of the goods in question) justifies God in permitting E1 and E2. It should be obvious that I am trying to pose a serious difficulty for the theist by picking a difficult case of natural evil, El (Bambi), and a difficult case of moral evil, E2 (Sue). Should no good we know of justify God in permitting either of these two evils, P is true. What counts as a "good we know of"? I do not mean to limit us to goods that we know to have occurred. Nor do I mean to limit us to those goods and goods that we know will occur in the future. I mean to include goods that we have some grasp of, even though we have no knowledge at all that they have occurred or ever will occur. For example, consider the good of Sue's experiencing complete felicity in the everlasting presence of God. Theists consider this an enormous personal good. 1 have no doubt that it is. So, even though we don't have a very clear grasp of what this great good involves, and even though we don't know that such a good state of affairs will ever obtain, we do mean to include the good of Sue's experiencing complete felicity in the everlasting presence of God as among the goods we know of. Of course, if the good in question never does occur, then it is not a good that justifies God in permitting El or E2. So if some good state of affairs we know of does justify God in permitting El or E2, that good state of affairs must become actual at some point in the future, if it is not already actual. Under what conditions would P be true? P says that there is no good we know of that justifies God in permitting El and E2. One condition that would render P true is the nonoccurrence of the known good (supposing there is just one) whose occurrence would justify God in permitting El or E2. Suppose that among all known goods only Sue's experiencing eternal felicity in the presence of God is such that its occurrence would justify God in permitting E2 (Sue's suffering on being brutally beaten, raped, and strangled). If this good never occurs, P is true. As I indicated earlier, a good state of affairs justifies God in permitting some actual evil only if that good state of affairs occurs. Second, we should note that the nonexistence of God is also a sufficient condition of the truth of P. For the realization of a known good justifies God in permitting El or E2 only if God exists. To see this, consider the negation of P. The negation of P asserts that God exists and that some

165

166

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

265

good known to us justifies him in permitting E1 and E2. Since the negation of P is false if God does not exist, P will be true if God does not exist.8 Having spent some time clarifying P, we can be brief with Q. Q: No good at all justifies an omnipotent, omniscient, perfectly good being in permitting E1 and E2. So long as we keep in mind the features of the being in question, we can abbreviate our formulation as follows: Q: No good justifies God in permitting El and E2. As with P, we should note that if God does not exist, Q is true. For given that El and E2 exist, if God does not exist then it is not the case that there is an omnipotent, omniscient, perfectly good being who is justified in permitting El and E2 by virtue of realizing some good. Another way of seeing this point is to note that the negation of Q asserts that God exists and that there is a good that justifies God in permitting E1 and E2. Since the negation of Q is false if God does not exist, Q is true if God does not exist. I turn now to the background information k on which we will rely in forming judgments about how likely P, Q, and G (God exists) are. What will k include? I take it as important here that k be restricted almost entirely to information that is shared by most theists and nontheists who have given some thought to the issues raised by the problem of evil.9 To this end, we will want to include in k our common knowledge of the occurrence of various evils in our world, including El and E2, as well as our knowledge that the world contains a good deal of evil, k will also include our common understanding of the way the world works, the sorts of things we know to exist in the world, along with our knowledge of many of the goods that occur and many of the goods that do not occur. Of course, k will not include the information that God exists or the information that God does not exist. If we conceive of k in the way just suggested, what assignment should be given to the probability that God exists, given k, Pr(G/k)? Many nontheists hold that the enormous amount of evil in our world, particularly instances of horrendous human or animal suffering such as El and E2, make the existence of the theistic God unlikely. Many theists and some nontheists, however, will disagree with this assessment. On the other hand, many theists will argue that the mere existence of a world (or the order in the world) makes the existence of God likely. But some theists and many nontheists will disagree with this assessment. In order not to beg any of these questions, I will assign a probability of 0.5 to Pr(G/k), and, of course, 0.5 to Pr(~G/k). We will say that k by itself makes neither God's existence nor his nonexistence more likely than not. This need not be understood as denying what some nontheists hold concerning the possible negative evidential impact of the existence and multitude of horrendous evils in the world. Nor need it be understood as denying what some theists hold about the possible positive evidential impact of the existence of a world exhibiting order, etc. What it does indicate is that these different aspects of k—if they do impact positively or negatively on the

THE PROBLEM OF EVIL 266

/

W I L L I A M L. R O W E

likelihood of God's existence—in some way balance out so that the totality of k leaves the probability of the existence of God at O.5.10 Will k include the information that ordinary religious experiences and mystical religious experiences occur? Insofar as the inclusion of such information raises the probability of G on k above 0.5 we will have to exclude it. This may seem arbitrary and harmful to the theist's position. But it need not be construed in that way. If it should turn out that we have reason to believe that P is true and that P lowers the probability of God's existence, it is open to the theist to reply that the addition to k of our information concerning the occurrence of ordinary and mystical religious experiences restores the balance or even tips the scales in favor of theism. As I endeavored to make clear in "The Problem of Evil and Some Varieties of Atheism," I have not argued that no matter what other evidence a person has, the argument from evil will still make it unreasonable for that person (who understands the argument from evil and accepts the grounds for its premises) to believe in God. For one might have stronger evidence for the existence of God than is provided by the problem of evil for the nonexistence of God.11 k will not, of course, include either P or Q. Moreover, so that we do not beg an important question central to the criticism that P is not a good reason for Q, k will not include any explicit claim as to whether the goods we know are representative of all the goods there are.

/// We want to discover the answer to four questions: 1. Does P make Q more likely than it would otherwise be? That is, is Pr(Q/P&k) > Pr(Q/k)? 2. Does P make Q more likely than not?That is, is Pr(Q/P&k) > 0.5?12 3. Does P make G less likely than it would otherwise be? That is, is Pr(G/P&k) > Pr(G/k)? 4. Does P make G less likely than not?That is, is Pr(G/P&k) > 0.5?13 In this section we will give reasons for an affirmative answer to the first question and note a difficulty in one attempt to provide an affirmative answer to the second question. To begin our investigation, let's consider what Bayes's Theorem tells us about Pr(Q/P&k), the probability of Q given P and k. According to Bayes's Theorem, Pr(Q/P&k) = Pr(Q/k) X IJjggj^ • By reflecting on this equation, can we make any progress toward answering either of our first two questions? I believe we can. First, all of us will certainly agree that Pr(Q/P&k) < 1. Indeed, according to Alston and others, we have no reason at all

167

168

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument Jrom Evil

I

267

to think that Pr(Q/P&k) isn't less than 0.5. Moreover, since Q entails P, Pr(P/Q&k) = 1. Now, since Pr(Q/P&k) < 1 and Pr(P/Q&k) = 1, it follows that Pr(Q/k) < Pr(P/k).14 It is also clear that our background information k does not entail P, so Pr(P/k) does not equal 1. And from the conjunction of Pr(P/Q&k) = 1, Pr(Q/k) < Pr(P/k), and Pr(P/k) < 1, it follows that Pr(Q/P&k) > Pr(Q/k). So, we have reached a definitive answer to our first question. P does make Q more likely than it would otherwise be. But what of our second question? Does P make Q more likely than not? In an earlier paper151 gave an affirmative answer to this question and endeavored to support that answer with an argument I've since come to believe is inadequate. I noted first that the inference from P to Q is like an inference from "All the A's we've observed are B's" to "All A's are B's." I then argued that if we have observed many A's and found all of them to be B's, we have a prima facie good reason to believe that the A's we haven't observed will likely be B's as well. Thus we have a prima facie good reason to believe that all A's are B's. I now think this argument is, at best, a weak argument.16 To shore it up we would need some reason to think it likely that the goods we know of (the A's we've observed) are representative of the goods there are (the A's there are). Noting the variety of goods we know of would be relevant to this task. Having a good argument to think that most goods are known to us would also be relevant.17 But I now propose to abandon this argument altogether and give what I believe is a better argument for thinking that P makes Q more likely than not. Consideration of this new argument, however, must be postponed until we have discovered the answers to questions 3 and 4, the questions that are of ultimate interest to us.

IV If we substitute G (God exists) for Q (No good justifies God in permitting El and E2) in our earlier representation of Pr(Q/P&k) according to Bayes's Theorem, we get the following: Pr(G/P&k) = Pr(G/k) X ** • Recall that from our earlier discussion of k, Pr(G/k) = 0.5. Also note that if Pr(P/k) > Pr(P/G&k) then Pr(G/P&k) < Pr(G/k). (The general point here is that G makes P less likely than it would otherwise be if and only if P makes G less likely than it would otherwise be.) So, if we can succeed in showing Pr(P/k) > Pr(P/G&k) then we will have established an important point about what can be inferred from P. For we will have shown, first, that given k, P makes G less likely

THE PROBLEM OF EVIL 268

/

W I L L I A M L. R O W E

than it would otherwise be, Pr(G/P&k) < Pr(G/k), and, second, that given k, P makes G less likely than not, Pr(G/P&k) < 0.5. (This part follows from the first part and the fact that Pr(G/k) = 0.5.)18 But can we establish that Pr(P/k) > Pr(P/G&k)? I believe we can. Let's begin by considering Pr(P/G&k). In the end we shall discover that it will not matter much what value is assigned to Pr(P/G&k), so long as it is less than I. 19 But for the moment it will be instructive to assign it a value of 0.5. Theodicists who believe that there are goods that we know of that could justify God in permitting El and E2 may think that Pr(P/G&k) < 0.5. Other theists who find themselves quite incapable of thinking of any good whose realization might be God's reason for permitting El and E2 would undoubtedly be sympathetic to the assignment of 0.5, and may even think that it should be somewhat higher than that.20 But for instructive purposes, let's assume that Pr(P/G&k) = 0.5. If so, can we determine Pr(P/k)? Or, if we cannot determine Pr(P/k), can we at least determine that it is > 0.5? The truth is that given that Pr(G/k) = 0.5 and given that Pr(P/G&k) is 0.5, we can determine exactly what Pr(P/k) is.To see this, let's first establish that Pr(P/k) must lie somewhere between a low of 0.25 and a high of 0.75. We establish this by using the rule of elimination to determine the value of Pr(P/k): Pr(P/k) = [Pr(G/k) X Pr(P/G&k)] + [Pr(~G/k) X Pr(P/~G&k)l. Since Pr(G/k) and Pr(P/G&k) are both 0.5, we have 0.25 on the left side of the plus sign. And since Pr(~G/k) = 0.5 and Pr(P/~G&k) must lie somewhere between zero and 1, it follows that on the right side of the plus sign the number must be somewhere between zero and 0.5. Therefore, Pr(P/k) is somewhere between 0.25 and 0.75. Now if we could go no further toward establishing Pr(P/k), very little of interest could be established concerning Pr(G/P&k). But we can go further. In fact, we can establish that if both Pr(G/k) and Pr(P/G&k) = 0.5, then Pr(G/P&k) = 0.333.21 This can be shown once we note that Pr(P/~G&k) = 1. For, as we earlier noted, P is entailed by ~G. So, since P is entailed by ~G, Pr(P/~G&k) = 1. If both Pr(G/k) and Pr(P/G&k) = 0.5, Pr(P/k) = 0.75, with the result that Pr(G/P&k) = 0.333. And this shows us a general truth: given that Pr(G/k) = 0.5 and Pr(P/G&k) < 1, P not only lowers the probability of G, it also makes G lower than 0.5. But we need to distinguish these two results.The first, that P lowers the probability of G, makes G less likely than it would otherwise be, is bound to be true provided that Pr(P/G&k) < 1. For given that Pr(P/G&k) < 1 (as almost all would agree), it must be that Pr(P/k) > Pr(P/G&k). And, if Pr(P/k) > Pr(P/G&k) then Pr(G/P&k) < Pr(G/k)—no matter what Pr(G/k) happens to be. The second, that P makes G lower than 0.5, does depend to some degree on our original assignment of 0.5 to Pr(G/k). If we sufficiently increase Pr(G/k), then although Pr(G/P&k) will always be lower than Pr(G/k), it will rise above 0.5. And this is what we should expect. For if the existence of God is sufficiently probable apart from the negative impact of P, then the existence of God may still

169

170

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

I

269

be more probable than not even when the negative impact of P is taken into account. No one should disagree with this result. To sum up then: Since Pr(Q/P&k) < 1 and Pr(P/k) < 1, it follows that l.Pr(Q/P&k)>Pr(Q/k). And given that Pr(G/k) = 0.5 and that Pr(P/G&k) < 1, we have established two other important points. 2. Pr(G/P&k) < Pr(G/k). 3. Pr(G/P&k) < 0.5. What we have yet to establish is 4.Pr(Q/P&k)>0.5. Let's return to our first formula concerning Pr(Q/P&k): Pr(Q/P&k) = Pr(Q/k) X ^pff ) ' Since Pr(P/Q&k) — 1, we can eliminate it from the formula with the following result:

ft Pr(Q/k).23 So, given that Pr(P/k) < 1, if Pr(Q/k) = 0.5, it follows that Pr(Q/P&k) > 0.5. In fact, if our argument is correct, Pr(Q/P&k) = 1 - Pr(G/P&k), and this will be true no matter what non-zero assignments less than 1 are made to Pr(P/k), Pr(G/k), and Pr(Q/k). Returning to our summing up, we can now add 4. Pr(Q/P&k) > 0.5 to our list of propositions we have established given that Pr(G/k) = 0.5 and that Pr(P/G&k) < 1 . This is the better argument I mentioned earlier for the view that

THE PROBLEM OF EVIL 270

/

W I L L I A M L. R O W E

P not only makes Q more likely than it otherwise would be but also makes it more likely than not.24 Return now to the formulation of the argument at the beginning of this chapter. The argument proceeds inductively from P to Q and deductively from Q to ~G. Given the discussion in sections II and III, it is clear that we can simplify the argument considerably by bypassing Q altogether and proceeding directly from P to ~G. And that is what I now propose to do. Our evidential argument from evil, therefore, can now be stated more succinctly as P: No good we know of justifies God in permitting El and E2. therefore, it is probable that ~G: There is no omnipotent, omniscient, perfectly good being. So far, we have not considered what justification we might have for the initial premise P. As I noted earlier, we need to consider this issue in the light of objections raised by Alston in "The Inductive Problem of Evil and the Human Cognitive Condition." What we have been considering is the justification we have for the claim that P makes ~G probable. And what we have seen is that given that the probability of G on our background information k is 0.5, there is a compelling argument that P makes ~G more probable than G. So it does seem that we are justified in holding not only that P makes G less likely than it would otherwise be, but also justified in holding that P makes G less likely than ~G. But we need to look more deeply into the degree of support P provides for ~G, particularly in light of Wykstra's penetrating discussions of this point.

V I refer the reader to Wykstra's new essay (chapter 7 in this volume), "Rowe's Noseeum Arguments from Evil," for a full explication of CORNEA (Condition of Reasonable Epistemic Access) and his current understanding of its application to my inference from P to ~G.25 But in brief it comes to this: Wykstra's CORNEA, as he now explains it, tells us that we are entitled to believe Q (or ~G) on the basis of P only if it is reasonable for us to believe that If P is true, it is likely that Q (~G) is true. As he puts it: "So all CORNEA really says is that premise P justifies our believing conclusion Q only if it is reasonable for us to believe that if P is true, then Q is likely true." Since we have deleted Q from our statement of the evidential argument, we can take the proposition in question to be: If P [No good we know of justifies God in permitting El and E2], it is likely that ~G [There is no omnipotent, omniscient, perfectly good being]. In his 1984 paper, to which I responded, Wykstra argued that given God's omniscience and that he is the creator of all that is, it is quite likely that the goods for the

171

172

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

271

sake of which he permits many sufferings would be altogether beyond our ken. Noting my claim that in a great many cases of evil the justifying goods are nowhere within our ken, he remarked: The linchpin of my critique has been that if theism is true, this is just what one would expect; for if we think carefully about the sort of being theism proposes for our belief, it is entirely expectable—given what we know of our cognitive limits— that the goods by virtue of which this Being allows known suffering should very often be beyond our ken.26

In his new essay for this volume he argues that it is not necessary for him to take such a strong position. He now holds that it would be sufficient to take a more "modest" approach, arguing that it is just as likely as not that these goods should be beyond our ken. So, instead of arguing that it is quite likely that the goods for the sake of which God permits many sufferings would be beyond our ken, Wykstra now opts for the more modest claim that it is as likely as not that these goods would be beyond our ken. He also argues that I cannot simply take the inference from P to ~G as a good inference, one we are entitled to make unless someone shows that we are not. Finally, he finishes his new essay by reworking and defending his analogy between God and the good parent who often acts in ways the child cannot understand. Let's first consider his point that for my argument to work it must be reasonable to believe that if P is true then it is likely that ~G, that it is not sufficient just to claim that we are entitled to believe it unless we have a good reason not to believe it, thus throwing all the burden on the theist to argue that P s being true does not make it likely that God does not exist. I agree with Wykstra that it must be reasonable to believe that if P is true then it is likely that ~G. And I hope that the first several sections of this chapter show that relative to background information k, on which G is as likely as not, P does make it likely that ~G. But this is. a somewhat complicated matter and we need to look at it a bit more fully. Given the level playing field assumption that the probability of God's existence is 0.5 relative to the theist's and nontheist's shared background information k, we raised the question of whether P makes ~G unlikely. We saw that whatever probability (less than 1) we assign to G on k, it will be the case that Pr(G/P&k) is lower than Pr(G/k).27We then saw, of course, that Pr(G/P&k) < 0.5. Have we then satisfied Wykstra's requirement that it be reasonable to believe that if P is true then it is likely that ~G? This depends on how we understand Wykstra s requirement. We've shown that it is reasonable to believe that P lowers the probability of G. We've shown that on our background information k, it is reasonable to believe that if P is true ~G is more likely than not. But have we shown that relative to background information k it is reasonable to believe that if P is true G is unlikely? Well, it might seem that we have. But suppose we put Wykstra's requirement like this. It must be reasonable to believe that P is a good reason to believe that G isfalse. Have we shown that the conjunction P&k is a good reason to believe ~G? If we have, then we may conclude that it is reasonable to believe that P is a good prima facie reason to believe that G is false. But what we have shown is that the conjunction P&k is a

THE PROBLEM OF EVIL 272

/

W I L L I A M L. R O W E

good reason to believe that G is less likely than not. And to have a good reason to believe that G is less likely than not is not the same as having a good reason to believe that G is false. For suppose we establish that Pr(G/P&k) = 0.45. If we do establish this, then we may agree that P is a good reason to believe that G is less likely than not. But we would require something more than this for P to be a good reason to believe that G is false. 0.45 is so close to 0.5 that the rational thing to do would be to suspend judgment about the truth or falsity of G. And I think what Wykstra requires of me is some reason to think that the probability of G on P&k is low enough for P to be a good reason to believe that G is false. Have I provided any such reason? We've taken note of two major points inWykstra's essay. First, we've seen that he holds that my evidential argument succeeds only if it is reasonable for me to believe that if P is true it is likely that ~G. Second, we've seen that he thinks that if the considerations he brings forth (the parent-child analogy, the moral depth line, etc.) suffice to show that if G is true P is just as likely as not (the "modest" attack), this will render it unreasonable for me to believe that if P is true it is likely that ~G. Concerning the first point, I've responded by noting (a) that P lowers the probability of G, and (b) that given k, P makes G less probable than not. The remaining issue is how low G must be on P&k for P&k to be a good reason to believe ~G. On this last point, we might settle on something like ~G being twice as likely as G. If so, then given our level playing field assumption, Pr(G/k) = 0.5,1 believe I can satisfy Wykstra's requirement. For I think we have reason to believe that Pr(P/G&k) is no greater than 0.5. And given that Pr(G/k) = 0.5 and that Pr(P/G&k) is no greater than 0.5, it follows that Pr(G/P&k) is no greater than 0.333. It is clear, however, that Wykstra thinks that a probability of 1 /3 is not low enough to justify disbelief. He holds, I believe, that if the initial probability of G on k is 0.5 and the probability of P on G & k is 0.5, then, although P lowers G's likelihood to 1/3 and increases ~G's likelihood to 2/3, thus making ~G twice as likely as G, P does not sufficiently lower G's probability to justify abandoning agnosticism and taking up atheism. I have some sympathy for this view. Using some concepts employed by Chisholm, so long as we had only k to go on we might say that believing theism was not more reasonable than believing atheism, and believing atheism was not more reasonable than believing theism. Adding P to k, however, shifts things in favor of atheism. It is now more reasonable to believe atheism than it is to believe theism. But this may be true without it being true that it is now more reasonable to believe atheism than it is to withhold judgment on the matter of God's existence (agnosticism). How much must P lower G's probability in order for Wykstra to think that we are justified in moving from agnosticism to atheism, assuming that our initial position was one of agnosticism? It may well be that there is no sharp cut-off point here. But even if there is, Wykstra's view is that the cutoff point is less than 1 / 3. In the penultimate version of his paper he suggested that if the initial probability of a hypothesis is 0.5 and some evidence lowers its probability to 0.2, this entitles a shift from "square agnosticism" to "semi-square atheism"; whereas if the evidence lowers its probability to 0.17, this entitles a shift

173

174

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argumentfrom Evil

/

273

from "square agnosticism" to "square atheism."28 But in the final version he suggests that "levering evidence," evidence sufficient to lever a belief-state from square agnosticism to square atheism, must lower the probability of theism virtually to 0, "to something under .05." In support of this rather strict view of what constitutes "levering evidence,"Wykstra remarks: "In ordinary contexts, the things we typically squarely believe each day (say, that one is wearing shoes and socks, that it is cloudy outside, etc.) are, in effect, rated so near to 1 as to make no practical difference."29 I thinkWykstra's examples here are typically cases of knowledge, rather than belief in the absence of knowledge. Typically, I know that I'm wearing socks and shoes. Typically, when outdoors I'm in the position to know that it is cloudy (when it is). But when clouds are dark and I hear what sounds like thunder, I may form the belief that it will rain soon. Is my belief that it will rain soon rated something well above .95? I don't think so. But this is a minor point. The important point here is Wykstra's observation that evidence that lowers an initial probability from 0.5 to 1/3 falls somewhat short of "levering evidence." I believe Wykstra is right about this point. Given our judgments about the probability of G on k and the probability of P on G & k, P is evidence for atheism. But it is not "levering evidence"; it does not justify a shift from square agnosticism to square atheism. In reply to this interesting criticism, I have two points to make. First, it must be acknowledged that P does lower the probability of G. (Theists have been rather reluctant to acknowledge that the evils in our world, or what we have reason to believe about some of them, render God's existence less likely than it would otherwise be.) And if we start with our judgments about the probability of G on k and the probability of P on G & k, P lowers G's probability significantly, making ~G twice as likely as G. Although it does not lower G's probability sufficiently to move one from square agnosticism to square atheism, it does make it more rational to believe atheism than to believe theism. Second, we should remember that we are here considering one particular argument from evil, an argument based on the claim that no good we know of provides sufficient justification for God to permit two instances of evil, El and E2. As I suggested in an earlier essay,30 there are two basic types of arguments from evil, one beginning from the fact that there exist evils "that seem to us to serve no good whatever, let alone one that is otherwise unobtainable by omnipotence," and the other that starts "from the somewhat less complex fact that the world contains vast amounts of intense human and animal suffering." In reaching any overall conclusion as to the force of the evidential argument from evil on the rationality of belief in God, we need to consider arguments of the second type as well. And the important point for our purposes here is that unlike the argument I've presented—resting as it does on a statement, P, that is not a part of k—arguments based simply on the existence of evils we know to occur are based on statements already contained in k. For k includes information about the kinds, amounts, and distribution of evils and goods in the world. Since I do think arguments based on our shared information about the kinds, amounts, and distribution of evils and goods in the world do have merit, I earlier noted in an endnote that my own view is that Pr(G/k) < 0.5. Suppose that after a thorough investigation of arguments of this sort we were to come to the conclusion that

THE PROBLEM OF EVIL 274

/

W I L L I A M L. R O W E

Pr(G/k) < 0.5. What bearing would this have on the argument that I have given in this chapter? It would mean that Pr(G/P&k) < 0.333. How much less would depend on how low the probability of G on k is and how low the probability of P given G and k is. If Pr(G/k) = 0.2 and Pr(P/G&k) = 0.25, Pr(G/P&k) would be < .06. Of course, even if we were to agree to all this, we would not have answered Wykstra's point that my argument from P to ~G does not suffice to leverage us from square agnosticism to square atheism. For, since we are no longer taking Pr(G/k) to be 0.5, we are not beginning with square agnosticism. But if we were then to introduce k' as k minus the information about the kinds, amounts, and distribution of evils and goods in the world, we could then say that given Pr(G/k') as 0.5, the combination of the two sorts of arguments from evil may suffice to move us from square agnosticism to square atheism. Wykstra's second point, his "modest" attack, turns out to be too modest to do much good. If all Wykstra does is establish that if God exists it is just as likely as not that P, we will be left with the result that Pr(G/P&k) = 0.333. Of course, since we've agreed that a probability of 0.333 is not sufficiently low to warrant belief in atheism, as opposed to belief that atheism is epistemically preferable to theism, Wykstra can rest content that my argument is insufficient to justify a move from square agnosticism to square atheism. But, as I've pointed out, the argument would still show that P significantly lowers the probability of G and makes belief in atheism more reasonable than belief in theism. To do anything more, Wykstra needs to revert to his strong attack and argue that Pr(P/G&k) is very high. Even if he is entirely successful in doing so, however, so long as Pr(P/G&k) < 1, it will still be true that Pr(G/P&k) < 0.5, that atheism is more probable than not. But if Wykstra succeeds in showing that Pr(P/G&k) is very high, Pr(G/P&k) will be quite close to 0.5. And this, of course, is not unimportant. For it would show that P does not significantly lower the probability of G. Two questions remain: Are there good reasons for believing that Pr(P/G&k) is very high? Are there good reasons for believing that P is true? In the next two sections I will take up these questions.

VI Wykstra believes that if God exists it is quite likely that the goods for the sake of which he permits many instances of suffering (including Bambi's and Sue's) are beyond our ken. He argues from what he calls "the parent analogy." In his 1984 essay he claimed that our discerning most of these goods is "about as likely as that a one-month-old should discern most of his parents' purposes for those pains they allow him to suffer—which is to say, it is not likely at all."31 In his new essay, he further develops and defends this argument from analogy.32 Before examining Wykstra's analogical argument for the view that Pr(P/G&k) is quite high, it will be helpful to see the bearing of this point on our

175

176

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

27 S

evidential argument from evil. If God exists then some good justifies him in permitting Sue's horrendous suffering on being beaten, raped, and strangled. Either that good is a good we know of or it is not. Suppose that P is true, that no good we know of justifies God in permitting the sufferings of Bambi and Sue. Would this fact be a good reason for thinking it likely that God does not exist? As we've seen, this depends to a considerable extent on the degree to which P lowers the probability of G. Of course, P does lower the probability of G. But our question is whether it significantly lowers the probability of G. And the importance of Wykstra's parent analogy argument is that if it is correct we have a reason to think that P does not significantly lower the probability of G. So, much hangs on whether it can be successfully shown that if G is true, P is just what one would expect, that G makes P quite likely. We must now explore Wykstra's argument to this effect. Is our intellectual grasp of goods for the sake of which God (if he exists) permits horrendous human and animal suffering analogous to a one-month-old infant's intellectual grasp of his parents' purposes for those pains they allow him to suffer? It hardly seems so. For a one-month-old infant hasn't developed the concepts necessary for even contemplating the proposition that good purposes may justify parents in permitting pains.33 Adult human beings, on the other hand, have the intellectual equipment to distinguish intrinsic goods from extrinsic goods, to distinguish different kinds of. intrinsic goods, to recognize certain intrinsic goods as superior to others, to form an idea of goods that have never been experienced by living human beings on earth (e.g., total felicity in the eternal presence of God), and to make some reasonable judgments about what goods an omnipotent being would (or would not) be able to bring about without permitting various instances of horrendous suffering. Of course, we have to allow that there may be kinds of intrinsic goods we have not thought of. But we do have reason to believe both that every intrinsic good necessarily involves conscious experience and that the highest intrinsic good human beings are capable of involves conscious experience of God. So, we know of many goods and we know of some of the very highest goods that human beings can experience. Why then doesWykstra believe that the parent analogy provides a strong argument for the view that the goods that justify God in permitting much horrendous suffering will be goods of which we have no knowledge? Wykstra argues that the greater the degree of the parents' intelligence, care for the future life of the child, and ability, the more likely it is that their permitting present sufferings of the child serves goods in the distantfuture. Since God has unlimited intelligence, cares infinitely about the totality of each creature's life, and is unlimited in power, the argument from analogy implies that the goods justifying God in permitting horrendous human and animal suffering are often likely to be realized in the distant future. What are we to make of this argument? Well, I don't think we should dismiss it out of hand simply because a parent's intelligence, loving concern for her children, and ability to provide are finite, whereas these features in God are infinite. But we should note that these differences make for significant disanalogies between the loving parent and God. The following are often true of the loving parent but are very likely never true of God:

THE PROBLEM OF EVIL 276

/

W I L L I A M L. R O W E

A. The parent does not prevent the child's suffering (due to disease, etc.,) simply because the parent is unaware of the cause of the suffering or unable to prevent the suffering. B. The parent does not prevent the child's suffering because the parent has other duties to fulfill that preclude her from being in a position to prevent the suffering. Unlike God, parents cannot be everywhere at once. C. The parent permits present sufferings for distant goods not because these goods are incapable of existing sooner, or better for being distant, but because of insufficient intelligence and ability to realize the goods in the present or near future. But let's put aside these disanalogies and focus on what is, I believe, the major weakness of the argument based on the analogy between God and the loving parent. What happens when a loving parent intentionally permits her child to suffer intensely for die sake of a distant good that cannot otherwise be realized? In such instances the parent attends directly to the child throughout its period of suffering, comforts the child to the best of her ability, expresses her concern and love for the child in ways that are unmistakably clear to the child, assures the child that the suffering will end, and tries to explain, as best she can, why it is necessary for her to permit the suffering even though it is in her power to prevent it. In short, during these periods of intentionally permitted intense suffering, the child is consciously aware of the direct presence, love, and concern of the parent, and receives special assurances from the parent that, if not why, the suffering (or the parent's permission of it) is necessary for some distant good. If we do apply the parent analogy, the conclusion about God that we should draw is something like the following: When God permits horrendous suffering for the sake of some good, if that good is beyond our ken, God will make every effort to be consciously present to us during our period of suffering, will do his best to explain to us why he is permitting us to suffer, and will give us special assurances of his love and concern during the period of the suffering.34 Since enormous numbers of human beings undergo prolonged, horrendous suffering without being consciously aware of any such divine presence, concern, and explanations, we may conclude that if there is a God, the goods for the sake of which he permits horrendous human suffering are more often than not goods we know of. In any case, I think we are justified in concluding that we've been given no good reason to think that if God exists the goods that justify him in permitting much human and animal suffering are quite likely to be beyond our ken.35

VII At long last we come to the question of P itself. What reason do we have to believe that no good we know of justifies God in permitting El and £2? The main reason to believe P is this: When we reflect on some good we know of we can see that it is very likely, if not certain, that the good in question either is not good enough

177

178

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

277

to justify God in permitting El or E2 or is such that an omnipotent, omniscient being could realize it (or some greater good) without having to permit El or E2. Consider, for example, Sue's pleasure upon receiving some toys on her fourth birthday. Clearly that pleasure is not good enough to justify the permission by God of what we can only suppose to be her terror and pain on being brutally beaten, raped, and strangled when she was five years old. And if we begin to reflect on the various kinds of goods we know of, we will come to the sound judgment that many of them are not good enough to justify anyone in permitting El. Similar remarks can be made concerning various kinds and degrees of goods relative to the terrible pain Bambi endures for several days upon being badly burned in the forest fire. In short, we can see that various kinds of goods we know of simply aren't good enough to justify permitting horrendous evils such as El and E2. On the other hand, for those goods we know of that do seem to outweigh either El or E2, reflection on them leads us to the judgment that it is very likely, if not certain, that an omnipotent being could have realized the goods in question (or some better goods) without having to permit El or E2. Consider, for example, Sue's experiencing complete felicity in the eternal presence of God. While this good may justly be held to outweigh almost any horrendous evil that may befall Sue in her earthly life, it strains credulity to think that it is beyond the power of God to realize this good without having to permit Sue's being brutally beaten, raped, and strangled at the age of five years. It might be suggested that had Sue continued to live out her life she would have freely chosen to harden her heart against God with the result that she would have precluded herself from experiencing complete felicity in the eternal presence of God. Let this be so. If God knows this, he might then have some reason to permit her death at the age of five. On the other hand, it hardly makes sense for God to give us the freedom to develop into beings who have hardened our hearts against him and his eternal kingdom and then act to prevent us from having the chance to exercise that freedom. Moreover, we might expect that God could have brought about circumstances in which Sue would have freely refrained from hardening her heart against God. But suppose that God does have reason to permit Sue's death at an early age. Does God then have a sufficient reason to permit her to die by being brutally beaten, raped, and strangled? Clearly, given his omnipotence he could have realized her early death by more humane means. So, even when we consider some good we know of that does seem to outweigh E2, we have good reason to believe that if God could realize that good, he could realize it without permitting E2. And what of Bambi's terrible suffering? What good do we know of that outweighs it and is such that God could not have realized it without permitting El? When we reflect on any good we know of and consider Bambi's excruciating suffering, reason cries out that it is very likely, if not certain, that either the good is not good enough or God could have realized it without having to permit El (or something else as bad or worse). In "The Inductive Argument from Evil and the Human Cognitive Condition" (chapter 6 in this volume), William Alston does a masterful job of surveying the terrain of goods we know of in search of some good whose realization might justify God in permitting El or E2. Although the overall purpose of his essay is to

THE PROBLEM OF EVIL 278

/

W I L L I A M L. R O W E

establish that none of us is or can be justified in believing that no good justifies God in permitting El or E2, my interest here concerns only that part of his essay that focuses on P (No good we know of justifies God in permitting El and E2).36 In pursuing our question of whether we are justified in believing that P is true, we can do no better than to critically examine Alston's discussion of this issue. A major part of Alston's project is to consider familiar Christian theodicies and to explore their possible application to Bambi and Sue. These theodicies divide into those proposing divine reasons for suffering that are concerned with possible goods to those who endure the suffering and those proposing divine reasons for suffering concerned with possible goods not restricted to the sufferers. Examples he gives of the first sort are punishment for sin, soul-making, and having a vision of the inner life of God. Examples he gives of the second sort are the value of free will, benefits to those who cause or witness the suffering, and the value generated by a lawlike, natural order. Alston emphasizes that he is considering Christian theodicies only as live possibilities for divine reasons for permitting evil. He is not undertaking to show that any of these theodicies is correct. Since he does not spell out what he means by a live possibility in this context, I shall take a live possibility here to be something that we have no good reason to believe would not wholly or partially justify an omniscient, omnipotent, wholly good being in permitting some evil. The question then becomes whether any or all of the theodicies are live possibilities JOT Bambi and Sue. If they are not live possibilities, if they are implausible suggestions as to what might justify an omnipotent, omniscient being in permitting Bambi and Sue, then, insofar as his use of Christian theodicies is concerned, Alston will not have shown that no one is justified in believing P to be true. So far as goods for the sufferer are concerned, Alston believes that the prominent Christian theodicies fail to provide live possibilities for the cases of Bambi and Sue. But he thinks that he can establish that some goods we are familiar with (for example, the supreme fulfillment of one's nature) are such that we've no good reason to think that an omnipotent, omniscient being can obtain them without having to permit Bambi's and Sue's sufferings. But apart from noting that it is logically possible that Bambi's and Sue's sufferings are required for the fulfillment of their natures, Alston does little or nothing by way of "establishing" that we are not justified in precluding familiar goods like the supreme fulfillment of Bambi's nature or Bambi's having a vision of God from providing an omnipotent, omniscient being with a morally sufficient reason to permit Bambi's excruciating torment and death. The major theodicy concerned with a good that extends beyond the sufferer is the free-will theodicy. I will here consider Alston's efforts to use this theodicy to establish that we cannot be justified in believing P.37 According to this theodicy, God is justified in permitting evil actions and their consequences because he has bestowed on some of his creatures genuine freedom in a range of actions, and it is a conceptual impossibility for God to create a free agent with respect to some action and also determine the agent to choose (not choose) to perform the action. So, according to this theodicy, God permits certain horrors to occur

179

180

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

279

because to prevent them would be to prevent a certain degree of freedom in his creatures. Against this theodicy, Alston notes that it has been argued (1) that God could have created free creatures who always choose to do what is right, and (2) that permitting free will with respect to certain actions at certain times isn't worth the horrendous evils (for example, Sue's suffering) that result from the use of that freedom. Concerning the first objection, Alston points out that if we set aside middle knowledge, as he does in the chapter, God would not be able to create beings with genuine freedom and "guarantee" that they will always choose to do what is right. And he further notes that even if we grant middle knowledge, Plantinga "has established the possibility that God could not actualize a world containing free creatures that always do the right thing." Before considering his response to the second objection, we should note two points here. First, to conduct his case on the assumption that there is no middle knowledge considerably weakens Alston's argument. For there is no consensus on whether middle knowledge is possible. And if middle knowledge is possible, then we have reason to think that an omnipotent, omniscient being could have created a world with less evil, but as much good, as our world contains.38 Nor will it do to emphasize that Plantinga has established the possibility that even with middle knowledge God could not have created a world with free creatures that always do the right thing. For Alston needs to show that even with middle knowledge it remains a live possibility that God could not have created a world with free creatures who always do what is right. But all that Plantinga has shown is that it is a logical possibility that such would be the case. And here again, I'm afraid, we have a slide on Alston's part from what is a logical possibility to what is a live possibility, from what is broadly logically possible to what we have no good reason for thinking isn't so. Indeed, that such a slide has occurred is indicated by the sentence with which Alston begins his very next paragraph. "Thus we may take it to be a live possibility that the maintenance of creaturely free will is at least part of God's reason for permitting wrongdoing and its consequences" (emphasis mine). Suppose we endeavor to apply the free-will theodicy to the case of Sue's suffering on being beaten, raped, and strangled. The first question we need to ask is whether the possession of free will is something that is in itself of such great value as to merit God's permission of the horrendous moral evils in the world. I think the answer must be no. We should distinguish the intrinsic value of possessing free will from its extrinsic value. The mere possession of free will does not strike me as itself having much in the way of intrinsic value.39 But the possession of free will does seem necessary to attaining states that are of great intrinsic value. Thus, if we can agree that free will is necessary for the existence of things of great intrinsic value, we can agree that an omnipotent, omniscient, perfectly good being would likely endow his creatures (or some of them) with free will, provided that it does result to a sufficient degree in the things of great intrinsic value for which its possession was intended. But, of course, it is sometimes right to curtail a particular exercise of free will when one foresees or predicts that its exercise is evil and/or will result in considerable suffering. Since curtailing a particular exercise of free will does not significantly diminish a person's overall degree of freedom, the ques-

THE PROBLEM OF EVIL 280

/

W I L L I A M L. R O W E

tion at hand is whether it is rational to believe that an omnipotent, omniscient, perfectly good being would have prevented the particular exercise of free will (if that is what it was) Sue's attacker engaged in when he brutally beat, raped, and strangled the five-year-old child. As Alston puts the issue: presumably a tiny additional constriction such as would be involved in God's preventing Sue's attacker from committing that atrocity would not render things radically different, free-will-wise, from what they would have been without that. So God could have prevented this without losing the good emphasized by this theodicy. Hence we can be sure that this does not constitute a sufficient reason for His not preventing it, (113, above)

Alston, however, thinks that the preservation of Sue's attacker's free will is a live possibility for at least part of God's reason for permitting the suffering Sue undergoes at the hands of her attacker. For he reasons that the value of free will is such that God can intervene in only a small proportion of cases. Rowe's claim would then have to be that Sue's murder was so horrible that it would qualify for the class of exceptions. But that is precisely where the critic's claims far outrun his justification. How can we tell that Sue falls within the most damaging n percent of what would be cases of human wrongdoing apart from divine intervention? To be in a position to make such a judgment we would have to survey the full range of such cases and make reliable assessments of the deleterious consequences of each. Both tasks are far beyond our powers, (ibid.)

Alston's point seems to be this: God cannot intervene in all cases of the use of free will in doing evil that results in suffering, for this would severely limit human capacity to choose between good and evil. God, then, would intervene only in a certain percentage (n) of such cases. The cases in which God intervenes would be selected so as to minimize human suffering overall or to maximize human welfare. And we simply don't know enough about what would be cases of human wickedness (past, present, and future) apart from divine intervention to determine whether or not Sue's case would be included among the percentage of cases God would select to eliminate. Alston states: Hence, by the nature of the case, we are simply not in a position to make a warranted judgment that Sue's case is among the n percent worst cases of wrongdoing in the history of the universe. No doubt it strikes us as incomparably horrible on hearing about it, but so would innumerable others. Therefore, the critic is not in a position to set aside the value of free will as at least part of God's reason for permitting Sue's murder. (114)

Does this argument establish that we aren't justified in excluding the value of human free will as part of God's reason for permitting Sue's suffering? Well, if the basic premise of Alston's argument is correct, the argument does establish just that. For to be so justified, Alston requires that we compare the episode involving Sue with the whole range of cases of human wrongdoing in the universe (past,

181

182

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

I

281

present, and future) that would occur apart from divine intervention in order to determine whether Sue's case is sufficiently bad to warrant God in setting aside the value of free will in that case. And, of course, no human being knows enough to engage in such a comparison. But I do not think this knowledge is required. The free-will theodicy is built around the idea that the possession of freedom to do good and evil is a good in each person that God creates. What is important, therefore, is that each person have some measure of freedom to do good and evil. Now, as Alston will readily admit, (1) we don't possess unlimited freedom, and (2) it is sometimes right to curtail someone's freedom in order to prevent some horrendous evil act that results in considerable suffering to an innocent person. But if God were to select some person and effectively prevent that person from ever choosing to do an evil deed that results in suffering to an innocent person, we can agree that this might severely diminish the value of that person's freedom to do good and evil. So God has a reason to permit each person to effectively engage in doing good and evil acts. Of course, God would be able to intervene in some of those acts without significantly diminishing the person's general freedom to do good and evil. What, then, do we have to make a judgment about in order to determine whether the prevention of Sue's attacker's freedom to brutally beat, rape, and strangle Sue would have severely diminished the value of that person's freedom to do good and evil? Do we have to compare it with all the cases of human wrongdoing, past, present, and future, as Alston claims? Clearly not. What is at issue here is not some sort of amount of freedom to do good and evil in the entire universe, past, present, and future. What is at issue here is the degree and value of Sue's attacker's freedom to do good and evil. Would this particular intervention severely diminish the value of that individual's overall degree of freedom to do good and evil? And if it would, is the value of that individual's overall degree of freedom to do good and evil worth the price of permitting the act in question? Perhaps some rational judgments are required about these questions in order for us to be justified in taking Sue's suffering as an instance of gratuitous suffering. But this is a far cry from having to do a survey of all human acts of wickedness in the universe, past, present, and future.40 We've had a look at Alston's efforts to single out among goods we know of some live possibilities for a sufficient reason for God to permit Bambi's or Sue's suffering. Concerning goods we know of that are restricted to the sufferer, Alston concedes that "none of the sufferer-centered reasons I considered could be any part of God's reasons for permitting the Bambi and Sue cases." And because it would be wrong for God to permit horrendous suffering solely for the benefit of others, Alston concludes "that nonsufferer-centered reasons could not be the whole of God's reasons for allowing any case of suffering." So, Alston's final conclusion is that he must appeal to goods beyond our hen in order to argue that we cannot be justified in believing that God has no sufficient reason to permit £1 or E2. Thus, quite apart from the specific criticisms I have presented against his attempts to provide live possibilities for God's reasons to permit El and E2, Alston concedes that he hasn't shown that we cannot be justified in believing that no good we know of justifies God in permitting El and E2.

THE PROBLEM OF EVIL 282

/

W I L L I A M L. R O W E

Of course, it's one thing for the most talented philosophers and theologians to fail to show that we aren't justified in believing P, and another thing for us to be justified in believing P, and still another thing for us to show that we are justified in believing P. And I must confess that I know of no way to prove that P is true. What we do have is genuine knowledge that many goods we know of are insufficient to justify God in permitting El or E2. In addition we have very good reason to believe that many other goods we know of could be realized by an omnipotent, omniscient being without his having to permit El and E2 (or something just as bad). And, finally, we have the failure of theodicists to show how any of the goods we know of can plausibly be held, separately or collectively, to constitute a sufficient reason for God to permit El or E2. All this, I believe, gives us good reason to believe that P is true.

vn The evidential problem of evil derives its strength from our almost inescapable conviction that among the goods that fall within our intellectual grasp none can reasonably be thought to constitute God's justifying reason for permitting such horrendous evils as El and E2. For if we divide the possible justifying goods into those that fall within our intellectual grasp and those utterly beyond our ken, and then discover that none of the goods in the first category are justifying for God with respect to such horrendous evils, we significantly lower the likelihood of God's existence. This is particularly so if our antecedent expectations are that the justifying goods are as likely to fall in the first category as in the second. Tough-minded theists have held that the facts about evil in our world do not render God's existence less likely than not. Indeed, some have held that the facts about evil do not even make God's existence less likely than it would otherwise be. In this chapter I have argued that these views are seriously mistaken. Given our common knowledge of the evils and goods in our world and our reasons for believing that P is true, it is irrational to believe in theism unless we possess or discover strong evidence in its behalf. I conclude, therefore, that the evidential argument from evil is alive and well.41

NOTES

1. See Bruce Russell's chapter 10 in this volume for a classification of several kinds of evidential arguments from evil. 2. William L. Rowe, "The Problem of Evil and Some Varieities of Atheism" American Philosophical Quarterly 16 (1979): 335-41; chapter 1 in this volume. 3. See William L. Rowe,aEvil andTheo&cy? Philosophical Topics 16 (1988): 11932, and "Ruminations about Evil,** Philosophical Perspectives 5 (1991): 69-88. 4. William Alston uses "Bambf to refer to El and "Sue* to refer to E2. See

183

184

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

/

283

Alston, "The Inductive Argument from Evil and the Human Cognitive Condition," chapter 6 in this volume. 5. This objection can be remedied by complicating P and Q so that they include some clause concerning the prevention of some equal or worse evils. For sake of simplicity, I ignore this complication. See note 2 of "Evil and Theodicy," 131-32. 6. For an interesting development of this objection, see William Hasker, "The Necessity of Gratuitous Evil," Faith and Philosophy 9 (1992): 23-44. I discuss Hasker's view in "Ruminations About Evil." 7. See Stephen Wykstra, "The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of'Appearance,'1' International Journal for Philosophy of Religion 16 (1984): 73-93. This essay, along with my reply, is reprinted in The Problem of Evil, ed. M. M. Adams and R. M. Adams (Oxford University Press, 1990). Also see Wykstra's co-authored piece (with Bruce Russell), "The 'Inductive' Argument from Evil: A Dialogue," Philosophical Topics 16 (1988): 133-60, as well as his chapter 7 in this volume. Also see Terry Christlieb, "Which Theisms Face an Evidential Problem of Evil?" Faith and Philosophy 9 (1992): 45-64, Paul Draper, "Probabilistic Arguments from Evil," Religious Studies 28 (1992): 303-17, and James Sennett, "The Inscrutable Evil Defense," Faith and Philosophy 10 (1993): 220-29. 8. It is important not to confuse P with P*: No good we know of would justify God (if he exists) in permitting El and E2.

While P must be true if God does not exist, P* may be false if God does not exist. For even though God does not exist, it could still be true that if he did exist some good we know of would justify him in permitting El and E2. To avoid any confusion of P with P*, it will be helpful to keep in mind that the sentence "No good we know of justifies God in permitting El and E2" is here being used to express the negation of the following proposition: God exists & there exists a good we know of & that good justifies him in permitting El andE2.

9. The information that / exist near the earth's surface is not shared by most theists and nontheists. But we may well include such information in k, since most of us (at some time) know a proposition that corresponds to the one I know. It is also understood here that k includes certain stipulated information. In particular, k includes our information concerning El (the case of Bambi) and E2 (the case of Sue). 10. My own view is that Pr(G/k) < 0.5. For I think the information we possess concerning the abundance of various evils in the world renders G unlikely. And I do not think the other information in k manages to counterbalance the weight of our information about the abundance of evils in the world. But for purposes of finding a starting point for the "theist-nontheist dialogue" I am here putting this view aside, although I will return to it later. 11. In "The Problem of Evil and Some Varieties of Atheism," I made use of what I called "the G. E. Moore shift" to show how belief in God can be sustained as rational by having stronger evidence for God's existence than the evidence from evil constitutes for God's nonexistence. 12. Alternatively, is Pr(Q/P&k) < Pr(not-Q/P&k)? 13. Alternatively, is Pr(G/P&k) < Pr(not-G/P&k)? 14. If Pr(Q/k) = Pr(P/k) then Pr(Q/P&k) would be 1, which we know it not to be. If Pr(Q/k) > Pr(P/k) then Pr(Q/P&k) would be > 1, which is impossible. 15. "Evil and Theodicy," 123-24. 16. If we have observed many A's and all of them are B's, this fact will make the proposition "All A's are B's" more likely than it would otherwise be. But one proposition

THE PROBLEM OF EVIL 284

/

W I L L I A M L. R O W E

may make another more likely than it would otherwise be without making it more likely than not. 17. Michael Tooley advances such an argument in "The Argument from Evil," Philosophical Perspectives 5 (1991): 114-15. 18. This point is important, in part, because its first part has been denied by some in writing on the problem of evil. For example, Wykstra, in his 1984 essay, "The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of 'Appearance/" argued that propositions like P do not even "weakly disconfirm" G. A proposition weakly disconfirms a second proposition if, given the first, the second proposition is less likely than it otherwise would be. 19. That is, it won't matter if all we are trying to show is that Pr(G/ P&k) < 0.5. As we shall see later, for the resolution of other important issues it matters a great deal what value is assigned to Pr(P/G&k). 20. Some theists hold that although Pr(P/G&k) is at least 0.5, for all we know Pr(P/G&k) = 1. See, for example, Daniel Howard-Snyder, "Inscrutable Evil and the Silence of God," doctoral dissertation, Syracuse University, 1992. Assuming these theists will agree to an assignment of 0.5 to Pr(G/k), we can characterize this view as holding the following propositions: (1) unless the conjunction of G and k entails P, P lowers the probability of G, (2) if Pr(P/G&k) < 0.5 then P lowers the probability of G anywhere from 0.333 to 0, (3) we have a sufficient reason to believe that Pr(P/G&k) is not less than 0.5, (4) we are completely in the dark as to what assignment between 0.5 and 1 belongs to Pr(P/G&k), and therefore, (5) we can be confident that if P lowers the probability of G, it does not lower it beyond 0.333. This view clearly merits consideration. Here I will only note that it is somewhat odd to suggest that human reason is fully adequate to determine that Pr(P/G&k) cannot be a decimal point below 0.5, but utterly inadequate to judge whether it is closer to 0.5 than it is to 1. 21. Strictly speaking, this isn't exactly right. Pr(G/P&k) - 1 / 3. But for reasons of uniformity I will continue to use the approximate decimal equivalent. 22. Common to most theists and nontheists is a principle (a necessary truth) to the effect that if God exists and some horrendous evil exists then there is some good whose realization justifies God in permitting that evil. Given G&k, it follows that God exists and that El and E2 exist. And given the principle just noted, it will follow that there is some good that justifies God in permitting El and E2. But if this is so, Q is false. For Q says that no good justifies God in permitting El and E2. 23. "Evil and Theodicy," 123-24. 24. As with our conclusion that Pr(G/P&k) < 0.5, our conclusion that Pr(Q/P&k) > 0.5 depends to some degree on our original assignment of 0.5 to Pr(G/k). 25. Actually, Wykstra applies CORNEA to my inference from P to Q. But his point remains the same when extended to my more succinct statement of the argument which bypasses Q with a direct inference from P to ~G. For some important criticisms of Wykstra's CORNEA, see Daniel Howard-Snyder, "Seeing through CORNEA," International Journal/or Philosophy of Religion 32 (1992): 25-49. 26. Italics mine. I have taken this remark from the reprint of Wykstra's 1984 essay in The Problem of Evil, ed. M. M. Adams and R. M. Adams, 159. 27. It is taken for granted here that Pr(P/G&k) < 1. 28. See note 7 of "Rowe's Noseeum Arguments from Evil," chapter 7 in this book. 29. Ibid. 30. William L. Rowe, "The Empirical Argument from Evil," in Rationality, Religious Belief, and Moral Commitment, ed. Robert Audi and William J. Wainwright (Ithaca: Cornell University Press, 1986), 245-47. 31. Quoted in his new essay written for this volume (chapter 7). 32. See chapter 7. 33. We should note that Wykstra invites us to adjust the infant's age beyond his

185

186

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION The Evidential Argument from Evil

I

285

one-month suggestion. But it is striking, nevertheless, that he thinks the one-month age is appropriate for his analogy. 34. If the good is one we know of, then we have some chance of recognizing that good as one that may well be God's reason for permitting the suffering in question, thus reducing the need for God to attend to us directly, providing us with special assurances and explanations of why the suffering (or his permission of it) is necessary for some good. 35. Theists may say that there are special reasons applying to God, but not to loving parents, that prevent him from making his loving concern for our sufferings apparent to us, thus adding a further epicycle to the theistic response to evil.Tough-minded theists may insist that it is human perversity itself that prevents God from responding to human suffering in the way the loving parent responds to her child's suffering. Tenderminded theists may say that our freedom in relation to God would be destroyed if he were not to remain hidden during our times of apparently pointless travail and suffering. For helpful discussions of questions concerning the hiddenness of God, see Daniel HowardSnyder, "The Argument from Divine Hiddenness" and John Schellenberg, Divine Hiddenness and Human Reason (Ithaca: Cornell University Press, 1993). 36. I have critically discussed Alston's essay in Rowe, "William Alston on the Problem of Evil," in The Rationality of Belief and the Plurality of Faiths, ed. Thomas D. Senor (Ithaca: Cornell University Press, 1994). A good bit of my discussion here is taken from that paper. 37. For an examination of his discussion of other theodicies that emphasize a good not restricted to the sufferer, see my essay "William Alston on the Problem of Evil." 38. For a brief discussion of this point, see my essay "Ruminations about Evil," 74-76. 39. Thus, it seems to me that the free-will theodicy needs to be included within something like Hick's soul-making theodicy, a theodicy that stresses some intrinsic goods for which free will is a necessary condition. 40. One might argue that were God to intervene in Sue's case He would have to intervene in every similar case, with the result that no human being would be free to do evil acts of the kind represented by Sue. I respond to an argument of this sort in connection with Alston's discussion of the theodicy that rests on the need for the world to operate in accordance with relatively stable laws of nature. See "William Alston on the Problem of Evil." 41. For comments on earlier drafts of this chapter, I am grateful to Martin Curd, Paul Draper, William Gustason, Bruce Russell, Dan Howard-Snyder, Eleonore Stump, William Wainwright, and David Widerker.

CHAPTER TEN

In Defense of 'The Free Will Defense7

Alvin Plantinga's 'Free Will Defense' is intended to establish that it is logically possible both that God exists and that evil exists. Given certain moral and metaphysical assumptions (including incompatibilism), Plantinga's argument for this conclusion is, I believe, fairly compelling. Daniel HowardSnyder and John O'Leary-Hawthorne, however, think that his argument, even given these assumptions, does not succeed. Central to their discussion is the conception of transworld sanctity, a concept patterned after Plantinga's concept of transworld depravity. After making some remarks about these two concepts, I will state the central thesis of their paper and then offer an argument against that thesis. I

Each creature that exists in any possible world has a unique property E (that creature's essence) such that any existing thing (in any world) that has that property simply is that creature and any existing thing (in any world) that lacks that property isn't that creature. A creaturely essence, being a property, exists in every possible world. For properties are necessarily existing entities. However, the properties that are creaturely essences are not instantiated in every world. There are worlds in which I do not exist. But there is no world in which my essence does not exist. What is it for a creaturely essence to be transworld depraved or transworld sanctified? Consider all those worlds in which a creaturely essence E is instantiated, E's instantiation has libertarian freedom to do right or wrong, but always does what is right. We can think of these worlds as E perfect worlds, or for short, EPWs (worlds in which E is instantiated, E's instantiation is free to do right or wrong, but always does what is right). Consider now * This paper is a response to the previous published paper by Daniel Howard-Snyder and John O'Leary-Hawthorne, Transworld sanctity and Plantinga's Free Will Defense', IJPR, Volume 44, No. 1 (August 1998), pp. 1-21.

188

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 116

WILLIAM L. ROWE

some EPW and a world just like it but in which E's instantiation uses he freedom in a certain situation S to do one wrong action. We can think of that as an EP-lW,a world in which E's instantiation has a nearly perfect record. Now the interesting point is that it is not up to God whether E's instantiation will in situation S do right (with the result that we have an EPW) or wrong (with the result that we have an EP-1W). And it is logically possible that E's instantiation will go wrong in some situation S no matter what EPW God tries to create. If that logical possibility should be the way things are, E's essence would suffer from transworld depravity. Just as E is transworld depraved provided no EPW is creatable by God, so E is transworld sanctified provided every EPW is creatable by God.1 Suppose that before God creates any creature it is true that some creaturely essence E is transworld sanctified. If that is so, then God can bring about any of the EPWs. But then there also will be worlds that he cannot bring about, those worlds closest to EPWs but in which E's instantiation goes wrong with respect to some action she performs rightly in the relevant EPW.

II

Plantinga's celebrated free will defense is an ambitious project. Not content merely to display defects in arguments that purport to show the incompatibility of God exists and evil exists, he proposes to establish the compatibility of God exists and evil exists by finding a third proposition that is compatible with God exists and which in conjunction with God exists entails that evil exists. Following our authors's suggestion, we may take that third proposition to be R: God created a world containing moral good, but it was not within his power to create a world containing moral good and no moral evil Clearly, if R is possible then God exists is logically consistent with evil exists. But, as our authors observe - and as Plantinga is aware2 - we won't have established or shown that God's existence is compatible with the existence of evil unless we know or reasonably believe that R is possible. So, how does Plantinga establish that R is possible? Or, to be more exact, how does Plantinga establish that it is possible that it was not in God's power to create a world with moral good and no moral evil? He does so by developing the concept of transworld depravity and claiming that the following proposition is true. It is logically possible that (TD) every (creaturely) essence suffers from transworld depravity.

THE PROBLEM OF EVIL IN DEFENSE OF 'THE FREE WILL DEFENSE'

189 117

but is this claim of Plantinga's known by us or reasonably believed? Our authors maintain that it is not. They provide a definition of transworld sanctity,3 and ask us to consider the following proposition: It is necessary that some (creaturely) essence or other is blessed with transworld sanctity. They then note that this proposition is inconsistent with Plantinga's claim that TD is possible, insist that we have no good reason to think that this proposition is false, and conclude, therefore, that we have no good reason to think that Plantinga's claim is true. Having disposed of Plantinga's free will defense, our authors then suggest a much weaker argument than Plantinga's, an argument intended not to establish that the proposition God exists is compatible with the proposition evil exists, but only that it is not reasonable to believe that God exists and evil exists are incompatible, even if it should be true that they are. It is this consolation prize they offer to Plantinga. Ill

Contrary to their claim that we have no good reason to reject the proposition that It is necessary that some essence or other is blessed with transworld sanctity, I believe we do have good reason to reject that proposition. Perhaps I can approach my reason for this in the following way. Suppose there is an urn containing two balls numbered 1 and 2. Suppose I necessarily randomly select one of the two balls, and then, without returning it, select the remaining ball. Is it possible that my first selection is not ball number 2? Clearly the answer is yes. Is it possible that my second selection is not ball number 2? Again the answer is yes. Is it possible that neither of my selections is ball number 2? Clearly, it is not possible that neither selection is ball number 2. Suppose we change things slightly. Instead of not returning the first ball selected before making a second selection, the first ball selected is necessarily returned to the urn before making another selection. We ask again whether it is possible that my first selection is not ball number 2. We also ask whether it is possible that my second selection is not ball number 2. As before, the answer in each case is yes. But now, unlike before, it is possible that neither of my selections is ball number 2. That is, it is possible that both of my selections are ball number 1. What accounts for the difference? It is this. In the first situation, where, necessarily the first ball selected was not returned prior to the second selection, from the fact that the first ball selected is not number 2, it logically follows that the second ball selected is number 2. In the second situation, where, necessarily, the first ball was returned prior to

190

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 118

WILLIAM L. ROWE

the second selection, from the fact that the first ball selected is not number 2, it does not logically follow that the second ball selected is number 2. Given the necessity of the arrangements in the first situation (the first ball selected is not returned), while it is 0 (the first ball selected is not number 2) and 0 (the second ball selected is not number 2), it doesn't follow that 0 (the first ball selected is not number 2 & the second ball selected is not number 2). It does not follow because it is D (if the first ball selected is not number 2, then the second ball selected is number 2). But in the second situation (necessarily, the first ball selected is returned) the inference is correct. For in the second situation it is ~ D (if the first ball selected is not number 2, then the second ball selected is number 2). So, in the second situation from 0 (the first ball selected is not number 2) and 0 (the second ball selected is not number 2), it does follow that 0 (the first ball selected is not number 2 & the second ball selected is not number 2). I've labored over this point because I think it is directly relevant to the question of whether our authors are correct in claiming that we have no good reason for rejecting the following proposition: It is necessary that some essence or other is blessed with transworld sanctity. They allow that each essence is such that it is possible that it is not blessed with transworld sanctity.4 There is no particular essence, say Al Capone's essence, or even Gandhi's essence, such that it necessarily has transworld sanctity. For each such essence, our authors may allow that we have good reason to assert that it is possible that that particular essence is not blessed with transworld sanctity. What then of the claim that it is possible that no essence is blessed with transworld sanctity? Here they deny that we have good reason to make this assertion. I think we do have good reason to make this assertion, a reason similar to the reason we have for asserting that in the second urn situation it is 0 (the first ball selected is not number 2 & the second ball selected is not number 2). To parallel our earlier assumption that there are only two drawings from the urn, suppose that there are only two creaturely essences, El & E2. Our authors agree that it is 0 (El suffers from transworld depravity) and 0 (E2 suffers from transworld depravity), but may deny that we can infer from this that 0 (El suffers from transworld depravity & E2 suffers from transworld depravity). But, as we've seen, such an inference is invalid only if it is D [If El suffers from transworld depravity, then ~ (E2 suffers from transworld depravity)]. But clearly such a conditional cannot be necessary. For if it were and El were to suffer from transworld depravity, it would not be up to E2 whether it suffers from transworld depravity.5 So, in the case at hand, from 0 (El suffers from transworld depravity) and 0 (E2

THE PROBLEM OF EVIL IN DEFENSE OF 'THE FREE WILL DEFENSE'

191 119

suffers from transworld depravity) we can validly infer that 0 (El suffers from transworld depravity & E2 suffers from transworld depravity). Perhaps the point I've just made can be seen more clearly if we focus on two actual persons, Capone and Gandhi. Could the fact, assuming it is a fact, that Capone is transworld depraved logically necessitate the fact, assuming it is a fact, that Gandhi is not transworld depraved? No. For then some fact external to Gandhi himself would necessitate what he would do with his freedom if he were created in certain circumstances. Gandhi is not transworld depraved only if for some Gandhi perfect world there are no circumstances included in it such that if those circumstances were actual, Gandhi would be significantly free and go wrong. And whether that is so must depend on how Gandhi would use his freedom were any of those circumstances in that world actual. If the matter of how he would use his freedom if any of those circumstances in that world were actual were itself logically necessitated by Capone's being transworld depraved, then it would not really be up to Gandhi how he would use his freedom if any of those circumstances in that Gandhi perfect world were actual. For in those circumstances it cannot be up to Gandhi whether or not Capone is transworld depraved. Now this may be all well and good. But what if it is not just two creaturely essences we are considering? What if there are many perhaps infinitely many essences we are considering? How can we then be confident that 0 (every essence suffers from transworld depravity)? To see why we can be confident, let's go back to our two drawings from the urn. We saw there that given the conditions necessarily prevailing in the second situation, from the fact that 0 (the first ball selected in not number 2) and 0 (the second ball selected is not number 2), we can validly infer 0 (the first ball selected is not number 2 & the second ball selected is not number 2). But what if we make many, even an infinite number of drawings? Can we still validly infer 0 (no selection is of ball number 2)? Indeed we can. Of course, in fact the number 2 ball will be selected if the selection process is random and goes on long enough. But although with each selection it becomes increasingly unlikely that no selection in the series is ball number 2, it always will remain 0 (no selection in the series is ball number 2). Similarly, no matter how many conjuncts of the form 'Ex suffers from transworld depravity' fall within the scope of the possibility operator, the addition of any number of other such conjuncts cannot make it the case that the resulting conjunction is not logically possible. That being so, we can conclude that 0 [Every (creaturely) essence suffers from transworld depravity].6

192

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 120

WILLIAM L. ROWE

Notes 1. A creaturely essence E is not transworld depraved provided some EPW is creatable by God. And a creaturely essence E is not transworld sanctified provided some EPW is not creatable by God. 2. See, for example, The Nature of Necessity (Oxford: Clarendon Press, 1974), p. 185, and of God, Freedom and Evil (New York: Harper & Row, 1974), p. 45. Concerning R, Plantinga says that to show that R is possible 'we must demonstrate the possibility that among the worlds God could not have actualized are all the worlds containing moral good but no moral evil'. 3. According to our authors: 'An essence E is blessed with transworld sanctity iff for every world W such that E contains the properties is significantly free in W and always does what is right in W, for no action A and for no maximal world segment S such that (1) S includes E's being instantiated and E's instantiation being free with respect to A and A's being morally significant for E's instantiation, and (2) S is included in W but includes neither E's instantiation performing A nor E's instantiation's refraining from A, it is the case that (3) if S were actual, then the instantiation of E would have gone wrong with respect to A.' 4. I take our authors to allow that we may have good reason to claim that each creaturely essence lacks transworld sanctity in some world or other, but to deny that we have good reason to think that there is some world in which every creaturely essence lacks transworld sanctity. Similarly, I take them to allow that we may have good reason to claim that each creaturely essence is transworld depraved in some world or other, but to deny that we have good reason to think that there is some world in which every creaturely essence is transworld depraved. In case I am mistaken in this, let me add that I think a convincing argument can be given for the view that each creaturely essence is such that in some world it is transworld depraved. In addition, as Bill Hasker has pointed out to me, it is relatively easy to show that each creaturely essence is such that in some world it is not transworld sanctified. 5. Unless, of course, it were up to E2 whether or not El suffers from TD. 6. An ancestor to this paper was delivered as comments at the 1997 meetings of the Society for Philosophy of Religion. I'm grateful to Mike Bergmann, Bill Hasker, and Scott Davidson for helpful comments on the ideas in this paper.

CHAPTER ELEVEN

Grounds for Belief Aside, Does Evil Make Atheism More Reasonable than Theism?

THE ISSUE The question assigned to us for discussion is this: "Grounds for belief in God aside, do the evils in our world make atheistic belief more reasonable than theistic belief?" The initial clause in this question is important. For it is one thing to argue that the evils in our world provide such compelling reasons for atheism that the reasons for the existence of God are insufficient to swing the pendulum back in favor of the existence of God, and another thing to argue that, putting aside whatever reasons there may be for believing that God exists, the evils that occur in our world make belief in atheism more reasonable than belief in theism. If we put aside grounds for belief in the existence of God, the likelihood that God exists cannot reasonably be assigned any probability beyond .5 - where 1 represents God's existence as certain, and 0 represents certainty that God does not exist. So, if we start from an initial point of God's existence having a likelihood of .5 or less, and restrict ourselves to the evidence generated by the enormous amount of horrendous evil that occurs daily in our world, it should strike anyone that the likelihood of God's existence can only go downward from .5.1 To reach such a judgment is perfectly consistent with holding that once the reasons supportive of the existence of God are brought into the equation the likelihood of God's existence is in fact positive, somewhere between .5 and 1. So, we should not confuse arguing that the negative evidence of evil shows God's existence to be unlikely, even taking into account the positive reasons there are to think that God exists, with

194

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE

125

arguing that putting aside the positive reasons there are to think that God exists, the evils that occur in our world make atheistic belief more reasonable than theistic belief. The issue in this discussion is only the latter: Apart from taking into account the positive reasons there are to think that God exists, do the evils that occur in our world make atheistic belief more reasonable than theistic belief? I shall argue that they do. Before proceeding to argue that point, however, it is important to be clear on what theism is. Theism is the view that there exists an allpowerful, all-knowing, perfectly good being (God). We can call this view restricted theism. It is restricted in that it does not include any claim that is not entailed by it.2 So, theism itself does not include any of the following claims: God delivered the ten commandments to Moses, Jesus was the incarnation of God, Muhammad ascended into heaven. These are claims made in specific theistic religions; thus they are a part of an expanded form of theism: Judaic theism, Christian theism, or Islamic theism. The importance of not taking theism to include the claims held by only one particular religion among the three major theistic religions of the West is that the inclusion would make theism less likely; for if we identify theism with a particular one among the great theistic religions, then the truth of theism itself is made to depend on all the essential beliefs of that particular theistic religion. The other side of this coin is that philosophers who wish to defend theism ought not to suppose that the assumption of theism entities them to assume any of the special claims associated with their own particular theistic religion. Since most of the philosophers in the Anglo-American tradition who defend theism are adherents of some version of Christian theism, they should beware of confusing the assumption that theism is true with the altogether different, and less likely assumption, that Christian theism is true.

THE ARGUMENT Do the evils that occur in our world significantly lower the likelihood of God's existence?3 Let's begin thinking about this problem by considering a simple argument from the existence of some of the evils in our world to the nonexistence of God.

THE PROBLEM OF EVIL 126

THE EVIDENTIAL PROBLEM OF EVIL

(1) There exist horrendous evils that an all-powerful, all-knowing, perfectly good being would have no justifying reason to permit. (2) An all-powerful, all-knowing, perfectly good being would not permit an evil unless he had a justifying reason to permit it. therefore, (3) God does not exist. If theists reject this argument for the nonexistence of God, they must either reject the first premise or the second premise. Most theists accept the second premise, as do nontheists. So, most theists must reject the first premise, holding instead that God has a justifying reason for permitting each and every horrendous evil that occurs. But what would be a justifying reason for God to permit some terrible evil he could prevent? Since an evil is something that by its very nature is bad, God's justifying reason for permitting it would have to include something else - either some outweighing good that, all things considered, he wishes to realize and cannot realize without permitting that evil,4 or some equal or worse evil that, all things considered, he wishes to prevent and cannot prevent without permitting that evil. And the question we must ask ourselves is whether it is rational for us to believe that all the terrible evils that occur daily in our world are like that? Is it rational to believe that each evil is such that were an all-powerful, allknowing being to prevent it, he would have to forfeit some outweighing good?5 Perhaps it will make the issue before us a bit more concrete if we focus on some examples of terrible evils, rather than just terrible evils in the abstract. Here are two examples: A fawn is horribly burned in a forest fire caused by lightning. It lies on the forest floor suffering terribly for five days before death relieves its suffering. A five-year-old girl is brutally beaten, raped and strangled in Flint, Michigan on New Year's eve a few years ago. The theist must believe that for each of these evils there is some greater good to which it leads, a good that an all-powerful being simply could not realize without permitting that evil. But is what the theist believes about these two evils really so? Is there really some great good that an all-powerful being could bring about only by permitting that fawn to be badly burned and to suffer intensely for five long days before death

195

196

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE 127

relieves its torment? And is there really some great good that an allpowerful being could bring about only if he permits that little five-yearold girl in Mint, Michigan to be savagely beaten, raped, and strangled? And even if it should somehow be so in these two cases, is it true that all the instances of intense human and animal suffering occurring daily in our world lead to greater goods in such a way that even an all-powerful, all-knowing being could not have achieved any of those goods without permitting the instances of suffering that supposedly lead to them? In light of our knowledge of the scale of human and animal suffering occurring daily in our world, the idea that none of those instances of suffering could have been prevented by an all-powerful being without the loss of a greater good must strike us as an extraordinary idea, quite beyond our belief. And if it does strike us in this way, the first premise of the argument we are considering - There exist evils that an all-powerful, all-knowing perfectly good being would have no justifying reason to permit - is bound to strike us as plausible, something quite likely to be true. But since the second premise is generally agreed to be true, we should then conclude that it is likely that our conclusion is true, that God does not exist. It is important here to understand two points about the argument just presented. First, the argument is not, nor is it meant to be, a proof that God does not exist. To be a proof of its conclusion an argument must be such that its conclusion logically follows from its premises and its premises are known with certainty to be true. The argument we are considering meets the first condition, but not the second. The conclusion deductively follows from the two premises, but its first premise is not known with certainty to be true. The claim is only that the first premise is one we are rationally justified in believing to be true. And since our confidence in the truth of the conclusion should not exceed our confidence in the premises from which it follows, the claim is only that the premises provide sufficient rational support for that conclusion. Second, the truth of the first premise does not logically depend on any claim about the two examples of the fawn and the five-year-old girl. The examples are meant to illustrate the profound difficulty in really believing that an all-powerful, all-knowing being is incapable of achieving his noble ends without having to permit such horrendous, undeserved suffering. But if there were only a few such examples as these, perhaps it would not be unreasonable to believe that somehow even an infinitely intelligent, all powerful being could not achieve his good ends without permitting them. But, of course, our world is not like

THE PROBLEM OF EVIL 128

THE EVIDENTIAL PROBLEM OF EVIL

that. It is the enormous amount of apparently pointless, horrendous suffering occurring daily in our world that grounds the claim in the first premise that there are pointless evils in our world, evils that an allpowerful being could have prevented without forfeiting some outweighing good. But, again, it is not being asserted that the existence of pointless evils is known with certainty, only that it is quite likely that pointless evils occur.

EVALUATING Two RESPONSES

/ Having looked at a particular argument from evil against theistic belief, we can now consider and critically evaluate two theistic responses to this argument. The first response the theist may put forth goes something like this. "The first point I want to make is that thus far we have been given no reason at all to think that premise (1) is true. For all you have pointed out is that we don't know what the good is that justifies God in permitting any of these horrendous evils, like the fawn's suffering or the little girl's suffering. But to argue from the fact that we don't know what the good is that justifies God in permitting a certain evil to the conclusion that there is no such good is to engage in a fallacious argument from ignorance: we don't know of any justifying good, therefore there isn't any. So, you haven't really given any good reason at all to think that there are terrible evils for which there are no God-justifying goods. All that you have shown, if you have shown anything, is that if these evils do serve some God-justifying goods, we don't know what they are. And the interesting question to ask about our ignorance of these justifying goods is this: Given that God's mind infinitely transcends ours, is it really at all likely that the goods for the sake of which he permits much horrendous suffering will be goods we comprehend? After all, isn't God in relation to us like good, loving parents in relation to their small child? Such parents may permit their very young child to suffer a painful surgical procedure for a good the child simply cannot comprehend. So too, we should expect that if God exists he may permit many instances of human or animal suffering so as to realize goods our minds simply cannot comprehend. And if that is so, the fact that we don't know the goods that justify God in permitting much horrendous suffering cannot really be a

197

198

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE

129

reason for thinking he doesn't exist. For it is just what we should expect to be true if he does exist."6 What are we to make of this response by the theist? Are we really just arguing from ignorance? Perhaps we can come to see that we are not by first distinguishing between goods we know about (goods within our ken) and goods beyond our ken. Consider the suffering of the five-yearold girl as she was brutally beaten, raped, and strangled on New Year's Eve a few years ago in Flint, Michigan. I believe that no good we know about justifies God in permitting that suffering. By "goods we know about" I mean goods that we have some cognitive grasp of, even though we may have no knowledge at all that they have occurred or ever will occur. For example, consider the good of the little girl experiencing complete felicity in the everlasting presence of God. Theists consider this an enormous personal good, perhaps the greatest personal good possible for the little girl. So, even though we don't have a very clear grasp of what this great good involves, and even though we don't know that such a good will ever be actualized, I include the good of her experiencing complete felicity in the everlasting presence of God among the goods we know about Of course, if some good we know about does justify God in permitting her suffering, that good must have already been actualized or be actualized at some point in the future. But the notion of a good we know about extends to many future goods and to goods that never have and never will occur. And what we have good reason to believe is that none of the goods we know about justifies God in permitting the horrendous suffering of that little girl. For with respect to each such good we consider, we have reason to believe either that it isn't good enough to justify God in permitting that evil, or that it could likely be actualized by God without his having to permit the horrendous suffering of that little girl, or that some equal or better good could likely be actualized by God without his having to permit the horrendous suffering of that little girl. Of course, even granting that we know of many great goods and have reason to think that none of these goods justifies God in permitting the little girl's suffering, there still remains the possibility that some good we cannot even conceive does so. And it is here that the theist may appeal to the analogy between the good parent and God. For we cannot deny that some good the child's mind cannot even conceive may justify the parents in permitting the child to suffer. And by analogy won't the same be true of God in relation to us as his children? Indeed,

THE PROBLEM OF EVIL 130

THE EVIDENTIAL PROBLEM OF EVIL

since the disparity between his mind and ours may greatly exceed that of the good parents' minds to the mind of their child, isn't it likely that the goods that justify him in permitting us to suffer will often be beyond our comprehension? But against this argument from analogy, two points need to be made. First, although arguments from analogy are rather weak, the analogy in question has some merit if drawn between a good parent and a good deity of considerable but nevertheless finite power and knowledge. For, like the good parent, a deity with great but finite powers may reasonably believe that he cannot realize some important future good for some of his creatures without permitting a present evil to befall them. And there may be occasions when, like the good parent, the finite deity is simply unable to prevent a dreadful evil befalling his creatures even though there is no good at all served by it. But the theistic God has unlimited power and knowledge. A good parent may be unable to prevent some suffering her child undergoes, or even the child's death from some painful disease. Can we seriously think that an infinitely powerful, allknowing deity was powerless to prevent the horror of Auschwitz? A good parent may see that she cannot realize some important future good for her child without permitting some present evil to befall the child. Can we seriously think that there is some far off future good for the victims of Auschwitz, a good that a deity of infinite power and knowledge judged to be worth the horror of Auschwitz, and was powerless to achieve without permitting that horror? Perhaps we can if we turn from reason to faith. But the infinite distance between the God of traditional theism and the good mother with the sick child doesn't, in my judgment, provide human reason with good grounds for thinking that such a being would be powerless to prevent many of the countless, seemingly pointless horrors in our world without losing some goods so distant from us that even the mere conception of them must elude our grasp. But suppose we do reason from the good-parent analogy to the behavior of an all-powerful, all-knowing, infinitely good deity. I think we shall see that the good-parent analogy leads in a different direction from what its proposers desire. We know that when a good, loving parent permits her child to suffer severely in the present for some outweighing good the child cannot comprehend, the loving parent then makes every effort to be consciously present to the child during its period of suffering, giving special assurances of her love, concern, and care. For the child may believe that the parent could prevent her present suffering. So, of course, the parent will be particularly careful

199

200

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE 131

to give her child special assurances of her love and concern during this period of permitted suffering for a distant good the child does not understand. And indeed, what we know about good, loving parents, especially when they permit their children to suffer intensely for goods the children cannot comprehend, is that the parents are almost always consciously present to their children during the period of their suffering, giving special assurances of their love and care. So, on the basis of the good-parent analogy, we should infer that it is likely that God too will almost always be consciously present to humans, if not other animals, when he permits them to suffer for goods they cannot comprehend, giving special assurances of his love for them. But since countless numbers of human beings undergo prolonged, horrendous suffering without being consciously aware of God's presence or any special assurances of his love and comfort, we can reasonably infer either that God does not exist or that the good parent analogy is unable to help us understand why God permits all the horrendous suffering that occurs daily in our world. Our conclusion about the theist's first response is this. The argument in support of premise (1) is not an argument from ignorance. It is an argument from our knowledge of many goods and our reasonable judgment that none of them justifies God in permitting instances of horrendous evil. It is also an argument from our knowledge of what a being of infinite power, intelligence, and goodness would be disposed to do and would be capable of doing. Of course, there remains the logical possibility both that some goods incomprehensible to us justify God in permitting all these horrendous evils that occur daily in our world and that some further goods incomprehensible to us justify God in not being consciously present to so many who endure these horrendous evils. So, we cannot prove that premise (1) is true. Nevertheless, the first response of the theist should, I believe, be judged insufficient to defeat our reasons for thinking that premise (1) is probably true. Before turning to the theist's second response, we should note that some theists will protest the conclusion we've come to about the first response. Here is what such a theist may say: "Your distinction between goods we know about and goods beyond our ken is well-taken. Moreover, you are right to insist that your argument is not a flagrant example of an argument from ignorance. But there is one quite important point you have failed to establish. It is crucial to your

THE PROBLEM OF EVIL 132

THE EVIDENTIAL PROBLEM OF EVIL

argument that we should expect to know the goods for the sake of which God permits much terrible suffering or, failing such knowledge, be particularly aware of God's presence and his love for us during the period of intense suffering for goods we cannot comprehend. For if we have no good reason to expect to know these goods, or to experience God's presence and love during our suffering, then the fact that we don't know them and don't experience God's presence and love won't really count against the existence of God. And my point is that God may have good reasons (unknown to us) for not revealing these goods to us. And he also may have good reasons (unknown to us) for not disclosing himself and his love during the period when many suffer terribly for goods they cannot comprehend. How are you able to show that this point of mine is just a mere logical possibility and not the way things really are? I think you need to treat more seriously than you do the distinct possibility that God's reasons for permitting so much horrendous suffering, and his reasons for not being consciously present to us during our suffering, involve goods that are presently incomprehensible to us." The theist here raises an important point. Using the theist's own good-parent analogy, I argued that there is reason to think that when we don't know the goods for the sake of which God permits some horrendous suffering, it is probable that, like the good parent, he would provide us, his children, with special assurances of his love and concern. Since many endure horrendous suffering without any such special assurances, I suggested that we have further reason to doubt God's existence. And the theist's only reply can be that there are still further unknown goods that justify God in not being consciously present to us when we endure terrible suffering for goods beyond our ken. And I've allowed that we cannot prove that this isn't so. It remains a logical possibility. I've said, however, that we can conclude that premise (1) is probably true. But the theist says that I'm not justified in concluding that premise (1) is probably true unless I give a reason to think it likely that there are no unknown goods that justify God in permitting much horrendous suffering or no unknown goods that justify God in not being present to us when we endure suffering for unknown goods. The theist may grant me that no goods we know of play this justifying role. But before allowing me to conclude that it is probable that premise (1) is true and, therefore, probable that God does not exist, the theist says I must also provide some grounds for thinking that no unknown goods play that justifying role.

201

202

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE 133

Suppose we are unsure whether Smith will be in town this evening. It is just as likely, say, that he will be out of town this evening as that he will be in town. Suppose, however, that we do know that if Smith is in town it is just as likely that he will be at the concert this evening as that he won't be. Later we discover that he is not at the concert. I conclude that given this further information (that he is not at the concert) it is now less likely that he's in town than that he's out of town, that given our information that he is not at the concert, it is more likely that he is out of town than that he is in town. I do admit, however, that I haven't done anything to show that he is not actually somewhere else in town. All I've established is that he is not at the concert. I acknowledge that it is logically possible that he's somewhere else in town. Nor do I know for certain that he is not somewhere else in town. All I claim is that it is probable that he is not in town, that it is more likely that he is not in town than that he is in town. Those who want to believe that Smith is in town may say that I'm not justified in concluding that it is probable that he's out of town unless I give some reason to think that he is not somewhere else in town. For, they may say, all I've done is exclude one of the places he will be if he is in town. Similarly, the theist says that if God exists then either all the horrendous evils we consider serve unknown goods or some of them serve goods we know of. We might even agree that if God exists it is equally likely that some of the justifying goods will be known to us as that all of the justifying goods will be beyond our ken. After all, when we understand why God may be permitting some terrible evils to occur, those evils will be easier to bear than if we haven't a clue as to why God is permitting them to occur. Suppose we then consider the goods we know of and reasonably conclude that none of them justifies God in permitting any of these horrendous evils that abound in our world. The theist may even agree that this is true. I then say that it is probable God does not exist. The theist says I'm not justified in drawing this conclusion unless I give some reason to think that no unknown goods justify God in permitting all these terrible evils. For, he says, all I've done is exclude one sort of good (goods known by us) as God's justification for permitting any of these terrible evils. Who is right here? Let's go back to the claim that it is probable that Smith is not in town this evening. How can we be justified in making that claim if we've learned only that he is not at the concert? The reason is this. We originally knew that it was equally likely that he would be out of town as in town. We also agreed that jf he is in town it is equally likely

THE PROBLEM OF EVIL 134

THE EVIDENTIAL PROBLEM OF EVIL

that he will be at the concert as that he won't be. Once we learn he is not at the concert, the likelihood that he is out of town must increase, as does the likelihood that he is somewhere else in town. But since it was equally likely that he is out of town as in town, if the likelihood that he is out of town goes up it then becomes greater than .5, with the result that it is probable that he is not in town. Turn now to the existence of God and the occurrence of horrendous evils. Either God exists or he does not. Suppose for the moment that like the case of Smith being or not being in town, each of these (God exists, God does not exist) is equally likely on the information we have prior to considering the problem of evil.7 Consider again the many horrendous evils that we know to occur in our world. Before we examine these evils and consider what sort of goods (known or unknown) justifies God (if he exists) in permitting them, suppose it is as likely that the justifying goods for some of these evils are known to us as that the justifying goods for all of these evils are unknown to us. We then examine the known goods and those horrendous evils and come to the conclusion that no known good justifies God in permitting any of those horrendous evils. That discovery parallels our discovery that Smith is not at the concert. And the result is just the same: it is then more probable than not that God does not exist.

// The second response the theist can give to the challenge of the problem of evil is the following: "It is a mistake to think that the goods for which God permits these horrendous evils are totally beyond our ken. For religious thinkers have developed very plausible theodicies that suggest a variety of goods that may well constitute God's reasons for permitting many of the horrendous evils that affect human and animal existence. When we seriously consider these theodicies we can see that we have good reason to think that premise (1) is false. For these theodicies provide us with plausible accounts of what may be God's justifying reasons for permitting the evils that occur in our world." The theist's first response was to argue both that we have given no reason at all for thinking that premise (1) is true and that our ignorance

203

204

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE 135

of many goods that God's mind can comprehend prevents us from being able to establish that premise (1) is probably true. In the second response, the theist proposes to give a good reason for thinking that premise (1) is false. And, of course, to the extent that theodicies do provide a good reason for rejecting premise (1), to that extent the theist will have pointed the way to reconciling the existence of God with the fact that our world contains the horrendous evils that it does. But do these theodicies really succeed in providing a good reason for rejecting premise (1)? I believe they do not. But to demonstrate this we would have to show that these theodicies, taken together, are really unsuccessful in providing what could be God's reasons for permitting the horrendous evils in the world. Although I believe this can be done, I propose here to take just one of these theodicies, the one most commonly appealed to, and show how it fails to provide a good reason for rejecting premise (1). I refer to the free will theodicy, a theodicy that has played a central role in defense of theism in the theistic religions of the West. Developed extensively by St. Augustine (AD 354-430), the free will theodicy proposes to explain all the evils in the world as either directly due to evil acts of human free will or to divine punishment for evil acts of human free will. The basic idea is that rather than create humans so that they behave like automatons, acting rightly of necessity, God created beings who have the power to act well or ill, free either to pursue the good and thereby enjoy God's eternal blessing or to pursue the bad and thereby experience God's punishment. As things turned out, many humans used their free will to turn away from God, freely choosing to do ill rather than good, rejecting God's purpose for their lives. Thus, the evils in the world that are not bad acts of human free will, or their causal effects, are due to God's own acts of punishment for wrongful exercises of human free will. The cornerstone of this theodicy is that human free will is a good of such enormous value that God is justified in creating humans with free will even if, as Augustine held, God knew in advance of creating them that certain human beings would use their freedom to do ill rather than good, while knowing that others would use their freedom to do only (or mostly) what is good. So, all the horrendous evils occurring daily in our world are either evil acts of free human beings and their causal effects or divine punishments for those acts. And the implication of this theodicy is that the good of human free will justifies God in permitting all these horrendous acts of evil and their causal effects, as well as the

THE PROBLEM OF EVIL 136

THE EVIDENTIAL PROBLEM OF EVIL

other evils resulting from plagues, floods, hurricanes, etc., that are God's ways of punishing us for our evil acts. While this theodicy may explain some of the evil in our world, it cannot account for the massive amount of human suffering that is not due to human acts of free will. Natural disasters (floods, earthquakes, hurricanes, etc.) bring about enormous amounts of human and animal suffering. But it is obvious that such suffering is not proportionate to the abuses of free will by humans. So, we cannot reasonably think that such disasters are God's way of punishing human free choices to do evil. Second, while being free to do evil may be essential to genuine freedom, no responsible person thinks that the good of human freedom is so great as to require that no steps be taken to prevent some of the more flagrant abuses of free choice that result in massive, undeserved suffering by humans and animals. Any moral person who had power to do so would have intervened to prevent the evil free choices that resulted in the torture and death of six million Jews in the Holocaust. We commonly act to restrict egregious abuses of human freedom that result in massive, undeserved human and animal suffering. Any moral being, including God, if he exists, would likely do the same. And since the free will theodicy is representative of the other attempts to justify God's permission of the horrendous evils in our world, it is reasonably clear that these evils cannot be explained away by appeal to theodicies. In this essay I have argued that, putting aside whatever reasons there may be to think that the theistic God exists, the facts about evil in our world provide good reason to think that God does not exist. While the argument is only one of probability, it provides a sound basis for an affirmative answer to the question that is the focus of this exchange.

NOTES TO "GROUNDS FOR BELIEF ASIDE'5 1 At best it can but remain the same. For no reasonable person would argue that all the horrendous evils that occur daily in our world are to be counted as evidence for the existence of God. 2 Theism itself does not include the claim that God created a world. For theists hold that God was free not to create a world. They hold that there is a possible world in which God exists but creates nothing at all. What theism may be taken to include is the claim that any contingent things that exist depend for their existence on God's creative act.

205

206

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE

137

3 Portions of the following are drawn from my essay "God and Evil," The Annual Proceedings of the Centerfor Philosophic Exchange, No. 28, 1997-8, 4-15. 4 It could be that the outweighing good cannot be realized by God without his permitting that evil or some other evil just as bad. But for ease of understanding the fundamental issue I will ignore this complication. 5 To avoid needless complexity, I'll not mention the other possibility; that God permits the evil in question so as to prevent some equal or greater evil. 6 This response has been elegantly developed by Stephen Wykstra in "The Human Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of 'Appearance,5" International Journal for Philosophy of Religion 16 (1984): 73-93. Also see "Evil and the Theistic Hypothesis: A response to Wykstra," International Journal for Philosophy of Religion 16 (1984): 95-100. 7 As we noted earlier, given that we are putting aside reasons for the existence of God, the existence of God is, at best, no more likely than is the nonexistence of God.

CHAPTER TWELVE

Friendly Atheism, Skeptical Theism, and the Problem of Evil

In 1979, a quarter of a century ago, I published a paper entitled "The Problem of Evil and Some Varieties of Atheism."1 It received a good deal of attention in the professional journals, and was frequently included in anthologies for use in the classroom. Indeed, nothing I've written before or since has received anything near the attention that was given to that paper. So, in that respect, my career as a philosopher has been downhill ever since. In that paper I focused on a particular example of evil: a fawn being horribly burned in a fire caused by lightning, and suffering terribly for 5 days before death ended its life. Unlike humans, fawns are not credited with free will, and so the fawn's suffering cannot be attributed to a misuse of free will. Why then would God permit it when, if he exists, he could have so easily prevented it? It is generally admitted that we are simply unable to imagine any greater good whose realization can reasonably be thought to require God to permit that fawn's terrible suffering. And it hardly seems reasonable to suppose there is some greater evil that God would have been unable to prevent had he not permitted that fawn's terrible suffering. Suppose that by a "pointless evil" we mean an evil that God (if he exists) could have prevented without thereby losing an outweighing good or having to permit an evil equally bad or worse. Is the fawn's suffering a pointless evil? Clearly, it certainly seems to us to be pointless. On that point there appears to be near universal agreement. For given God's omniscience and absolute power it would be child's play for him to have prevented either the fire or the fawn's being caught in the fire. Moreover, as we've noted it is extraordinarily difficult to think of, or even imagine, a greater good whose realization can sensibly be thought to require God to permit that fawn's

208

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 80

WILLIAM L. ROWE

terrible suffering.2 And, it is just as difficult to imagine an equal or even worse evil that God would be required to permit were he to have prevented the fawn's suffering. It therefore seems altogether reasonable for us to think that the fawn's suffering is likely to be a pointless evil, an evil that God (if he exists) could have prevented without thereby losing some outweighing good or having to permit some other evil just as bad or worse. In light of such examples of horrendous evils, evils that occur all too frequently in our world, I proposed an argument similar to this: 1. Probably, there are pointless evils, (e.g., the fawn's suffering). 2. If God exists, there are no pointless evils, therefore, 3. Probably, God does not exist. How can a theist respond to this argument? Since the conclusion logically follows from its two premises, and since the second premise is generally admitted to be not only true but necessarily true, theists are limited, I believe, to basically three different responses. The first is the response that I, rather naively, thought is the most reasonable response for the theist to give. It consists in simply accepting the argument as showing that we have a reason to think it unlikely that God exists. For given that no good we can imagine can be reasonably thought to justify God in permitting that fawn's terrible suffering, it does seem unlikely that there is a good that in fact does justify God in permitting that suffering. After all, we can think of many goods, even the greatest good of all - life eternal in the loving presence of God. And none of the goods we can think of appears to require the fawn's terrible suffering as a condition of its realization.3 Of course, it remains possible that some good that is unknowable by us justifies God in permitting the fawn's suffering and death. So, even though the argument may make it somewhat unlikely that God exists, if, as some theists will surely claim, we have stronger reasons to think he does exist - for example, a personal experience of an all powerful, all-knowing, perfectly good creator - we will be justified in believing that some good unknown to us does justify God in permitting the fawn's 5 days of terrible suffering. On the other hand, if someone has no outweighing reason to believe that God exists, the fawn's suffering, along with numerous other instances of seemingly pointless evils, may well justify such a person in thinking that it is unlikely that there is any such being as God.

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL 81

In a relatively short time, however, I came to see that I was mistaken in thinking that philosophers, whether theists or nontheists, would agree with me in accepting the first premise of the argument, the premise that says it is likely that pointless evils occur. Indeed, I came to see this as early as 1982, just 3 years later, when I conducted an NEH Summer Seminar for College Teachers. I had a group of young, able philosophers who held teaching positions in various colleges. We covered several topics during the 6 weeks they were at Purdue, and toward the end we spent a week on the problem of evil. Among the group was a chap named Stephen Wykstra who had accepted a teaching position in philosophy at Calvin college. Wykstra talked only occasionally in the seminar, but when he became excited about some point or argument he would talk a good deal, sometimes having difficulty stopping talking, even after having fully made his point. At such times he would finally become aware that he had gone on too long, stop for moment, and then say, "Shut up Wykstra!" And when he said that, to our surprise he would stop talking. When the 6 weeks were up, many of those in the seminar departed. But Wykstra remained at Purdue for about two additional weeks, coming in to see me and discussing further the problem of evil. I have only dim recollections of those discussions, but I rightly sensed that Wykstra was very focused on providing a philosophical critique of my argument from evil. I also correctly sensed he would not rest until he had done so. A year later, he presented an important paper on the problem of evil at the Pacific Division of the American Philosophical Association, to which I was invited to respond. I recall that meeting of the Pacific Division very well. For it took place on board the Queen Mary, and each person's private room on the vessel was about the size of a rather spacious coat closet. I don't recall much at all of what went on at the session with Wykstra. I do recall, however, what was then referred to as 'the smoker', an evening session at all APA meetings in which the main activity is sipping wine and talking with one's friends in the profession. My recollection is that I felt it altogether proper and good to be on the Queen Mary, drinking wine while talking with other philosophers. And it did not seem to matter that the Queen Mary was going nowhere at all, being permanently in dock. In 1984 Wykstra's paper was published in the International Journal for Philosophy of Religion, along with my reply.4 The publication of his paper significantly advanced a position in philosophy that has come to be known as 'skeptical theism.' For Wykstra presented a skeptical argument against the justification I gave for the crucial

209

210

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 82

WILLIAM L. ROWE

premise at work in my argument from evil, the premise stating that probably there are pointless evils. Why did I believe that the fawn's suffering is likely to be pointless? It was because we cannot think of or even imagine a good that would both outweigh the fawn's suffering and be such that an all-powerful, all-knowing being could not find some way of bringing about that good, or some equal or better good, without having to permit the fawn's suffering. For think for a moment of the fawn's suffering. It is not only terribly burned, but it lies for 5 days on the forest floor in agony, before death finally ends its life. Is there some great good that an all-powerful, all-knowing being could bring about only by allowing that fawn to suffer for five full days, rather than say, four, three, two, one, or even not at all - say, by bringing it about that its death is instantaneous? It baffles the human mind to think that an all-powerful, all-knowing being would find itself in such a predicament. But the skeptical theist's response is that, for all we know, the reason why the human mind is baffled by this state of affairs is simply because it doesn't know enough. For if we were to know what God knows then our human minds would know that God really had no choice at all. For, according to the skeptical theist, God knew that if he prevented that fawn's being terribly burned, or prevented even 1 day of the fawn's 5 days of terrible suffering, he either would have to permit some other evil equally bad or worse or forfeit some great good, without which the world as a whole would be worse than it is by virtue of his permitting that fawn to suffer intensely for five full days. And the fact that we have no idea of what that good might be is not at all surprising, given the disparity between the goods knowable by our minds and the goods knowable by a perfectly good and all-knowing creator of the world. So, according to the skeptical theist, we simply are in no position to reasonably judge that God, if he exists, could have prevented the fawn's 5 days of terrible suffering without losing some outweighing good or having to permit some equally bad or worse evil. For our limited minds are simply unable to think of the goods that the mind of God would know. And since we are simply unable to know many of the goods God would know, the fact that no good we know of can reasonably be thought to justify an infinitely good, all-powerful being in permitting the fawn's terrible suffering is not really surprising. In fact, given the enormous gulf between God's knowledge and our knowledge, that no good we know of appears to in any way justify God in permitting the fawn's terrible suffering is perhaps just what we should expect if such a being as God actually exists.

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL

211 83

In developing his view Wykstra argued that to reasonably believe that the fawn's suffering is likely to have been pointless, we must have a positive reason to think that if some good should justify God's permitting the fawn's suffering it is likely that we would know of that good. He then claimed that goods knowable to God are quite likely not going to be knowable to us. To support his claim Wykstra pointed out that upon looking in his garage and seeing no dog, we would be entitled to conclude that there is no dog in the garage. But upon looking in his garage and seeing no fleas, we would not be entitled to conclude that there are no fleas in his garage. For we have reason to think that if there were any fleas in his garage, it would not be likely that we would see them. And similarly, he argued, our not being able to think of a good that might justify God in permitting the fawn's suffering is no reason to think there isn't such a good. For, on Wykstra's view, were there such a God-purposed good for permitting the fawn's suffering it is altogether likely that we would not know of it. So, the fact that we cannot even imagine what such a good would be, far from being a reason to think it unlikely that God exists, is just what we should expect to be true if God does exist. Toward the end of his paper Wykstra notes that I am right in holding that a wholly good God would allow suffering, such as the fawn's terrible suffering, only if "there is an outweighing good served by so doing." He also agrees with me "that such goods are, in many cases, nowhere within our ken." But he then says: The linchpin of my critique has been that if theism is true, this is just what one would expect: for if we think carefully about the sort of being theism proposes for our belief, it is entirely expectable - given what we know of our cognitive limits - that the goods by virtue of which this Being allows known suffering should very often be beyond our ken. Since this state of affairs is just what one should expect if theism were true, how can its obtaining be evidence against theism? (p. 91)

n Wykstra's elegant and carefully argued paper forced me to make some distinctions that are all too easily overlooked in responses to objections to the claim that the theistic God exists.5 In making these distinctions, I let the capital letter 6O' abbreviate 'an omnipotent,

212

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 84

WILLIAM L. ROWE

omniscient, wholly good being'. I then suggested that standard theism is "any view which holds that O exists." Thus, traditional Christianity, Islam, and Judaism are all examples of standard theism. However, within standard theism we can distinguish restricted standard theism and expanded standard theism. Expanded theism is the view that O exists, conjoined with certain other significant religious claims, claims that are not entailed by the proposition that there exists an omnipotent, omniscient, wholly good being who has created the world. For that proposition is common to traditional Christianity, Islam, and Judaism. What are some of the independent religious ideas that have been added to standard theism so as to produce the version of expanded theism that we now know as Christianity? There is, of course, the idea that God is a trinity: God the Father, God the Son, and God the Holy Ghost, an idea that took several centuries to be worked out, and is held only in Christianity. It was first established at the Council of Nicea in 325 AD when Athanasius, accused of heresy and exiled several times in his lifetime, won out over Arius who denied that Jesus of Nazareth was a divine being. In addition, of course, there are claims about original sin, Adam and Eve in the garden, redemption, a future life, a last judgment, and the like. Orthodox Christian theism is a version of expanded theism, for the doctrine that God took on human form and died for our sins is essential to orthodox Christianity, but not deducible merely from the claim that the theistic God exists. Since some might be led to think that if the theistic God exists, then certain other logically independent religious claims must also be true - claims about sin, redemption, heaven, the divinity of the son of Mary and Joseph, etc. - in my reply to Wykstra I insisted on using the expression 'O exists' rather than the expression 'God exists'. I did this to assist the reader in recognizing that from the assumption that there exists an omnipotent, omniscient, wholly good being who created the world nothing can be logically deduced concerning whether certain other religious claims held by Judaism, Islam or Christianity are true or false.6 And this means, of course, that from the fact that O exists we cannot logically deduce that there is a life beyond our three score years and ten, that there is a heaven or a hell, that Jesus of Nazareth is the Son of God, or that the Bible is divinely inspired. With this in mind I noted that in answer to my claim that it appears that the fawn's suffering is pointless, Wykstra's principle response is that we have no reason at all to think that were O to exist things would strike us any differently. And we can formulate his argument for this point as follows:

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL

213 85

4. O's mind grasps goods beyond our ken. 5. It is likely that the goods in relation to which O permits many sufferings are beyond our ken. therefore, 6. It is likely that many of the sufferings in our world do not appear to have a point - we can't see what goods justify O in permitting them. I then pointed out that the fact that O's mind grasps goods beyond our ken does not entail that it is likely that the goods in relation to which O permits many sufferings are beyond our ken. For once such goods occur we have reason to think that we would know them, given that they are goods involving those who have suffered. Moreover, the mere fact that O exists gives us no reason to think that the goods for the sake of which O permits horrendous human and animal suffering are goods that occur only in some far distant future, perhaps in some other form of existence altogether unknown and perhaps unknowable by us. But what about expanded theism? Reading the writings attributed to Saint Paul, one may conclude that the goods for the sake of which O permits vast amounts of human and animal suffering will be realized only at the end of the world, or in some state of existence quite unknown to us. If we conjoin this claim with the horrendous suffering that occurs in our world, we do seem to have some reason to think just what Wykstra claims: that it is likely that the goods in question would be beyond our ken. But as we've noted, the mere fact that O exists gives us no reason to think that what we find in the Bible or the Koran is anything more than what it probably is, the writings of various human beings scattered over time who, for whatever reasons (if any) came to believe that a divine being exists and is the creator of everything else, and that this being has a plan for his creatures (at least, human beings) and will provide them with some sort of existence after bodily death, etc. In his paper Wykstra claimed that the theistic hypothesis "contains" the claim that the goods for the sake of which O permits the sufferings in the world are, to a large extent, quite beyond our ken. He speaks of it as a "logical extension of theism." He says this claim is not an "additional postulate" but instead was "implicit" in theism all along. In my reply I said that Wykstra is mistaken about this. What is implicit in theism is that O's mind grasps goods that are beyond our ken. That does seem to be a "logical extension" implicit in the-

214

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 86

WILLIAM L. ROWE

ism. For O is omniscient, and we quite clearly are not. However, the claim that these goods are realizable only in a world beyond the world of our earthly existence is not a part of the hypothesis that O exists. It is an additional postulate that produces a form of expanded theism, a version that is not rendered unlikely by the facts about suffering that I claim to render restricted standard theism unlikely. Indeed, I now rather suspect that Wykstra was supposing that the Bible, or at least much of the New Testament, is somehow guaranteed to be true by virtue of the assumption that O exists. And that is a supposition that I, along with a number of biblical scholars, am unwilling to concede. Wykstra noted in his essay that among believers, as well as nonbelievers, there is a "persistent intuition that the inscrutable suffering in our world in some sense disconfirms theism." This observation by Wykstra strikes me as exactly right. It is not just unbelievers who tend to see the inscrutable suffering in our world as in some sense disconfirming theism; it is a near universal, natural phenomenon. Believers too, as Wykstra notes, have a strong, natural tendency to see inscrutable suffering, especially as it affects those they dearly love, as an intellectual difficulty or obstacle to belief, something that in the absence of a sensible explanation tends to count against theism. He, nevertheless, thinks that this persistent intuition of believers and nonbelievers is a mistake. For given our cognitive limitations and O's omniscience and omnipotence, Wykstra believes that it should be expected that much of the suffering in our world will be inscrutable to us. So, he concludes that believers and nonbelievers simply fail to see what is really contained in the theistic hypothesis. But if I am right, what Wykstra has unwittingly done is change the question. He has supplemented the theistic hypothesis that O (an omnipotent, omniscient, wholly good, perfectly loving being) exists with other propositions such that the supplemented result is not discontinued by the facts that are claimed to count against the hypothesis that O exists. I ended my response, however, by pointing out that Wykstra would likely disagree with me about whether he has really supplemented the hypothesis that O exists. For Wykstra seems to think that what I hold to be supplements to the hypothesis that O exists are in some way already logically contained in that very hypothesis. But clearly, they are not. An analogy to which theists often appeal in defending the reasonableness of supposing that the goods justifying the horrendous evils in our world are unknowable by us is the good-parent analogy. The

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL

215 87

idea is that God is to us humans as good parents are to their children whom they love. And just as their children often cannot comprehend the goods for which their loving parents permit things to happen to them, so too we humans cannot comprehend the goods for which God permits us, his created children, to endure the evils that happen to us. My own judgment is that this analogy, very much favored by theists, is actually unfavorable to theism. It is true that good, loving parents may have to permit their ailing child to be separated from them, confined to a hospital, forced to swallow evil tasting medicines, and put in the care of strangers in order to cure the child of some illness. The very young child, of course, may not understand why his parents have removed him from his home and put him in the care of strangers. So too, the theist may say, our sin, or something else, may have separated us from God. But the analogy is a dismal failure. When children are ill and confined to a hospital, the loving parents by any means possible seek to comfort their child, giving special assurances of their love while he is separated from them and suffering for a reason he does not understand. No loving parents use their child's stay in the hospital as an occasion to take a holiday, saying to themselves that the doctors and nurses will surely look after little Johnny while they are away. But many human beings have endured horrendous suffering without any awareness of God's assurances of his love and concern during their period of suffering. If you are in doubt of this, try reading the literature concerning the holocaust victims. Unlike the good parents, God has been on holiday for centuries. Indeed, even in the llth century, during the age of faith, the great Christian saint, St. Anselm of Canterbury, lamented: I have never seen thee, O Lord my God;... What O most high Lord, shall this man do, an exile far from thee? He longs to come to thee, and thy dwelling place is inaccessible.... I was created to see thee and not yet have I done that for which I was made.7 In an age of faith, before the growth of the scientific knowledge that produced alternative, credible explanations of the emergence of human life, Anselm's lament, quite reasonably, would not lead to disbelief. But the age or faith has been replaced by an age of reason and science. And in this age of reason and science, for many human beings the idea of God no longer plays an essential, rational role in explaining the world and human existence. The idea that human suffering may be divine punishment for human sin and wickedness is

216

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 88

WILLIAM L. ROWE

no longer a credible explanation for many educated human beings. My own inclination is to think that given the horrendous evils in our world, the absence of the God who supposedly walked with Adam and Eve in the garden is evidence that there is no God. For surely, if there were a God he would wish to provide us with strong reasons to think that he exists, given that the horrendous evils in our world, both natural and moral, seem to provide us with reason to doubt his existence. Of course, one can come up with elaborate stories in which God's hiddenness, even given all the horrendous evils that occur in our world, is not altogether implausible. John Hick and Richard Swinburne, for example, have endeavored to provide such stories. They are called theodicies. Agnostics and atheists tend to view these efforts at explaining God's permission of evil as rather unconvincing. And skeptical theists, following Alvin Plantinga, tend to dismiss theodicies as unnecessary, weak, and unpersuasive. My own judgment is that the theodicies provided by Swinburne and Hick do go some way toward reconciling theism with the horrendous evils that afflict us. But given the enormity of evil in our world, both natural and moral, coupled with the inexplicable absence of the God of traditional theism, I continue to think that the horrendous evils in our world provide evidence against the existence of God.

ra Of course, believing, as I and many others do, that the horrendous evils in our world count against the existence of the theistic God is not the same as being rationally justified in holding that belief. And I must confess that my earlier confidence in the inference from no good we know 0/justifies God in permitting many of the evils in our world to the conclusion probably no good justifies God in permitting those evils has been somewhat diminished by the objections raised by skeptical theists.8 The crucial objection is this: In order to have confidence in the inference, it seems that I must suppose that the goods we know of are representative of the goods there are.9 But given that God, if he exists, would likely know of goods beyond our wildest dreams,

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL

217 89

why should we think that the goods we know of are a representative sample of the goods there are? Of course, we do know of some very great goods, even the greatest good of all - life eternal in the loving presence of God. But that isn't the same as being justified in thinking that the goods we know of are representative of the goods there are. In response to this serious objection posed by skeptical theists I have tried to do two things. First, I have supplemented my original argument in an effort to blunt the challenge raised by the objection that the goods we know of are not known to be representative of the goods there are. And second, I have sought to show that the skeptical theist's objection, if correct, leads into a black hole for those theists who endeavor to support and defend some of their religious beliefs by philosophical arguments. The supplement I have added to my original argument is the good parent analogy that I relied on in responding to Wykstra's challenging objection to my original argument. So, I can be now be very brief in showing how it lends support to my original argument. God, if he exists, is all-powerful, all-knowing, perfectly good and a supremely loving being. He is to us as loving parents are to their children. It is not for nothing that the common prayer begins with the words: "Owr father which art in heaven." Morever, he is our loving father. What do loving parents do when their children are suffering for reasons they cannot comprehend? Loving parents do their best to relieve the suffering of their children. And if the suffering should be the result of necessary discipline of the child, the loving parents endeavor as best they can to help the child understand what the discipline is for; and they strive to enable the child to be aware of the constancy of their love. And should their children suffer from injury or illness, the loving parents make special efforts to be consciously present to them, showing their love and concern. Moreover, they do their best to help the child understand the illness and what needs to be done to cure it. But in a world supposedly created by their loving heavenly father, countless people suffer horrendously without any sense of his comforting presence or his helping them to understand why he permits them to endure such suffering. The point is this: love entails doing the best one can to be consciously present to those one loves when they are suffering, and particularly so when they are suffering for reasons they do not or cannot comprehend. And given the absence of any loving, heavenly father, the evil and suffering in our world only increases the likelihood that God does not exist. Emphasizing these implications of God's love for his human creatures, however, does not refute the line of reasoning advanced by skep-

218

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 90

WILLIAM L. ROWE

tical theists. For since we are unable to prove that the goods we know of are representative of the goods known by God, there remains the logical possibility that some outweighing good would be lost were God to yield to his perfect love and be consciously present to his human creatures, comforting them while permitting them to suffer for reasons they are unable to comprehend. IV

The skeptical theist's emphasis on our inability to be confident that the goods we know of are representative of the goods there are does raise important questions about the inference in the argument from evil. But when we apply their skeptical thesis to the religious beliefs that they and other believers hold, the position of the skeptical theist is seen to lead to conclusions that are very unfriendly to traditional theism. Christian theists, for example, believe that there is a glorious life after death, at least for those who accept Christ as their savior and endeavor to live in accordance with his teachings. What rational grounds do Christians have for these beliefs? They believe, not unreasonably, that God's ultimate purpose for their lives is seldom, if ever, fulfilled in this earthly life; and that, being perfectly good and loving, God will surely grant them the greatest good conceivable to mankind: everlasting life in the presence of God himself. In addition, they also find some support for these beliefs in the Bible. On the basis of these two sources - the recognition that God's purpose for their lives is not fulfilled on earth, and the teachings of the Bible - they not unreasonably conclude that they have rational grounds for their belief that there is life after death, a glorious everlasting life in the presence of God himself. As forceful as this line of reasoning may seem to the faithful, the position of the "skeptical theists" shows it to be utterly inadequate. For given their skeptical theses, we human beings are simply in the dark about how likely it is that God will bring it about that faithful believers will have a glorious, everlasting life in his presence. We humans are in the dark about this important matter because, as the skeptical theists tell us, we simply are in the dark about the goods that God will know, and the conditions of their realization. For we have no sufficient reason to suppose that the goods we know are representative of the goods there are. And for all we know there is some good far greater than the good of eternal life for the faithful on earth, a

THE PROBLEM OF EVIL FRIENDLY ATHEISM, SKEPTICAL THEISM, AND THE PROBLEM OF EVIL 91

good the realization of which precludes God's granting eternal life to the faithful on earth. For example, consider the Christian belief that there are fallen angels: Satan and his cohorts. They are judged to be higher beings than mere humans on earth. And, if we follow the skeptical theists we must conclude that for all we know, the good of the fallen angels being redeemed by God far exceeds the good of faithful human beings being granted eternal life. And for all we know the cost of God's redeeming the fallen angels precludes God from granting eternal life to any humans. For remember, according to the skeptical theists we are simply in the dark not only about the goods there are but also the conditions of their realization. And it clearly follows from this skeptical view that our being unable to think of any possible good that would justify God in not permitting faithful Christians to experience an afterlife provides no adequate reason to conclude that probably there is no good that justifies God in precluding faithful Christians from experiencing an afterlife. Moreover, as we've seen, for all we know, a condition of the realization of the salvation of Satan and the fallen angels may require God to permit the faithful on earth to perish along with atheists, agnostics, and those who have lived without even forming an idea of such a being as God. Skeptical theists choose to ride the trolley car of skepticism concerning the goods that God would know so as to undercut the evidential argument from evil. But once on that trolley car it may not be easy to prevent that skepticism from also undercutting any reasons they may suppose they have for thinking that God will provide them and the worshipful faithful with life everlasting in his presence. Of course, they may still appeal to some special divine revelation in which God himself supposedly informs them that he will provide faithful, Christian believers with life everlasting in his presence. But to the rest of us, particularly philosophers who find ourselves without the benefit of such special revelations, such a carefully crafted philosophical skepticism will surely appear to be something less than genuine skepticism about whether the goods we know of are representative of the goods there are. Morever, while philosophers may respect religious appeals to special divine revelations, such appeals can hardly provide a philosophical response to the skeptical implications of the philosophical claims of skeptical theism.

219

220

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 92

WILLIAM L. ROWE

Notes 1. "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16 (1979): 335-41. 2. Skeptical theists will say that we are in no position to assert this. For they claim that several of the goods we can think of may, for all we know, be goods, any one of which in fact may be the good that does justify God, if he exists, in permitting the fawn's terrible suffering. They would prefer that the sentence in question be revised as follows: "Moreover, it is extraordinarily difficult to think of, or even imagine, a greater good whose realization we know would require God to permit that fawn's terrible suffering." 3. If some good G outweighs an evil e, then the good state of affairs G&e outweighs e and cannot be obtained without permitting e. I take it as given that if G can exist without e, a perfectly good being (other things being equal) would prevent e. 4. "The Humean Obstacle to Evidential Arguments from Suffering: On Avoding the Evils of "Appearance," International Journal for the Philosophy of Religion, 16 (1984): 73-93. "Evil and the Theistic Hypothesis: A Response to Wykstra," International Journal for the Philosophy of Religion, 16 (1984): 95-100. 5. Since there is a natural tendency to associate the word 'God' with various religious beliefs about when the goods for the sake of which God permits horrendous evils will be realized, beliefs that are not themselves logically derivable from the concept of God, I used the letter 'O' to free the mind from the tendency to think that those religious beliefs are logically entailed by the mere concept of an omnipotent, omniscient, creator of the world. 6. Of course, since these three religious systems contradict one another, we can be sure not all these religious claims can be true. 7. Proslogium, Ch. 1, pp. 3-7 in Anselm: Basic Writings, trans. S. N. Deane. (LaSalle, IL: Open Court, 2n ed., 1966). 8. Two objections have been raised. The first is that the premise is too strong. Instead, it should read as follows: "no good we know of (is known by us to justify) God in permitting many of the evils in our world." The second, more crucial objection is as I go on to state. 9. The best presentation of Skeptical Theism I am aware of is by my colleague, Michael Bergmann. See his important paper, "Skeptical Theism and Rowe's New Evidential Argument from Evil," Nous 35 (2001): 278-296.

PART III REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL

This page intentionally left blank

CHAPTER THIRTEEN

Evil and Theistic Hypothesis: A Response to Wykstra

Wykstra's paper, "The Humean Obstacle to Evidential Arguments from Suffering," is both conceptually rich and an important advance to our understanding of the evidential argument from evil. Its conceptual richness lies in its analysis of the epistemic notion of "appears" and its advancement of a necessary condition (CORNEA) for being entitled to make the claim: "It appears that p (not-p)." Its advance in our understanding of the evidential argument from evil consists in his application of CORNEA to the very starting point in the justification of the first premise of the deductive argument against theism I presented in the article of which his essay is a critique. My response will assume the conceptual contribution Wykstra has made, focussing mainly on the question of whether he has shown that CORNEA is in fact not satisfied when we make claims such as, "It appears that there are instances of suffering that do not serve outweighing goods otherwise unobtainable by an omnipotent, omniscient being." To this end it is important to begin with a distinction that will play a crucial role in my discussion. Let's call standard theism any view which holds that there exists an omnipotent, omniscient, omnigood being who created the world. Letting "0" abbreviate "an omnipotent, omniscient, omnigood being," standard theism is any view which holds that 0 exists. Within standard theism, we can distinguish restricted theism and expanded theism. Expanded theism is the view that 0 exists, conjoined with certain other significant religious claims, claims about sin, redemption, a future life, a last judgment, and the like. (Orthodox Christian theism is a version of expanded theism.) Restricted theism is the view that O exists, unaccompanied by other, independent religious claims. I will return to this distinction, but at the outset I want to focus on restricted standard theism - the view that 0 exists. Consider the following proposition: 1. There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse.

224

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 96

In the paper Wykstra criticizes, I argued that to the extent that we have rational grounds for accepting (1), to that extent we have rational grounds for rejecting the claim that 0 exists. Furthermore, I argued that we do have rational grounds for accepting (1). What grounds? I there claimed that (1) is a rational belief given three things: first, our knowledge of the vast amount of intense human and animal suffering that occurs daily in our world; second, our understanding of the goods that do exist and that we can imagine coming into existence; and third, our reasonable judgments as to what an omnipotent being can do. (To these three I now would add a fourth: our reasonable judgments of what an all-wise omnigood being would endeavor to accomplish with respect to human (and animal) good and evil in the universe.) To illustrate this claim, I set forth an example of intense suffering (the fawn's suffering) and observed that as far as we can determine it serves no greater good at all, let alone one that is otherwise unobtainable by an omnipotent being. Recognizing, however, that, appearances to the contrary, it might nevertheless serve such an outweighing good, I then claimed that it seems quite incredible that all the instances of suffering which serve no greater good we know or can think of, should, nevertheless, be such that none could have been prevented by an omnipotent being without loss of greater good. These claims about what it is that rationally justifies us in accepting (1) have been challenged by several philosophers; but no one, in my judgment, has raised such an important point (and clarifications) as has Wykstra. So let's now turn to consider his basic objection. In my earlier paper I made ample use of the phrase "it appears that" and its substitutes; and since Wykstra's objection is couched in such phrases, it will be best to retain them in discussing our disagreement. Suppose then that we are simply unable to think of any good that exists or might come into existence that both outweighs the fawn's suffering and couldn't be obtained by an omnipotent, omniscient being without permitting that suffering. We might then claim that 2. It appears that the fawn's suffering is pointless - i.e., it appears that the fawn's suffering does not serve an outweighing good otherwise unobtainable by an omnipotent, omniscient being. We can now understand Wykstra's central objection, based as it is on his view of the importance of CORNEA in our being entitled to make claims such as (2). Put in its simplest terms it comes to this. If someone claims that it appears that S is not P, that person is entitled to that claim only if she has no reason to think that if S were P things would strike us pretty much the same. Thus on detecting no sour odor, the person with the cold is not entitled to the claim that it appears that the milk is not sour because he presumably has a reason (the cold) to think that if the milk were sour things would strike him pretty much the same - he would still detect no odor of sourness. Now let's return to (2). Wykstra believes that we are not entitled to affrim (2) because, in his judgment, we have good reason to think that were the fawn's suffering actually to serve an outweighing

REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL

225 97

good, otherwise unobtainable by O, things would strike us in pretty much the same way - we would still be in the position of not being able to think of any good that exists or might come into existence that both outweighs the fawn's suffering and couldn't be otherwise obtained by an omnipotent, omniscient being. The reason he gives is that were O to exist, it would be likely that the outweighing good in relation to which 0 must permit the fawn's suffering would be a good quite beyond our ken. And if such good were beyond our ken, then we would still be in the position of not being able to see what good is served by the fawn's suffering. Thus Wykstra concludes that were 0 to exist, the fawn's suffering would likely have just the feature I claim it to have — of serving no outweighing good we know to exist or can think of which is otherwise unobtainable by an omnipotent, omniscient being. It's helpful, I think, to view Wykstra's objection as having two steps. The first step is the claim that in the situation described we are entitled to affirm proposition (2) only if the following proposition is true: 3. We have no reason to think that were O to exist things would strike us in pretty much the same way concerning the fawn's suffering. The second step in Wykstra's objection is the claim that (3) is false. For were 0 to exist, Wykstra thinks it likely that the outweighing good in relation to which 0 must permit the fawn's suffering would be a good beyond our ken. My response to Wykstra will focus entirely on the second step in his objection: the step which claims that were 0 to exist things would strike us in pretty much the way they do so far as instances of human animal suffering are concerned. Before criticizing this step, however, it's important to note that Wykstra could hold that in the situation described we are entitled to affirm proposition (2) only if (3') is true. 3'. We have reason to think that were O to exist things would strike us differently concerning the fawn's suffering. He might then argue that (3') is false and that, therefore, it is not reasonable to believe that were the fawn's suffering actually to serve an outweighing good, otherwise unobtainable by an omnipotent being, things would likely strike us differently concerning the fawn's suffering. In the discussion that follows I have taken Wykstra to be arguing that (3) is false and that, therefore, his condition CORNEA is not satisfied. (A careful reading of pages 87-89 makes it clear that it is (3), not (3'), that he holds must be true if we are entitled to affirm (2).) My comments, therefore, concern only his attempt to show that (3) is false. It would be another matter to determine whether (3') is also necessary for being entitled to claim (2), and another matter yet to determine whether (3') is true or false. Given the history of humans and animals, and the sorry tale of their sufferings

226

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 98

through the ages, and given our inability to discover among the goods we know or can imagine, goods that both outweigh these sufferings and are unobtainable by omnipotence unless these sufferings are permitted, why should we think, as Wykstra does, that this is just how things would likely be, if O exists? It's true, as Wykstra observes, that O's mind can grasp goods that are beyond our ken. The idea, then, is that since 0 grasps goods beyond our ken, we've reason to think it likely that the goods in relation to which 0 permitts many sufferings that occur would be unknown to us. Let's look at Wykstra's reasoning here. He starts with 4. O's mind grasps goods beyond our ken. moves to 5. It is likely that the goods in relation to which 0 permits many sufferings are beyond our ken. and concludes with 6. It is likely that many of the sufferings in our world do not appear to have a point - we can't see what goods justify 0 in permitting them. The difficulty with this reasoning is that the move from (4) to (5) presupposes that the goods in question have not occurred, or, at the very least, that if they have occurred they, nevertheless, remain quite unknown to us (in themselves or in their connections with the sufferings in our world). And, so far as I can see, the mere assumption that 0 exists gives us no reason to think that either of these is true. If 0 exists it is indeed likely, if not certain, that O's mind grasps many good states of affairs that do not obtain and which, prior to their obtaining, are such that we are simply unable to think of or imagine them. That much is reasonably clear. But the mere assumption that 0 exists gives us no reason whatever to suppose either that the greater goods in virtue of which he permits most sufferings are goods that come into existence far in the future of the sufferings we are aware of, or that once they do obtain we continue to be ignorant of them and their relation to the sufferings. The fact that O's mind can apprehend non-actual good states of affairs that we are not able to think of gives us reason to assent to the following proposition: 7. If O exists then the outweighing goods in relation to which some sufferings are permitted by 0 are, antecedent to their obtaining, beyond our ken. But this proposition is insufficient to justify the claim that if O were to exist the sufferings in our world would appear to us as they in fact dp. I conclude, therefore, that Wykstra has provided no adequate justification for the second step in his objection. And without that step, the objection fails.

REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL

227 99

Earlier, I distinguished restrictive standard theism from expanded standard theism. I now want to bring this distinction to bear on our problem. Standard theism implies (roughly) that the sufferings that occur are permitted by 0 by virtue of outweighing goods otherwise unobtainable by 0. Restricted standard theism gives us no reason to think that these goods, once they occur, remain beyond our ken. Nor does restricted standard theism give us any reason to think that the ocurrence of the goods in question lies in the distant future of the occurrence of the sufferings that 0 must permit to obtain them. So I conclude, contrary to Wykstra, that the mere hypothesis of O's existence gives us no good reason to think that things would appear to us just as they do so far as the sufferings of animals and humans in our world are concerned. I conclude, therefore, that we've been given no convincing reason to reject my view that the items mentioned at the outset give us reason to believe that (1) is true and, therefore, that 0 does not exist. But what about expanded theism? Suppose, for example, that we add to the hypothesis that 0 exists the claim made by Saint Paul in his letter to the Church at Rome: "For I reckon that the sufferings we now endure bear no comparison with the splendor, as yet unrevealed, which is in store for us." Playing fast and loose with biblical hermeneutics, suppose we set forth the Pauline doctrine as follows: 8. The goods for the sake of which O must permit vast amounts of human and animal suffering will be realized only at the end of the world. We now have a version of expanded standard theism, a version consisting of the conjunction of the proposition that 0 exists and proposition (8). Let's call this particular expansion of standard theism EST (expanded standard theism). EST is not rendered unlikely by the items that render RST (restricted standard theism) unlikely. And this, for precisely the reason Wykstra so clearly and carefully sets forth. Given EST we have some reason to think that the fawn's suffering might well appear to us just the way it does appear. So the fact that it appears to us as it does in no way renders EST more unlikely than it otherwise is. Since the facts about suffering do not tell against EST, if RST implied EST (or something close to it) then Wykstra's objection would be decisive against my claim that the facts about suffering render RST unlikely. But it is clear, I believe, that RST does not imply EST. It is not so clear, however, that Wykstra would accept this point. He seems to think that the theistic hypothesis somehow "contains" the claim that the goods for the sake of which 0 permits the sufferings in the world are, to a large extent, quite beyond our ken. On p. 89 he speaks of this claim as a "logical extension of theism." On p. 91 he says that the claim is not an "additional postulate," but instead was "implicit" in theism all along. But this is a mistake. What is "implicit" in theism (RST) is that O's mind grasps goods that are beyond our ken (proposition (4) above). And, as we've seen, (4) is significantly different from the proposition that Wykstra claims to be a part of the theistic hypothesis. My own best judgment is that the crucial proposition Wykstra claims

228

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 100

to be implicit in theism is in fact an added postulate that produces a version of expanded theism, a version that is not rendered unlikely by the facts about suffering that I claim to render restricted standard theism unlikely. Wykstra notes that among believers as well as nonbelievers there is a "persistent intuition that the inscrutable suffering in our world in some sense disconfirms theism." I would say that there are two distinct intuitions at work concerning suffering and theism. First, there is the intuition that the magnitude and intensity of human and animal suffering disconfirms theism. It is hard to see that suffering of this magnitude and intensity is just what we should expect on the hypothesis that 0 exists. Second, there is the intuition that the existence of so much suffering for which we are unable to see any point at all disconfirms theism. It is hard to see that the existence of so much suffering for which we can see no point whatever is just what we should expect on the hypothesis that 0 exists. Both of these intuitions are, I believe, essentially correct. Given our cognitive limitations and O's omniscience and omnipotence, Wykstra has argued that a good deal of the suffering in our world can be expected to have the characteristic of being inscrutable to us. He concludes that our second intuition is based on a failure to see what is really contained in the theistic hypothesis. But, if I am right, what Wykstra has done is "read into" the theistic hypothesis a proposition that is part of the story of traditional theism, thus creating a version of expanded theism. He has, unwittingly, changed the question. The crucial question is whether the facts about suffering in our world tend to disconfirm the hypothesis that 0 exists. That question cannot be shown to deserve a negative answer merely by showing that we can supplement the hypothesis that 0 exists with other propositions such that the supplemented result is not disconfirmed by the facts that are claimed to disconfirm the hypothesis that 0 exists. Wykstra, I believe, would agree with this last point. What he would disagree with me about is whether he has really supplemented the theistic hypothesis in the way I have suggested he has done. If he has not, then he has found a fundamental error in my original paper. And even if he has unwittingly supplemented the hypothesis that 0 exists in the way I have suggested, his paper remains an important contribution to our understanding of the basic issues involved in evaluating the evidential argument from evil.

CHAPTER FOURTEEN

Response to Linda Zagzebski

I want to discuss two points Linda Zagzebski makes: one in criticism of me and one she makes by the way. The point she makes by the way is one I agree with and constitutes, I think, the basis of another objection to Hick's efforts to support his soulmaking theodicy. I shall return to this point. Let me begin, however, with her criticism. 1 Hick acknowledges that it appears reasonable to believe that there are evils an omnipotent being could have prevented without loss of soul-making. Let us call any such evils "excess evils". Since it is rational to believe there are excess evils, doesn't this make it unlikely that the following theses is true? I It is to achieve the good of soul-making that God permits the world (with all its evils) to be as it is. In response to this objection, Hick proposes what I shall call Hick's paradox: II Unless it were reasonable to believe that evils occur in excess of what God would have to permit for soul-making, significant soul-making would be diminished. I have noted the merit of this response but have pointed out that in setting it forth Hick uses an illicit "all-or-nothing argument". For it is rational to believe that evils that appear not to serve soul-making occur in far greater excess than what is required for it to be rational for us to believe that there are excess evils. (Who would say that if only 5 million had been permitted by omnipotence to perish in the Holocaust it would not have been rational to believe that evils occur that omnipotence could have prevented without loss of soul-making?) Zagzebski argues that I cannot discredit thesis I by using the point that it is rational to believe that evils not needed for

230

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William Rowe

131

soul-making occur far in excess of what is required for it to be rational for us to believe that there are excess evils. Her point is this. Since the evils required to render it rational for us to believe there are excess evils are, by that very fact, required for soulmaking, it follows that "it is rational to believe that there is evil [not needed for soul-making] in excess of what must exist in order to make it rational to believe that there is excess evil". And, of course, if Thesis I implies (a) It is rational to believe that there is evil [not needed for soul-making] in excess of what must exist in order to make it rational to believe that there is excess evil then we cannot use (a) to discredit thesis I. But does thesis I imply (a)? I don't think so. With the help of Hick's paradox, we can reason our way from thesis I to the claim (b) It is rational to believe that excess evils occur. But from thesis I we cannot infer (a). From thesis I it does follow that whatever evils are required to render it rational to believe there are excess evils are not themselves excess, for in making (b) true these evils serve the soul-making process. But nothing else about what it is rational to believe follows from thesis I. Perhaps Zagzebski is arguing as follows. (1) It is rational to believe there are excess evils. (2) If thesis I is true then the evils that make (1) true are not excess evils - they are needed for soul-making. Therefore, (3) If thesis I is true then it is rational to believe there are evils not needed for soul-making that are in excess of the evils required to make it rational to believe there is excess evil. I believe this argument is invalid: (3) does not logically follow from (1) and (2). So I continue to hold that

REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL

132

Response to Linda Zagzebski

(c) It is rational to believe that there are excess evils in amounts far beyond what is required for it to be rational to believe there is excess evil counts against the soul-making theodicy (i.e. thesis I). 2 In the course of her penetrating discussion Zagzebski remarks, "My point, then, is that, if the soul-making hypothesis is true, it would be rational for us to believe that it is false. That is, it would be rational for us to believe that there is a great deal of excess evil beyond that needed for soul-making." I think she is exactly right in this remark. Or at least she is right if Hick is right in holding that soul-making would be diminished unless it were rational for us to believe that there are evils not needed for soul-making. Her point means, however, that thesis I has a peculiar feature: it is true only if it is rational for us to believe it is false. Zagzebski hastens to observe, however, that we cannot use (d) It is rational to believe that there is evil beyond what omnipotence need permit for soul-making to disconfirm the soul-making hypothesis (thesis I), for (d) is what is predicted by the hypothesis. Again I agree. What I used to disconfirm thesis I is not (d) but (c). But, although (d) won't disconfirm the soul-making hypothesis, it will, I believe, undermine Hick's central purpose in proposing the soul-making hypothesis. For we must remember that Hick's purpose is more than to state the soul-making hypothesis: it is to argue that it is reasonable (or at least not unreasonable) to believe that the hypothesis is true. What Hick aims to do is show that it is reasonable to believe that it is to achieve the good of soul-making that God permits the world (with all its evil) to be as it is. But, once we see that the soul-making hypothesis is true only if it is rational to believe that it is false, we cannot consistently go on to argue that it is reasonable to believe the soul-making hypothesis. The general point underlying the objection just presented is this. If our aim is to show that it is reasonable to believe p, we cannot consistently proceed to argue that the truth of p depends on it being rational to believe that p is false. The trouble with Hick's paradox is that, although it neatly disposes of a serious objection

231

232

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William Rowe

133

("So much evil appears to be unrelated to, or even to frustrate, soul-making"), it does so at the enormous price of rendering it unreasonable to believe the soul-making hypothesis to be true.

CHAPTER FIFTEEN

Reply to Plantinga

Consider two instances of evil. E1: a five year old girl's being brutally beaten, raped, and strangled in Flint, Michigan on New Year's eve a few years ago. E2: a fawn's dying a lingering and painful death due to being horribly burned in a forest fire occasioned by lightning. Theists believe that God exists and is justified in permitting El and E2. Since evil is intrinsically bad, what justifies God in permitting El and E2 is something else. To simplify matters we will suppose here that the something else is a good state of affairs (perhaps a conjunctive state of affairs) that God brings about and could bring about only by permitting these evils. Now that good is either a good we know of or a good beyond our ken. [Goods we know of will include both actual and nonactual goods, and goods that can be actual only if God exists—for example, the little girl's enjoying eternal felicity in the presence of God.] The theist, then, believes the following proposition. God exists and (some known good justifies him in permitting El and E2 or some unknown good justifies him in permitting El and E2). This proposition is equivalent to a disjunction whose first disjunct is: God exists and some known good justifies him in permitting El and E2.1 argue that this first disjunct is false.1 That is, I argue that ~[God exists and some known good justifies him in permitting El and E2.] Given that necessarily there is at most one God, this proposition is equivalent to P: God does not exist or (God exists and no known good justifies him in permitting El and E2). © 1998 Blackwell Publishers Inc., 350 Main Street, Maiden, MA 02148, USA, and 108 Cowley Road, Oxford OX4 1JF, UK.

234

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 546

NOUS

I try to convince the theist that P is true, or at least that it is quite likely that P is true. Of course, T could hardly persuade the theist of this by suggesting that the first disjunct of P is true. And, as a nontheist, I'm not prepared to assert the truth of its second disjunct. But we can know that a disjunction is true without knowing which disjunct is true.2 Suppose that Smith was killed in his home under circumstances that make it clear that either he took his own life, but made it look like murder, or he was in fact murdered. Of course, if he was murdered, he was murdered by an acquaintance or a stranger. Consider, then, the proposition: Some acquaintance murdered Smith. Suppose we learn that this proposition is false. We learn that A: Smith was not murdered or he was murdered but no acquaintance murdered him.3 We learn that A is true by discovering, say, that each of Smith's acquaintances was somewhere else at the time Smith was killed in his home. So, we know that the disjunction A is true. But we don't know which disjunct is true. Similarly, we may have grounds for believing P to be true without having grounds for a particular one of the two disjuncts. Suppose that prior to learning the whereabouts of Smith's acquaintances at the time Smith was killed, and thereby learning that A is true, it was equally likely, given our information, that Smith was murdered as that he was not (having committed suicide instead). Now that we have learned that A is true, how is the likelihood that Smith was murdered affected? It is now more likely that he was not murdered than that he was. For, since A is entailed by Smith was not murdered, A is more likely given that Smith was not murdered than it is given that he was murdered. So, if we learn that A is true, it becomes more likely than before that Smith was not murdered. And the same holds for God's existence and our learning that P is true. Suppose that prior to learning that P is true, our information made it equally likely that God exists as that God does not exist. Now that we have learned that P is true (supposing that we have learned this), how is the likelihood that God exists affected? It is now more likely that God does not exist than that he does exist. For, since P is entailed by God does not exist, P is more likely given that God does not exist than it is given that he does exist. So, if we learn that P is true it becomes more likely than before that God does not exist.4 Plantinga believes that there is something dreadfully wrong with my argument from P to it being more likely than not that God does not exist. First, he thinks that my efforts to persuade the theist that P is true are doomed to failure. For he thinks that "a theist is committed to denying that no known good justifies a perfect being in permitting El and E2." Second, he thinks there are other arguments that counterbalance my argument from P. Third, he thinks my argument from P is what he calls "an argument from degenerative evidence." And on the basis of these three objections, he concludes that the argument has no tendency at all to show "that it

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL REPLY TO PLANTINGA

235 547

is more rational to accept atheism than to accept theism." I will reply to each of these objections.

Plantinga gives two reasons for holding that a theist is committed to denying P. First, recognizing that apart from appealing to the existence of God we have little reason to think that the actual world, a, is itself a good state of affairs, Plantinga fastens on F, "the conjunction of all the goods, gl, g2,..., a includes or contains." He then takes it that F is a "known good"; and concludes that if God exists, then T justifies him in permitting El and E2. When I endeavored to distinguish between goods we know of and goods unknown to us (goods beyond our comprehension), I never imagined that someone would think that the conjunction of a known good with a good beyond our comprehension is somehow itself a known good.5 Nor did I imagine that someone would think that the conjunction of all the goods we know of with all the goods beyond our comprehension, would itself be a good we know of. By 'goods we know of I mean good states of affairs we know of in the sense that we are acquainted with them in some significant way beyond merely knowing that they (the states of affairs in question) are good. To merely know that F is a good state of affairs is insufficient for us to be acquainted with F. F is not a good we know of. A case can be made for saying that the conjunction of all the known goods in F is itself a known good. But the latter will not serve Plantinga's purpose. For if God exists, it may be that it is some unknown good that justifies him in permitting ElandE2. Plantinga's second point in support of the view that some known good justifies a perfect being in permitting El and E2 consists of an example. He suggests that the good of enjoying God's gratitude in eternal felicity may be what justifies God in permitting El. I think there are good reasons for rejecting this suggestion.6 But since the issue is to provide some known good that is a live possibility for justifying God in permitting El and E29 we need not pause over the only example Plantinga provides.7 Although Plantinga may be unconvinced that there are good reasons to accept P, or to think it more likely than its denial, I don't think he has provided any good reasons for his view.8 So, I will continue to present my argument in terms of P itself, rather than in terms of Plantinga's proposed revision of P.9 II

Plantinga proposes to counterbalance my P with the claim that P*: Neither El nor E2 is such that we know that no known good justifies a perfect being in permitting it.

236

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 548 NOUS

Just as my P is entailed by ~G (God does not exist), P*, Plantinga claims, is entailed by G (God exists). But this is a mistake. For it could be that it is some unknown good justifies a perfect being in permitting El or E2. So, God's existence is compatible with our knowing that no known good justifies him in permitting either El or E2.10 Nevertheless, Plantinga's example is easily repaired. P*': Neither E1 nor E2 is such that we know that no good justifies a perfect being in permitting it. P*' is entailed by G. But there is a proposition that is patterned after P*' that is entailed by ~G, and forms what I take to be the natural counterbalance to P*'. P**': Neither El nor E2 is such that we know that some good justifies a perfect being in permitting it. These two effectively counterbalance each other. So, Plantinga's argument has its own natural counterbalancesn What would be very relevant to my argument from P is to find a proposition that is a natural counterbalancer to it, that is entailed by the proposition that God exists, and that we (theists and nontheists) are able to show to be true, or at least to be very likely true, independent of any appeal to the proposition that God exits.12 We could try Some good we know of justifies God in permitting E. But clearly we could not justify that proposition independent of an appeal to the existence of God. We could try something like: No evil we know (to have occurred) is such that Godwouldn 't be justified in permitting it, if he were to exist.13 This proposition is entailed by God exists and, therefore, increases the likelihood that God exists. But it would be exceedingly difficult to reason our way to its being true apart from appealing to the proposition that God exists. Ill

Plantinga's third objection, his main challenge to my argument from P, has three steps. First he claims that P is logically equivalent to the following disjunction: P*: Either God does not exist or no known good stands in J to E.14 Second, he holds that what I reason toward is solely the second disjunct of P*. And, finally, he thinks that the way I arrive at P* is by using addition to disjoin God does not exist with no known good stands in J to E. Because he thinks these three things, Plantinga thinks my argument parallels "precisely" the reasoning in his barefoot argument.15 For we can just as easily disjoin no known good stands in J to E with God exists as we can with God does not exist. And these two disjunctions will counterbalance one another with respect to lowering the prob-

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL REPLY TO PLANTINGA

237 549

ability of God exists/God does not exist. But is Plantinga right in claiming that my argument parallels "precisely" the reasoning in his barefoot argument? Before we can answer this question we need to consider what conditions must exist if it is to be true that some good g justifies God in permitting El and E2. Plantinga provides two different accounts, accounts that are not equivalent. For his first account implies that a good g justifies an evil e only ifg is actual; whereas, the second allows that a good g may justify an evil e even ifg is not actual. According to his second account, a good g justifies e provided that if there were a perfect being b, and g and e were actual, b would be justified by g in permitting e. Clearly, if God doesn't exist, e is actual, and g is not actual, it still may be true that if God were to exist and g and e were actual, he would be justified by g in permitting e. My own approach to this matter is not to try to provide a set of conditions specifying relations between some good g and some evil e such that if all those conditions are met and God exists, God will be justified by g in permitting e. I don't think we are in any position to know the limits of such a set. What we do know, I believe, are some conditions that must hold between a good g and an evil e if it is to be true both that God exists and that g justifies God in permitting e. And I think Plantinga has elegantly set forth several such conditions. We know, for example, that if God exists and some evil e is actual, then g justifies God in permitting e only if (1) g is actual, (2) g outweighs e, (3) g cannot exist unless e is permitted to exist by a perfect being, and (4) no better world can be brought about ifg and e are prevented by a perfect being. Let us, then, following Plantinga, say that if any one of these conditions fails to hold between a good g and an evil e, g fails to stand in J to e. However, departing from Plantinga, because we don't know that there aren't other conditions that must obtain for a good to justify God in permitting e, we won't say that if God exists and some good g satisfies (1 )-(4) in relation to some evil e, then g justifies God in permitting e. I don't mean to positively deny that this is so. I mean only to point out that God may have requirements that we don't know to be requirements.16 At best, then, we can only state some conditions that are necessary for a good g to justify God in permitting and evil e. We can now return to the question of whether my argument parallels "precisely" the reasoning in his barefoot argument. As a start, we should note that what one comes to learn initially in the barefoot argument is something (B: I am now barefoot) that is evidentially irrelevant to whether God exists or does not exist. We can see that this is so by noting that k, our background information for the barefoot argument, contains the information that I am barefoot about one-half of the time.17 So, given k, it is just as likely that I am now barefoot as that I am not. Since God presumably has little interest in whether I am or am not now barefoot, the probability of (B/G and k) is 1/2, just as the probability of (B/~G and k) is 1/2. What happens next is that the nontheist uses addition so as to infer (~G or B) from B, thus producing a statement that lowers the probability of God's existence; while the theist returns the favor by using addition to infer (G or B) from

238

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 550

NOUS

B, thus producing a statement that lowers the probability of God's nonexistence. Compare now what we learn in my argument. Upon reflecting both on the goods we know of and El and E2, we come to see that that none of these goods satisfies the four conditions that must be satisfied if one of these goods has J to El and E2. Suppose I am right about this. Suppose both that no known good g satisfies conditions (1 )-(4) relative to E1 and E2 and we have very good reason to believe that this is so. We thus have very good reason to believe that X: No known good has J to El and E2. From this result (X) we directly infer No good we know of justifies God in permitting El and E2.18 Since this proposition is logically equivalent to P, it is entailed by ~G, and thus lowers the probability of G. It should be clear that the above reasoning does not proceed "precisely" as the barefoot argument proceeds. For unlike B (I am now barefoot), what we learn initially in this reasoning is something (X) that is evidentially very relevant to the question of whether God exists. As we saw in the barefoot argument, B (I am now barefoot) is as likely on (~G and k) as it is on (G and k), and for that reason it itself is evidentially irrelevant to the question of God's existence. But (X) itself makes ~G more likely than G. For X is more likely given ~G than it is given G. If God does not exist, it is quite likely that no known good g satisfies each of the four conditions necessary for g to have J to E1 and E2. That is, it is very likely that no known good is actual, outweighs El and E2, cannot possibly be realized by a perfect being without its permitting E1 and E2, etc. One reason this is so is that no good involving God, such as the little girl's enjoying eternal felicity in the presence of God or (Plantinga's suggestion) the little girl's enjoying God's gratitude in eternal felicity, is even a candidate for consideration, since such a good is actual only if God does exist. But if God does exist, these goods are at least candidates for the job of justifying him in permitting the little girl's suffering. So, for this reason and others, X is much more likely if God does not exist than it is if God does exist. However, B (I am now barefoot) is no more likely if God does not exist than it is if God does exist. But what if Plantinga takes (X), from which we directly inferred P, and uses the rule of addition so as produce the disjunction, G or X? Won't he then have counterbalanced my argument? For just as ~G entails P, so does G entail (G or X). The answer is no. For, as we've just seen, X itself is a reason to think that ~G (God does not exist) is more likely than G (God exists). Of course, nothing I have noted prevents one from using X in the fashion that Plantinga describes as "an argument from degenerate evidence." Someone may leam X as new evidence, add ~G so as to produce ~G or X, and then claim that what has been arrived at lowers the probability of G. But, of course, what is sauce for the goose is sauce for

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL REPLY TO PLANTINGA

239 551

the gander. The theist may accept X, use addition to arrive at G or X, and then claim that what has been arrived at lowers the probability of ~G.19 But my argument is not "an argument from degenerate evidence." It can be taken as an argument from X alone. For X itself, given our background information k, is sufficient to render God's existence less likely than it was before we acquired X.20 And that is sufficient to establish that my argument is not "an argument from degenerate evidence." For Plantinga stresses that in an argument from degenerate evidence a part of your evidence (e.g., B or G) makes a certain proposition (G) probable when there is "an isomorphic part of your evidence" (B or ~G) with respect to which (~G) is at least "equallyprobable." But, as we've seen, neither (P) nor (X), from which we directly inferred (P), is such that it makes G anywhere near as probable as ~G. For, unlike B, which leaves G as probable as ~G, X makes ~G more likely than G. I conclude that none of the three main points Plantinga advances to show that my argument "has no tendency at all to show that it is more rational to accept atheism than to accept theism" is successful.21 Notes 1. See "The Evidential Argument from Evil: A Second Look" in The Evidential Argument from Evil, ed. Daniel Howard-Snyder (Bloomington and Indianapolis: Indiana University Press, 1996), 262-285. 2. Given that we don't know the truth value of p, the obvious case is a proposition of the form pv~p. 3. I assume here for purposes of illustration that necessarily at most one person murdered Smith. 4. See "The Evidential Argument from Evil: A Second Look" for proof of this point. 5. Suppose a man who is known to me is put into a room with some other man who is unknown to me. I am then told that these two men (the one known to me and the one unknown to me) are in that room. Is it true that the two men in that room are known to me? I do know that two men are in the room. But the two men are not known to me. 6. Is it, for example, clear that it makes any sense at all for God to be grateful to the little girl for enduring suffering that she had no opportunity to avoid or not endure? Saint John of the Cross prayed to God to be allowed to suffer as Christ suffered. It would be absurd to think that the little girl freely sought to suffer in this way. 7. For a discussion of various proposals of known goods that might be thought to justify God in permitting El and E2, see William Alston's illuminating essay, "The Inductive Argument from Evil," Philosophical Perspectives, 5, Philosophy of Religion, ed. James Tomberlin (Atascadero: Ridgeview Publishing Co., 1991), 29-67. Also see my essay, "William Alston on the Problem of Evil," The Rationality of Belief and the Plurality of Faith, ed. Thomas Senor (Cornell University Press, 1996), 71-93. 8. Moreover, it is evident, I believe, that the theist is not committed to denying that the good that justifies God in permitting El and E2 is an unknown good (a good beyond our comprehension). 9. See "The Evidential Argument from Evil: A Second Look" for a discussion of P and the reasons in favor of accepting it. 10. I assume that Plantinga is here still thinking that T is a known good. 11. A further difficulty with P*' is that its probability is very high, given k. For k clearly allows that there are unknown goods. And so long as there are unknown goods, we cannot know that no good justifies God in permitting El and E2.

240

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 552

NOUS

12. An important feature of P is that we may reason our way to accepting it without any appeal to the claim that God does not exist. 13. This example surfaced in an email discussion with Plantinga. 14. I've made some slight changes in Plantinga's formulation here so as to render it closer to my actual formulation of P. These changes do not affect the point Plantinga is making. 15. In the barefoot argument what one initially learns, B (I am barefoot), does not lower the probability of G (God exists). But by using addition to infer (~G or B), one can then use (~G or B) as a basis for increasing the probability of ~G. For ~G entails (~G or B). Unfortunately for the nontheist, nothing prevents the theist from using addition to infer (G or B) and using it to show that G is more likely than ~G. 16. For example, for all we know God may require that any good for which he is prepared to permit El (the little's girl's suffering upon being brutally beaten, raped, and murdered) must involve some good for the little girl. 17. Unless k is understood to contain this information, we won't be able to hold that Pr(B/G & k) = 1/2. 18. Given the uniqueness of God, this proposition is logically equivalent to P: God does not exist or (God exists and no known good justifies him in permitting El and E2). 19. It is worth noting, however, that even here the parallel with the barefoot argument is not exact. For, unlike B (I am barefoot), X is not neutral with respect to what is derived by adding G (i.e., G or X); for, as we've seen, X itself, given what we already know, makes it less likely that G is true. 20. It should be noted, however, that if ~G does not logically imply X, then even though X reduces the likelihood that God exists, it won't reduce it to quite the degree that P does. For P is logically implied by ~G. 21. I'm grateful to Jan Cover, Jeff Jordan, Alvin Plantinga, Stephen Wykstra, and especially Mike Bergmann for helpful comments on some of the issues discussed in this paper.

CHAPTER SIXTEEN

Skeptical Theism: A Response to Bergmann

In his essay "Skeptical Theism and Rowe's New Evidential Argument from Evil"l Michael Bergmann argues that my 1996 effort to strengthen the evidential argument from evil2 fails because it presupposes the falsity of one or more plausible skeptical theses. The skeptical thesis that he cites most frequently is ST1: We have no good reason for thinking that the possible goods we know of are representative of the possible goods there are. Bergmann suggests that ST1 is reasonable—or at least not unreasonable— owing to our "awareness of our cognitive limitations and the vastness and complexity of reality." And he notes that I seem to have some sympathy with STL He then considers my "new evidential argument from evil" and argues that it depends on "a rejection of this skeptical thesis and, therefore, suffers from the same problem that afflicts his [Rowe's] original argument." In what follows, after a preliminary comment, I will respond to Bergmann's discussion of my 1996 argument from evil. The constant theme in my discussions of the problem of evil is our awareness that no good within our ken can reasonably be thought to justify an allpowerful, all-knowing, perfectly good being in permitting any particular instance of the vast number of instances of horrific suffering (both animal and human) that occurs daily in our world. Skeptical theists like Bergmann do not directly dispute this point, although they think my claim exceeds what we are fully justified in asserting.3 Rather, they contend that given the disparity between our knowledge of goods and the conditions of their realization and the knowledge of these matters possessed by God, if he exists, we simply aren't epistemically qualified to make any reasonable judgments whatever about the amount of seemingly pointless, horrific evil such a being would need to allow in the world. For all we know, they say, the world could contain a great deal more seemingly point© 2001 Blackwell Publishers Inc., 350 Main Street, Maiden, MA 02148, USA, and 108 Cowley Road, Oxford OX4 1JF, UK.

242

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 298

NOUS

less evil and still be reasonably (or, at least, not unreasonably) believed to be under the care and control of an infinitely powerful, all-knowing, perfectly loving being. For, to repeat their constant refrain, since we don't know that the goods we know of are representative of the goods there are, we cannot know that it is even likely that there are no goods that justify God in permitting whatever amount of apparently pointless, horrific evil there might occur in our world. Indeed, if human life were nothing more than a series of agonizing moments from birth to death, their position would still require them to say that we cannot reasonably infer that it is even likely that God does not exist.4 But surely such a view is unreasonable, if not absurd. Surely there must be some point at which the appalling agony of human and animal existence as we know it would render it unlikely that God exists. And this must be so even though we all agree that God's knowledge would far exceed our own. I believe my theistic friends have gone considerably beyond that point when in light of the enormous proliferation of horrendous evil in this world they continue to insist that we are unjustified in concluding that it is even the least bit unlikely that God exists. But the issue to be addressed here is whether my 1996 version of the evidential argument from evil really depends on rejecting the skeptical thesis (ST1) stated above. In the essay in question, instead of focusing on the seemingly limitless amount of horrific evils that serve no good within our ken, I developed the argument in terms of two particular examples of such evils, El and E2.5 How does Bergmann endeavor to show that my 1996 argument depends on my rejecting ST1? First, he correctly notes that my argument is successful only if we are justified in believing that in relation to our background information and the claim that God exists, the likelihood that the goods that justify God in permitting El and E2 are beyond our comprehension is not high.6 That is, my argument will work only if we are justified in believing it likely, or as likely as not, that the good that justifies God, if he exists, in permitting El and E2 will be a good we know of, a good within our ken. For then, if we discover that none of the known goods is such that it justifies God in permitting El and E2 we will be justified in concluding that it is unlikely that God exists. Second, he claims that the only way I have of justifying the claim that it is likely, or at least not unlikely, that the God justifying goods for El and E2 will be goods known to us is by assuming as a premise the denial of ST1. It is Bergmann's justification for this second point that I wish to discuss. He arrives at this second point in the course of considering my discussion of the good parent analogy, an analogy favored by theists as a way of explaining our cognitive limits in terms of understanding the goods for the sake of which God permits all this horrendous human and animal suffering. The trouble with the analogy between the good parent and God is that it leads to a conclusion inimical to theism. Bergmann quotes the relevant passage from my 1996 essay. If we do apply the parent analogy, the conclusion about God that we should draw is something like the following. When God permits horrendous suffering for the sake of some good that is beyond our ken, God will make every effort to be con-

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL Skeptical Theism: A Response to Bergmann 299 sciously present to us during our period of suffering, will do his best to explain to us why he is permitting us to suffer, and will give us special assurances of his love and concern during the period of the suffering. Since enormous numbers of human beings undergo prolonged, horrendous suffering without being consciously aware of any such divine presence, concern, and explanations, we may conclude that if there is a God, the goods for the sake of which he permits horrendous human suffering are more often than not goods we know of. (276)

He then suggests that the relevant premise at work here is the following. (1) If God exists and the goods that justify God in permitting El and E2 are beyond our ken, then it is likely that we would not have divine silence (i.e., it is likely that we would at least have assurances of God's love and of the fact the there is a good that justifies God in permitting such horrendous evils even though we don't know what that good is.) But why should we accept (1)? Well, as seems clear from the context of my discussion, the plausibility of (1) derives from the widely accepted analogy of God to the good, loving parent. When a good, loving parent permits her child to suffer for a good the child cannot comprehend, that parent makes every effort to be consciously present to the suffering child, giving special assurances of her love and concern during the period the child is suffering. Accordingly, Bergmann remarks. "As Rowe points out, the plausibility of (1) has to do with an analogy theists like to employ—the analogy between God and human parents." But immediately after noting that premise (1) finds its plausibility in the generally accepted good parent analogy between God and the loving parent, Bergmann tells us that "premise (1) depends on a prior rejection of ST1." That is, he tells us that the only way we can be justified in accepting (1) is by first being justified in rejecting ST1. Why so? Why couldn't someone just rely on the good parent analogy, so widely accepted in theistic religions, and justifiably conclude that (1)? Bergmann is silent on this question. Instead, he offers his own account of the reasoning process by which people accept (1). He says: It begins with a recognition of the plausibility of: (2) If God exists and the goods that justify God in permitting El and E2 are beyond our ken, then either (a) we wouldn't have divine silence or (b) there is some good that justifies God in permitting divine silence. Next Bergmann tells us that the advocate of (1) "assumes" that (2a) is much more likely than (2b). We then learn that in order to assume that (2a) is much more likely than (2b), we must think it likely that (3) No good justifies God in permitting divine silence.

243

244

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 300

NOUS

Bergmann then asks: "But how could a proponent of (1) come to any reasonable conclusion about how likely it is that there is no such good?" Well, let me make a suggestion. If someone who accepts (1) were to reason in the way in which Bergmann suggests we all do (if we accept (1)), she might say (3) is likely because God is to us as the good loving parent is to her child whom she must permit to undergo suffering (e.g., from a painful medical procedure) for a good the child greatly needs but cannot understand. In situations where the child must suffer for a good the child cannot understand the loving parent justifiably believes it is best to be consciously present to her child, expressing her love and concern.7 Failure to be consciously present to the child she loves while the child is suffering for reasons the child cannot understand is almost always due to human limitations of power and/or knowledge. Thus, again, the good parent analogy itself provides the grounds for thinking that (3) is likely. But Bergmann needs to convince us that our acceptance of (1) "depends" on the rejection of ST1. So, instead of considering the support given by the good parent analogy, Bergmann responds to his question ("But how could a proponent of (1) come to any reasonable conclusion about how likely it is that there is no such good?") by saying: "She could rely on (4) No good we know of justifies God in permitting divine silence."8 But, of course, to infer (3) from (4) presupposes a rejection of ST1. And from this Bergmann concludes that this way of supporting the conclusion [the probability that no known good justifies God in permitting El and E2, given our background knowledge and the claim that God exists, is not high] "doesn't avoid what Rowe seems to want to avoid—namely, reliance on a rejection of the skeptical theses of the skeptical theist." I think we can credit Bergmann with showing that one way of trying to justify (1) If God exists and the goods that justify God in permitting El and E2 are beyond our ken, then it is likely that we would not have divine silence (i.e., it is likely that we would at least have assurances of God's love and of the fact that there is a good that justifies God in permitting such horrendous evils even though we don't know what that good is.) is to rely on (4) No good we know of justifies God in permitting divine silence. But, of course, relying on (4) in order to support (1) does presuppose a rejection of STL And that is not uninteresting. But what Bergmann set out to establish is that this argument of mine, an argument that can be thought of as supporting the conclusion that Pr(P/G&k) is not high, depends for its success "on a rejection of one or more of the skeptical theist's skeptical theses." What he does show, at best, is that one way of supporting (1) involves a rejection of ST1. What he needs to show is that the only way of supporting (1) involves a rejection of STL For only the latter will suffice to show that the argument depends for its success on rejecting ST1. It is as though Bergmann has discovered a road that starts out in the direction of Rome but doesn't in fact lead to Rome. This is not uninteresting. But to conclude from this that there is no road

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL Skeptical Theism: A Response to Bergmann 301

that leads to Rome is not quite right. One can conclude the latter only by showing that if this road doesn't lead to Rome, none other does. Similarly, he shows that one way of trying to justify (1) involves a rejection of STL What he needs to show is that there is no way of justifying (1) that doesn't involve a rejection of ST1. And we may justifiably be skeptical that Bergmann is in any position to show that. Let's now consider more carefully Bergmann's proposition (2). (2) If God exists and the goods that justify God in permitting El and E2 are beyond our ken, then either (a) we wouldn't have divine silence or (b) there is some good that justifies God in permitting divine silence. Bergmann's challenge is to provide a reason for thinking that (a) is more likely than (b).9 Can we do this? I believe we can. But to do so we need to keep before our minds just what is implied by the idea of divine silence, given that God exists and permits seemingly countless instances of horrendous evil such as a five year old girl's being savagely beaten, raped, and strangled. To suppose that God exists and divine silence is what occurs in response to the seemingly countless instances of horrendous suffering in our world is to suppose all of the following: 1. A being of infinite wisdom and power is unable to prevent any of those instances of horrendous suffering without thereby forfeiting a goodl so great that the world would be worse without goodl, even given the instance of horrendous suffering that must be permitted by the infinitely powerful being if that being is to realize goodl. 2. A being of infinite wisdom and power is unable to enable those who undergo such horrendous suffering to understand just what the goodl is for which this infinitely powerful being is required to permit that horrendous suffering without this being thereby forfeiting a good2 so great that the world would be worse without good2, even given the additional suffering occasioned by the sufferers being unable to understand what the goodl is for which an infinitely powerful being permits them to undergo their horrendous suffering.10 3. A being of infinite wisdom and power is unable to be consciously present to those who suffer horrendously, expressing his love and concern during their period of horrendous suffering for a goodl that is beyond their ken, without thereby forfeiting still another good, good3, such that the world would be worse without good3, even given the despair and loneliness of those who undergo seemingly pointless horrendous suffering without any conscious sense of God's being present, expressing his love and concern during their period of seemingly pointless horrendous suffering for a goodl that is beyond their ken. 4. A being of infinite wisdom and power is unable to enable those who undergo horrendous suffering without any sense of God being consciously present expressing his love for them to have any understanding of just what the

245

246

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 302

NOUS

goodS is for which this being is required to permit them to suffer without any conscious awareness of his love and concern without thereby forfeiting still another good, good4, such that the world would be worse without good 4, even given the additional suffering occasioned by the sufferers and their loved ones being unable to understand what the goodS is for which an infinitely powerful being permits them to undergo their horrendous suffering.11 Now my position is that anyone who seriously reflects on (l)-(4) will see the inherent implausibility in the idea that (l)-(4) is the way things are. For the idea that outweighing good states of affairs are so related to the vast array of horrendous evils like El and E2 that an infinitely powerful, infinitely loving being is unable to prevent any of those horrendous evils without (a) forfeiting a greater goodl, (b) forfeiting a greater good2 if he enables the sufferers to understand goodl, (c) forfeiting a greater good3 if he is even consciously present, expressing his love and concern to those who suffer horrendously for a goodl they cannot understand, and (d) forfeiting a greater good4 if he enables the sufferers to understand goodS, etc., etc., etc., is an inherently implausible idea. The skeptical theist, however, may agree with me about the implausibility of this idea. But she will say that its implausibility is derivative, not inherent. And she will argue that we take the idea to be implausible only because we are assuming that the way the goods we know of are known to be related to the evils we know of is representative of the way the goods there are are related to the evils there are. And she will then, following Bergmann's lead, construct an argument for the implausibility of the idea that (l)-(4) is the way things are that employs as a premise the claim that the way the goods we know of are known to be related to the evils we know of is representative of the way the goods there are are related to the evils there are. "See," she will say, "I've shown a way in which you can get to that conclusion by assuming the falsity of a plausible skeptical thesis." My reply is that the idea that (l)-(4) is the way things are is an inherently implausible idea, not dependent for its implausibility on a prior rejection of one or more skeptical theses. Tt is worth noting that an idea may be inherently implausible and yet be justifiably believed. If, for example, we have quite strong evidence for the existence of God we can reasonably believe that (l)-(4) is the way things are. But if, as I assumed in my 1996 essay, our situation at the start is one in which God's existence is no more likely than not, and we then become aware that no good we know of justifies God in permitting El and E2, we will be in a position to conclude that God's existence is less likely than not. For, as I've argued here, to suppose otherwise is to embrace an inherently implausible idea in the absence of a justifying reason to do so. Notes 1

Notts 35 (June 2001), pp. 278-296. "The Evidential Argument from Evil: A Second Look," in The Evidential Argument From Evil, Indiana University Press, ed. Dan Howard-Snyder, 1996. 2

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL Skeptical Theism: A Response to Bergmann 303 3

See Bergmann's footnote 9. In discussions with Bergmann he has indicated that his position is not as extreme as the view I here ascribe to skeptical theists. He allows that there may be some point at which the amount of seemingly pointless evil in the world would justify someone in believing that God does not exist. But this justification, Bergmann believes, would not depend on any principle to the effect that the goods we know of are representative of the goods there are. 5 E1 is a fawn's being trapped in a forest fire and undergoing several days of terrible agony before dying; E2 is a five year old girl's being savagely beaten, raped, and strangled. 6 I am indebted to Bergmann for pointing out to me that in my 1996 paper I did not give any argument for the claim that Pr(P/G&k) is not high. All I did was to show the weaknesses of arguments for the claim that Pr(P/G&k) is high. But, of course, as Bergmann notes, to show that there is no good argument for the claim the Pr(P/G&k) is high does not suffice to show that Pr(P/G&k) is not high. 7 Of course, even here we can think of desert island cases. A mother comes down with a contagious disease that the child may contract in its present state. So, all things considered that loving mother must stay away from her child while he suffers for a good he cannot understand. 8 See Bergmann, p. 283. (Italics mine.) 9 I think Bergmann is right to challenge me to provide such a reason. For providing such a reason seems to be an important step to establishing that Pr(P/G&k) is not high. 10 Of course, it also follows that God is unable to enable the sufferers to understand good2 without thereby having to forfeit some other yet greater good, and so on, and so on. 11 Of course, it also follows that God is unable to enable the sufferers to understand good4 without thereby having to forfeit some other yet greater good, and so on, and so on. 4

247

This page intentionally left blank

CHAPTER SEVENTEEN

Reply to Howard-Snyder and Bergmann

My friends, Dan Howard-Snyder and Mike Bergmann, think that the enormous amount of seemingly pointless, horrendous evil occurring daily in our world gives us no good reason at all to think it unlikely that God exists. For, on the assumption that God exists, they believe we have no good reason to think it probable either that there would be any less horrendous evil or that God would help us understand what some of the justifying goods are that he is powerless to bring about without

250

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 156

THE EVIDENTIAL PROBLEM OF EVIL

permitting all this horrendous evil. In support of their view they liken my argument for the probable nonexistence of God to the reasoning of someone who concludes that there is probably no extraterrestrial life because we don't detect any communications from extraterrestrials. I believe they are right to reject the inference to the likely nonexistence of extraterrestrials from our failure to detect communications from them. For, as they point out, we have no good reason to think that extraterrestrials would know that we exist, or would care about us enough to want to communicate with us, or would have anything like sufficient power and knowledge to devise a way to communicate with us. So, given these considerations, we cannot reasonably infer the nonexistence of extraterrestials from our not having detected any communications from them. As opposed to what we don't know about extraterrestrials, however, we do know that God, if he exists, most certainly knows that we exist, most certainly loves us and cares for us, and, being infinitely powerful, is able to prevent any of the horrendous evils that befall us. Furthermore, given his infinite knowledge, God would know how to achieve the very best lives possible for us with the minimum of horrible suffering. My friends, however, believe that we have no sufficient reason at all to think it even likely that God could achieve the very best for us (humans and animals) were he to have prevented the Holocaust, the terrible suffering of the fawn, the horrible suffering of the little girl, or any of the other countless evils that abound in this world. Why on earth do they believe this? The basic reason is this: God's knowledge of goods and the conditions of their realization extends far beyond our own. Because God's knowledge extends far beyond our own they think it just may be that God would know that even he, with his infinite power, cannot achieve the best for us without permitting all the horrendous evils that occur daily in our world. And they also think it just may be that God can achieve the best for us only if he keeps us in the dark as to what the good is that justifies him in permitting any of these horrendous evils. But what their view comes to is this. Because we cannot rule out God's knowing goods we do not know, we cannot rule out their being goods that justify God in permitting any amount of evil whatever that might occur in our world. If human and animal life on earth were nothing more than a series of agonizing moments from birth to death, the position of my friends would still require them to say that we cannot reasonably infer that it is even likely that God does not exist. For, since we don't know that the goods we know of are representative of the goods there are, we can't know that it is likely that there are no goods

— REPLIES TO CRITICS ON THE EVIDENTIAL PROBLEM OF EVIL HOWARD-SNYDER, MICHAEL BERGMANN, AND WILLIAM L. ROWE

157

that justify God in permitting human and animal life on earth to be nothing more than a series of agonizing moments from birth to death. But surely such a view is unreasonable, if not absurd. Surely there must be some point at which the appalling agony of human and animal existence on earth would render it unlikely that God exists. And this must be so even though we all agree that God's knowledge would far exceed our own. I believe my theistic friends have gone considerably beyond that point when in light of the enormous proliferation of horrendous evil in this world they continue to insist that we are unjustified in concluding that it is unlikely that God exists. They characterize my argument as a "noseeum" argument. But this is not quite correct. There are lots of things we can conceive of occurring in our world which we don't see occurring. My argument is basically a "noconceiveum" argument, not a "noseeum" argument. We cannot even conceive of goods that may occur and would justify God in permitting the terrible evils that afflict our world. Of course, being finite beings we can't expect to know all the goods God would know, any more than an amateur at chess should expect to know all the reasons for a particular move that Kasparov makes in a game. But, unlike Kasparov who in a chess match has a good reason not to tell us how a particular move fits into his plan to win the game, God, if he exists, isn't playing chess with our lives. In fact, since understanding the goods for the sake of which he permits terrible evils to befall us would itself enable us to better bear our suffering, God has a strong reason to help us understand those goods and how they require his permission of the terrible evils that befall us. My friends, however, do seem to think we can conceive of goods that may require God to permit at least some of these awful evils. They suggest that for all we know the following complex good may occur: the little five-year-old girl meets up with her rapist-killer somewhere in the next life, and he then repents and begs her forgiveness for savagely beating, raping, and strangling her, and she then forgives him with the result that both of them live happily ever after in the presence of God. What are we to make of this suggestion as to why God permitted the little girl to be brutally beaten, raped, and strangled? Well, they are right in holding that even God cannot bring about this complex good without permitting that individual to brutally beat, rape, and strangle the little girl. But that alone won't justify God in permitting that to happen to her. For it is eminently reasonable to believe that God could win the soul of the little girl's rapist-killer without having to permit him to do what he did to her. And even if

251

252

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 158

THE EVIDENTIAL PROBLEM OF EVIL

he can't, is it right for any being to permit the little girl to be robbed of her life in that way just so her killer could have something bad enough on his conscience to ultimately seek forgiveness? It is one thing to knowingly and freely give up one's life for the sake of another and quite another thing to have it ripped away, against one's will, just so someone else can later be led to repentance. If this is the best that can be done to find a good we know of that may justify God in permitting the little girl to be brutally beaten, raped, and strangled, the evidential argument from evil will surely remain a thorn in the side of theism for some time to come.

PART IV DIVINE FREEDOM

This page intentionally left blank

CHAPTER EIGHTEEN

The Problem of Divine Perfection and Freedom

Though God is a most perfectly free agent, he cannot but do always what is best and wisest in the whole. Samuel Clarke

Many thinkers in the theistic tradition have held that in addition to omnipotence and omniscience God's attributes must include perfect goodness and freedom. For the theistic God deserves unconditional gratitude, praise, and worship. But if a being were to fall short of perfect goodness, it would not be worthy of unreserved praise and worship. So, too, for divine freedom. If God were not free in some of his significant actions, if he always lacked the freedom not to do what he in fact does, we could hardly thank him or praise him for anything that he does. He would not be deserving of our gratitude and praise for the simple reason that he would act of necessity and not freely. So, along with omnipotence and omniscience, perfect goodness and significant freedom are fundamental attributes of the theistic God. Some attributes are essential to an object. That object could not exist were it not to possess those attributes. Other attributes are such that the object could still exist were it to lack them. If the theistic God does exist, to which class do his attributes of omnipotence, omniscience, perfect goodness, and freedom belong? Are they essential to that being? Does that being possess those attributes in every possible world in which he exists? Or are they not essential? Most thinkers in the theistic tradition have held the view that these attributes are constitutive of God's nature; they are essential attributes of the being that has them. With this view in mind, my aim in this essay is to consider the question of whether God's perfect goodness, specifically his moral perfection, is consistent with his being free in many significant actions. Throughout, we will suppose that if

256

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 224

William Rowe

God exists he is essentially omnipotent, omniscient, perfectly good, and free in many of his actions. What I want to determine is whether there is a serious difficulty in the endeavor to reconcile God's essential goodness and moral perfection with any significant degree of divine freedom.1 1 begin the investigation with the question of whether God is ever free to do an evil (morally wrong) act. The answer, I believe, is no. Of course, being morally perfect, omnipotent, and omniscient, God will never in fact do an evil act. No being who knowingly and willingly performs an evil act is morally perfect. Since being free to do an evil act is consistent with never in fact doing an evil act, it may seem initially plausible to think that God could be free to perform such an act. But if God is free to perform an evil act, then he has it in his power to perform that act. And, if God has it in his power to perform an evil act, then he has it in his power to deprive himself of one of his essential attributes (moral perfection). But no being has the power to deprive itself of one of its essential attributes.2 Therefore, God does not have it in his power to perform an evil act. The reasoning in this argument proceeds as follows. 1. God has it in his power to bring it about that he performs an evil act. (Assumption to be refuted) 2. From God's performing an evil act it follows that God is not morally perfect. 3. If X has it in its power to bring about p, q follows from p, and q does not obtain, then X has it in its power to bring about q. 4. God has it in his power to bring it about that he is not morally perfect. (From i to 3.) 5. Being morally perfect is an essential attribute of God. Therefore, 6. God has it in his power to bring it about that he lacks one of his essential attributes. Because (6) is clearly false, we must deny the initial assumption that God has power to bring it about that he performs an evil act. But if God does not have it in his power to perform an evil act, then performing an evil act is not something God is free to do. It may seem that my argument to show that God is not free to do 1

1 take God's moral perfection to be logically implied by his perfect goodness. It is understood here that a being has the power to deprive itself of a property only if it is possible for that being to lose that property and continue to exist. 2

DIVINE FREEDOM

The Problem of Divine Perfection and freedom

257

225

an evil act has already produced a serious difficulty in the theistic concept of God. For if God cannot do evil, what becomes of his omnipotence? After all, even we humans, with our quite limited power, are able to perform evil deeds. If God does not have the power to do what even we can do, how can we reasonably hold that he is essentially omnipotent? So long as we hold that omnipotence does not imply power to do what is not possible to be done, we need not conclude that God's lacking power to bring it about that he do something morally wrong renders him less than omnipotent.3 For, as we've seen, it is strictly impossible for a being who logically cannot be other than morally perfect to do something evil. If God is not free to do a morally wrong action, might he be free to do a morally right act? 'Morally right act' may mean either what is morally obligatory or what is morally permissible. If we are willing to countenance refraining from performing a certain action as an "action", then it is clear that God is not free with respect to performing any action that is morally obligatory for him to perform. For refraining from that action would be morally wrong, and, as we've seen, God is not free to do anything that is morally wrong for him to do. For God to be free in performing any action, it must be both in his power to perform it and in his power to refrain from performing it. But since refraining from doing what is morally obligatory is morally wrong, and being morally perfect is essential to God, he does not have the power to refrain from doing what is morally obligatory. In short, God does what is morally obligatory of necessity, not freely. The way I've just put the point about God doing what is morally obligatory for him to do is not quite right. For it suggests that even though God does not freely do what he is morally obligated to do, he nevertheless does (of necessity) what he has a moral obligation to do. But the truth is that no action is such that God can have a moral obligation to perform it. For one cannot have a moral obligation to do what one cannot do freely. If a person freely does some act, then it was in the person's power not to do it.4 Since it would be morally wrong for God not to do what he is morally obligated to do, it follows from my previous argument that God cannot do freely what he 3

For a discussion of this point see Joshua Hoffman, "Can God Do Evil?" Southern Journal of Philosophy 17 (1979): 213-20. 4 To avoid Frankfurt-type counterexamples, we should say that when an agent freely performs some action she caused her action and had the power not to cause that action. The power not to cause one's action is not the same as the power to prevent one's action. I ignore this complication in the text.

258

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 226

William Rowe

is morally obligated to do. But, since one cannot be morally obligated to do what one is not free to do, there are no actions God has a moral obligation to perform. At best we can say that God does of necessity those acts he would be morally obligated to do were he free to do them. We've seen that because he is essentially a morally perfect being, God is neither free to do a morally wrong action nor free in doing a morally obligatory action. We've also seen that no action can be such that God has a moral obligation to do it (or not to do it). It looks, then, as though my initial efforts have yielded the result that God's absolute moral perfection places significant restraints on the scope of divine freedom. However, the fact that God is neither free to do what is wrong nor free in doing what is morally obligatory (what would be morally obligatory were God free with respect to doing or not doing it) may still leave considerable scope for God's freedom to be exercised.5 For, so long as some of his important actions are morally permissible but not morally obligatory, we thus far have no reason to deny that God is free with respect to all such actions, that he has it in his power to do them and in his power not to do them. For example, it has long been held that God's action in creating the world was a free action, that God was free to create a world and free not to create a world.6 Creating the world is certainly a very significant act, involving, as it does, a vast number of divine acts in actualizing the contingent states of affairs that constitute our world. If God enjoys freedom with respect to the world he creates, then, although his freedom is constrained in ways that ours is not, there would not appear to be any insurmountable problem to reconciling perfection and freedom.7 To pursue the investigation of a possible conflict between God's moral perfection and his freedom, therefore, it will be helpful to turn our attention to God's action in creating the world. Specifically, we need to consider whether God's moral perfection leaves God free with respect to his creation of the world. 5

If we take significant freedom, as Alvin Plantinga does, to be the freedom to do or refrain from doing what is morally obligatory, it can be shown that God cannot be significantly free. See Wes Morriston, "Is God 'Significantly Free'?" Faith and Philosophy 2 (1983): 257-63. 6 For an excellent discussion and critical evaluation of the classical Judeo-Christian views on God's freedom with respect to creation see Norman Kretzmann, "A General Problem of Creation", and "A Particular Problem of Creation", in Being and Goodness, ed. Scott MacDonald (Ithaca and London: Cornell University Press, 1990). 7 There is a problem of understanding what God's moral perfection comes to, given that he can have no moral obligations. But this problem may not be unresolvable. See Thomas Morris, Anselmian Explorations (Notre Dame: University of Notre Dame Press, 1987), pp. 31-41.

DIVINE FREEDOM

The Problem of Divine Perfection and freedom

259

227

It is important to distinguish two questions concerning God's freedom in creating a world. There is the question of whether God is free to select among creatable worlds the one he will create.8 There is also the question of whether God is free not to create a world at all. That these are quite different questions can be seen as follows. Suppose that among worlds creatable by an omnipotent being there is one that is morally better than all other worlds. On this supposition, one can imagine arguments for any of four positions. Someone might argue that although God is free not to create a world at all, if he chooses to create, he must create the best world he can. Hence, although he is free not to create the morally best world, he is not free to create any world other than it. Alternatively, someone might argue that God's perfect goodness absolutely necessitates that he create a good world. God is not free not to create a world. But God's perfection does not necessitate that he create the best world he can. He is free to create among the class of creatable good worlds. Third, someone might argue that God enjoys both sorts of freedom. He is free not to create at all. He is also free to create some good world other than the best that he can create. Finally, one might argue that God's being essentially perfect necessitates his creating that world which is superior to all others. God is not free to create some world other than the best, and he is not free not to create any world at all. In what follows, I focus primarily on my first question (whether God is free to select among creatable worlds). In an important article, Robert Adams has argued that it need not be wrong for God to create a world that is not as good as some other world he could create.9 Adams supposes that the world God creates contains creatures each of whom is as happy as it is in any possible world in which it exists. Moreover, no creature in this world is so miserable that it would be better had it not existed. Let's suppose there is some other possible world, with different creatures, that exceeds this world in its degree of happiness, a world that God could have created. So, God has created a world with a lesser degree of happiness than he could have. Has God wronged anyone in creating this world? Adams argues that God cannot have wronged the creatures in the other possible world, for merely possible beings don't have rights. Nor can he have wronged the creatures in the world he has created, for their lives could not be made more happy. Adams notes that God would have done something wrong in creating this 8 Since there may be possible worlds that an omnipotent being cannot create, we need to restrict our discussion to the class of worlds creatable by God. 9 Robert Adams, "Must God Create the Best?" Philosophical Review 81 (1972): 317-32.

260

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 228

William Rowe

world were the following principle true: "It is wrong to bring into existence, knowingly, a being less excellent than one could have brought into existence."10 But this principle, Adams argues, is subject to counterexamples. Parents do no wrong when they refrain from taking drugs that would result in an abnormal gene structure in their children, even though taking the drugs would result in children who are superhuman both in intelligence and in prospects for happiness. Suppose we agree with Adams on these points. Suppose, that is, that we agree that God is not morally obligated to create the best world that he can, that it would be morally permissible for God to create the best world he can, but also morally permissible for God to create any of a number of other good worlds of the sort Adams describes. If so, can't we conclude that there is no unresolvable conflict between God's being essentially morally perfect and his enjoying a significant degree of genuine freedom? For it now appears that God's moral perfection does not require him to create the best world. In short, he is free to create (or not create) any of a number of good worlds. As forceful and persuasive as Adams's arguments are, I don't think they yield the conclusion that God's perfect goodness imposes no requirement on God to create the best world that he can create. What Adams's arguments show, at best, is that God's moral perfection imposes no moral obligation on God to create the best world he can. His arguments establish, at best, that God need not be doing anything morally wrong in creating some world other than the best world. But this isn't quite the same thing as showing that God's perfect goodness does not render it necessary that he create the best world he can. For, even conceding the points Adams tries to make, there still may be an inconsistency in a morally perfect being creating some world other than the best world he can create. My point here is this. One being may be morally better than another even though it is not better by virtue of the performance of some obligation that the other failed to perform. It may be morally better by virtue of performing some supererogatory act that the other being could have but did not perform. Analogously, a being who creates a better world than another being may be morally better, even though the being who creates the morally inferior world does not thereby do anything wrong. Following Philip Quinn, I'm inclined to think that if an omnipotent being creates some world other than the best 10

Ibid., p. 329.

DIVINE FREEDOM

The Problem of Divine Perfection and Freedom

261

229

world it can create, then it is possible there should exist a being morally better than it is.11 For it would be possible for there to be an omnipotent being who creates the best world that the first being could create but did not. I conclude then that if an essentially omnipotent, perfectly good being creates any world at all, it must create the best world it can. For although a being may do no wrong in creating less than the best it can create, a being whose nature is to be perfectly good is not such that it is possible for there to be a being morally better than it. If, however, a being were to create a world when there is a morally better world it could create, then it would be possible for there to be a being morally better than it. What we have seen is that a being who is morally perfect and creates a world must create the very best world it can create. But what if there is no best world among those it can create? This would be so in either of two cases. First, it might be that for any world it creates there is a morally better world it can create. Second, it might be that there is no unique best world. Perhaps, instead, there are many morally unsurpassable worlds among the worlds God can create. Let's consider these two cases in turn. On the assumption that for any world God creates there is a morally better world he can create, it is clear that it is impossible for God to do the best that he can. Whatever he does, it will be the case that he could have done better. This being so, it would seem only reasonable that God's perfect goodness is fully satisfied should he create a very good world. And we may safely assume that there are a large number of such worlds that he can create. So long as he creates one of these worlds, he will have satisfied the demands of his morally perfect nature. For the idea that he should create the best world he can is an idea that logically cannot be implemented. Hence, on the assumption of there being no morally unsurpassable world among the worlds God can create, it would seem that God's absolute moral perfection is fully compatible with his freely creating any one of a number of good worlds that lie in his power to create. To complain that God cannot then be perfect because he could have created a better world is to raise a complaint that no creative action 11 Philip L. Quinn, "God, Moral Perfection, and Possible Worlds", in God: The Contemporary Discussion, ed. Frederick Sontag and M. Darrol Bryant (New York: The Rose of Sharon Press, Inc., 1982), pp. 197-213. Quinn remarks: "An omnipotent moral agent can actualize any actualizable world. If he actualizes one than which there is a morally better, he does not do the best he can, morally speaking, and so it is possible that there is an agent morally better than he is, namely an omnipotent moral agent who actualizes one of those morally better worlds" (p. 213).

262

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 230

William Rowe

God took would have enabled him to avoid. As William Wainwright notes: The critic complains that God could have created a better order. But even if God had created a better order, He would be exposed to the possibility of a similar complaint. Indeed, no created order better than our own is such that God would not be exposed to the possibility of a complaint of this sort. The complaint is thus inappropriate. Even though there are an infinite number of created orders better than our own, God can't be faulted simply because He created an order inferior to other orders that He might have created in their place.12 There is something forceful and right about this reasoning. If, no matter what world an omnipotent being creates, there is a morally better world that being can create, then, provided that the omnipotent being creates a significantly good world, it cannot be morally at fault for not having created a morally better world. But our question is whether a being in such a situation can be an absolutely perfect being. And for reasons I have already uncovered, I think the answer is no. A being is necessarily an absolutely perfect moral being only if it is not possible for there to be a being morally better than it. If a being creates a world when there is some morally better world that it could have created, then it is possible that there be a being morally better than it. Since our assumption implies that for any world an omnipotent being creates there is a morally better world it can create, it follows that any such being who creates even a very good world cannot be an absolutely perfect moral being. Although the omnipotent being in question could be a very good moral agent and enjoy a significant degree of freedom in creating among a number of very good worlds, it could not be an absolutely perfect moral being. The existence of the theistic God who creates a world is inconsistent with the supposition that among the worlds he can create there is no morally unsurpassable world. Let's now consider the second way in which it could be true that no creatable world is better than all others. Suppose that among the worlds God can create there are a number of worlds that are morally unsurpassable. For reasons we've already considered, if God creates a world, he cannot create some world that is morally inferior to some other that he can create. Therefore, if there are a number of morally unsurpassable worlds among the worlds he can create, then 12 William Wainwright, Philosophy of Religion (Belmont, Calif.: Wadsworth Publishing Company, 1988), p. 90.

DIVINE FREEDOM

The Problem of Divine Perfection and Freedom

263

231

if he creates at all he must create one of these worlds. But unlike the case when there is exactly one morally unsurpassable world among the worlds he can create, here we do seem to have found a wedge to open up space for some degree of divine freedom to exist in harmony with God's absolute moral perfection. For God would seem to be free to create any one of the morally unsurpassable worlds. In any case, God's absolute moral perfection imposes no requirement on his creation among the set of morally unsurpassable worlds. Among the worlds creatable by an omnipotent being, either (i) there is exactly one morally unsurpassable world or (2) there is not. If there is not, then either (2a) there is no morally unsurpassable^ world or (2b) there are a number of morally unsurpassable worlds. If (i) is the case, then God is not free to select among creatable worlds. If (2a) is the case, God's being essentially morally perfect is inconsistent with his act of creating a world. If (2b) is the case, God's moral perfection leaves him free to select among morally unsurpassable worlds the one he will create. Earlier, I distinguished two questions concerning God's freedom in creating a world. We have been discussing the question of God's freedom to select among creatable worlds the one he will create. It is now time to consider the other question of whether God is free not to create a world at all. Here we may limit our inquiry to the possibility that there is exactly one morally unsurpassable world among the worlds creatable by an omnipotent being. As we've seen, in this case God is not free to select any other world to create. If God creates, he must create the one world that is morally best. Our present question is whether in this situation God is free not to create at all. Some possible world must be actual. What possible world would be actual if God exists but does not create at all? Presumably, it would be a world in which no positive, contingent state of affairs obtains. By 'a positive, contingent state of affairs' I mean any state of affairs such that from the fact that it obtains it follows that some contingent being (other than God, if he should be contingent) exists. Apart from God and whatever necessarily existing entities there are, in a world God inhabits but does not create no other being would exist. To answer our question of whether God is free not to create a world at all, we must compare the best world God can create with a world whose inhabitants are simply God and whatever necessarily existing entities there are. Assuming such a world would not be morally incommensurate with the morally unsurpassable world among worlds God can create, it is plausible to think that God is not free with respect to whether he will not create at all. For either the

264

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 232

William Rowe

world he inhabits but does not create is better than the best world he can create or it is not. If it is better, then he is not free not to create a world at all; he necessarily refrains from creating. If it is worse, then he is not free not to create a world at all; he necessarily creates a world.13 Might the world he inhabits but does not create be on a moral par with the best world among those an omnipotent being can create? If so, then, as in the case where there are a number of morally unsurpassable worlds omnipotence can create, we again have a wedge for creating space for some degree of divine freedom. But, in this scarcely possible situation, God's freedom would be restricted to creating the morally unsurpassable world or not creating at all. I now can draw together the results of my study of the problem of divine perfection and freedom. The conclusions may be presented as follows. 1. God is not free to perform any evil act, nor is he free in doing what is morally obligatory. 2. If there is a world creatable by God that is morally better than any world he inhabits but does not create, God is not free not to create a world at all. If he exists, he is a creator of necessity. 3. If for any world creatable by an omnipotent being there is another creatable world that is morally better, the theistic God cannot exist and be a creator of anything. 4. If there is a single, morally best creatable world, God enjoys neither sort of freedom: he is not free not to create and he is not free to select among creatable worlds.14 5. If there are a number of morally unsurpassable, creatable worlds, then, although God necessarily creates one or the other of them, he is free (so far as his moral perfection is concerned) to select among the morally unsurpassable worlds the one he will create.15 Traditional theists who hold that God is essentially perfect and yet possesses libertarian freedom of will and action have neglected, I believe, some of the implications that appear to follow from God's 13

This point is a plausible extension of our earlier principle: If X creates a morally inferior world to one X can create, then it is possible that there be a being morally better than X. Extending this principle, it is plausible to hold that if X can but does not create a world that is morally better than the one X inhabits, then it is possible that there be a being morally better than X. 14 The plausible assumption here is that if there is a single, morally best creatable world, then that world is morally better than any world God inhabits but does not create. 15 See n. 13.

DIVINE FREEDOM

The Problem of Divine Perfection and Freedom

265

233

perfect goodness. For all we know about possible worlds, it may well be that God's perfect goodness is inconsistent with any degree of divine freedom in whether he creates or what world he creates. In any case, it would seem that his perfection places rather severe limitations on the scope of his freedom in creating a world. The problem we have been considering is rooted in two basic points: God's perfect goodness is such as to preclude the possibility of a morally better being; God's freedom is such that he acts and wills freely only if it is in his power not to so act and will. If either point is significantly qualified or given up, the conclusions I've drawn may no longer obtain. If, for example, we give up the libertarian idea of freedom, then, following Jonathan Edwards, we might hold that God's actions are free even though necessitated by his perfect goodness.16 On the other hand, we might endeavor to qualify God's perfect goodness so that it permits the possibility for God to have been better than he is.17 Neither of these alternatives has been addressed in this essay. 16 Jonathan Edwards, Freedom of the Will, ed. Paul Ramsey (New Haven, Yale University Press, 1957). 17 See, for example, the suggestion by William Wainwright (drawn from Charles Hartshorne) that God's perfection might require only that he be unsurpassable by some other being (Philosophy of Religion, p. 9).

This page intentionally left blank

CHAPTER NINETEEN

Evil and God's Freedom in Creation

I JLf we think of possible worlds in the abstract we might think of them as divided into worlds that are on the whole good, worlds that are on the whole bad, and worlds that are on the whole neutral, neither good nor bad.1 How is the actual world to be classified? If we are theists, we will say that the actual world is a good world. For theists hold both that God is essentially omnipotent, omniscient, and perfectly good2 and that God's creative power is not limited to less than good worlds. Since a perfect being would choose to create a good world, rather than a world that is not good, we can conclude that if theism is true, the actual world is a good world.3 What then of the evils in the actual world? If there were some other creatable world that contained as much good as our world, but less evil, it would be a better world.4 But surely, God, being absolutely perfect, would then have created a world like that, rather than our world. Also, if there were another creatable world that contained no more evil than our world, but more good, it would be a better world. And surely a perfectly good being would have created a world like that, rather than the world that happens to be actual. So, if we are theists, it won't do to believe just that the actual

world is a good world. We should also believe that no creatable world with less evil than the actual world is better than the actual world. And we should also believe that no creatable world with more good than the actual world is better than the actual world. In short we should believe (1) that the amount of evil in our world is such that no less than it would have resulted in a better creatable world, and (2) that the amount of good in our world is such that no more of it would have resulted in a better creatable world. If we believe that theism is true, we should believe that the actual world, with all its horrendous evil, is the best (or one of the best) of all worlds God can create. If we assume that the theistic God exists, this Leibnizian conclusion concerning the character of the actual world is very plausible, if not demonstrable. It is not, however, without challenge. In an important article,5 Robert Adams argues that it need not be wrong for God to create a world that is not as good as some other world he could create. Adams supposes that the world God creates contains creatures each of whom is as happy as it is in any possible world in which it exists. Moreover, no creature in this world is so miserable that it would be better had it not existed. Let's suppose there is some other possible world with different creatures that

268

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 102 / AMERICAN PHILOSOPHICAL QUARTERLY

exceeds this world in its degree of happiness, a world that God could have created. So, God has created a world with a lesser degree of happiness than he could have. Has God wronged anyone in creating this world? Adams argues that God cannot have wronged the creatures in the other possible world, for merely possible beings don't have rights. Nor can he have wronged the creatures in the world he has created, for their lives could not be made more happy. Adams notes that God would have done something wrong in creating this world were the following principle true. It is wrong to bring into existence, knowingly, a being less excellent than one could have brought into existence.6 But this principle, Adams argues, is subject to counter-examples. Parents do no wrong, he suggests, when they refrain from taking drugs that would result in an abnormal gene structure in their children, even though taking the drugs would result in children who are superhuman both in intelligence and in prospects for happiness. As forceful and persuasive as Adams's arguments are, I don't think they yield the conclusion that God's perfect goodness imposes no requirement on God to create the best world he can. What Adams's arguments show, at best, is that God's perfection imposes no moral obligation on him to create the best world he can. His arguments establish, at best, that God need not be doing anything morally wrong in creating some world other than the best world. But this isn't quite the same thing as showing that God's perfect goodness does not render it necessary that he create the best world he can. For, even conceding the points Adams makes, there still may be an inconsistency in a perfectly good being creating some world other than the best world he can create. My point here is this. One being may be better than another

even though it is not better by virtue of the performance of some obligation that the other failed to perform. It may be better by virtue of performing some supererogatory act that the other being could have but did not perform. Analogously, a being who creates a better world than another being may be better, even though the being who creates the inferior world does not thereby do anything wrong. Indeed, I'm inclined to think that if an omnipotent being creates some world other than the best world it can create, then it is possible there should exist a being better than it is. For it would be possible for there to be an omnipotent being who creates the best world that the first being could create but did not. I conclude then that if an essentially omnipotent, perfectly good being creates any world at all, it must create the best world it can. For although a being may do no wrong in creating less than the best it can create, a being whose nature is to be perfectly good is not such that it is possible for there to be a being better than it. If, however, a being were to create a world when there is a better world it could create, then it would be possible for there to be a being better than it. In light of God's perfect goodness and unlimited power, suppose we grant that our Leibnizian conclusion is plausible, if not demonstrable. Suppose we grant, that is, that if theism is true then the actual world is a superlatively good world, meaning that there is no world God could have created that is better than it. I say that this is a Leibnizian conclusion, but it differs in two respects from Leibniz's own conclusion that the actual world is the best of all possible worlds. First, it allows that there are or may be possible worlds that God cannot create. Second, it allows that there may be creatable worlds as good as the actual world. Leibniz rejected the first of these because he was a compatibilist about freedom, and thus did not think that there are

DIVINE FREEDOM

269

EVIL AND GOD'S FREEDOM IN CREATION / 103

possible worlds that God cannot create by virtue of the fact that it might be up to some free creature which of several worlds is creatable by God. He rejected the second because God would not then have a sufficient reason to choose this world for creation over the other equally good worlds. So, for him, the actual world is the best of all possible worlds. But although we have weakened his conclusion, our conclusion has serious implications, I believe, for the problem of evil. In what follows I will sketch out its implications for the problem of evil. I will then return again to the topic of God's perfections and argue that there is a serious conflict, if not incoherence, within the theistic picture of God. II

Consider how we generally think about the problem of evil. We note that there are horrendous evils in the world. Perhaps we then reason that it is a good deal more likely that the world would contain all this evil if it is the product of chance than if it is the product of an infinitely powerful, good deity. For a good, omnipotent deity has a prima facie reason to prevent evil and the power to do so. In addition, we may reason that if God exists there must be goods that justify him in permitting all this evil. And we add that a good will justify God in permitting an evil only if it outweighs that evil and is such that he could not bring it about without permitting that evil (or some other evil just as bad).7 We then consider whether we have any reason to think that for all these horrendous evils there are outweighing goods that have occurred or will occur, goods that even an omnipotent being could not have brought about without permitting all these horrendous evils. Since this is bound to seem rather unlikely to us, we may conclude that it is unlikely that God exists. All this is very

familiar ground. But when we consider our Leibnizian conclusion, we begin to see dimensions to the problem of evil that are simply not addressed in our conventional ways of thinking about the problem. Our Leibnizian conclusion goes well beyond telling us that there are outweighing goods that an omnipotent being could not bring about without permitting all the instances of evil that occur in our world. On the assumption that our world has been created, it may strike us as a staggering idea to suppose that some great good would have been lost had any of the horrendous evils in our world been prevented by its creator. But this idea, staggering as it may be, is child's play compared to what our Leibnizian conclusion compels the theist to try to grasp and accept. For, given our Leibnizian conclusion, we must consider all the possible worlds that God could have created instead of our world with its horrendous evils. And we must then try to convince ourselves that none of those creatable worlds is better than this world. Let's consider how this enlarges the problem of evil. It will help us, I believe, if we characterize the conventional approach to the problem of evil as the similar past worlds problem of evil We can then think of the Leibnizian approach as the all worlds problem of evil. Let's first consider the similar past worlds problem of evil. Suppose we believe that there are God-justifying goods for each of the evils in the actual world. In the actual world a little 5-year-old girl in Flint, Michigan, was brutally beaten, raped, and strangled on New Year's Eve a few years ago. What is it to believe that there is a God-justifying good for God's allowing this tragedy to occur? I suggested earlier that a necessary condition for there being a God-justifying good for God's allowing some evil to occur is that there be some good that outweighs that evil and which

270

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

104 /AMERICAN PHILOSOPHICAL QUARTERLY

God would have forfeited had he prevented that evil. But, although necessary, this condition is not a sufficient. For it could be that permitting a particular evil is necessary for God to bring about a good that outweighs it; whereas preventing that very evil would enable God to bring about an even greater good. And if this were so, the outweighing good that God could not bring about without permitting that evil would not be a God-justifying good for his allowing that evil. What then would be a sufficient condition—or at least a condition considerably closer to a sufficient condition—for there being a God-justifying good for God's allowing some evil to occur? Perhaps we can give a rough characterization of such a condition as follows. There is a God-justifying good for God's permitting the little girl to be brutally beaten, raped, and strangled just in case the actual world would have been no better had God prevented that evil.8 How are we to understand the idea that the actual world would have been less good had God prevented that evil? My suggestion is that we consider those possible worlds that share (as much as possible) the same past with our world up to some point reasonably close to when that evil actually occurred, but in which God exists and, by one way or another, prevents that evil from occurring. If any of these worlds is both better than the actual world and creatable by God, we will say that there is no God-justifying good for God's permission of that evil. This is an instance of what I am calling the similar past worlds problem of evil. It is a problem for theism in the following way. Consider all those possible worlds with (as much as possible) the same past as the actual world but in which God exists and brings it about that the little girl dies peacefully in her sleep, thus preventing her tragic experience of being savagely beaten, raped,

and strangled. Presumably, some of these worlds are worse than the actual world. But surely, some of these worlds in which God brings it about that she dies peacefully in her sleep are better than the actual world. And isn't it altogether reasonable to believe that an omnipotent being could have created at least one of those worlds, rather than the actual world? Or consider all those possible worlds with (as much as possible) the same past as our world but in which God brings it about that her attacker is seized by a fatal heart attack moments before beating, raping, and strangling the little girl. Again, some of those worlds will be worse than the actual world, but many of them are undoubtedly better worlds than the actual world. And doesn't it seem just incredible that an omnipotent, omniscient being should have been incapable of creating one of them, rather than the actual world? Or consider all those possible worlds with (as much as possible) the same past as the actual world but in which God brings it about that the little girl's attacker is so overcome with sorrow on seeing the terror in her face that he lets her go, physically unharmed. Undoubtedly, some of these worlds are worse than the actual world and many are better. Is it reasonable to believe that an omnipotent being was unable to create any of these better worlds, rather than the actual world? It seems incredible that this should be so. As difficult as the similar past worlds problem of evil may be for theism, I suggest that our Leibnizian conclusion points to an even more severe difficulty, what I have called the all worlds problem of evil. Our Leibnizian conclusion was that the actual world, with all its horrendous evil, is the best (or one of the best) of all worlds God can create. In light of this conclusion, what must we now believe about the tragic evil of the little girl's being savagely

DIVINE FREEDOM

271

EVIL AND GOD'S FREEDOM IN CREATION / 105 beaten, raped, and strangled? What we must believe is that no possible world in which that little girl is not brutally beaten, raped, and strangled is better than the actual world and creatable by God. In the similar past worlds problem of evil we only had to compare the actual world with worlds like ours up to the point of the evil, but in which God exists and prevents that evil. But in the all possible worlds problem of evil we must make a judgment about every possible world. Consider, then, the class of possible worlds. Within this class, of course, is the actual world with all its horrendous evil, including the evil of the little girl being beaten, raped, and strangled. I submit that the belief that no possible world is better than the actual world is an absolutely absurd belief. Indeed, the idea that our world with all its horrendous evil is just as good, if not better, than any other possible world is an idea that merits the ridicule Voltaire heaped upon it. And the same goes, I believe, for the view that no possible world creatable by God is better than the actual world. The idea that an omnipotent, omniscient being should be capable of creating the actual world but incapable of creating any of the possible worlds better than it is, I think, bound to strike us as absurd, at least when we seriously reflect on the horrific evils played out daily on the stage of the actual world. If some argument validly yields that idea, one can only be highly suspicious of one or more of its premises. According to Thomas Reid, absurd ideas are fit objects of ridicule. Nature, Reid suggests, has given us the emotion of ridicule for the purpose of ridding ourselves of ideas that are absurd. He tells us that This weapon, when properly applied, cuts with as keen an edge as argument. Nature hath furnished us with the first to expose absurdity; as with the last to refute error.

Both ... are equally friendly to truth when properly used.9 He warns us that absurdities that are tinged with religious feeling are resistant to the force of ridicule, but assures us that if ever we are able to view [an absurdity] naked and stripped of those adventitious circumstances from which it borrowed its importance and authority, the natural emotion of ridicule will exert its force.10 My own view is that the idea that the actual world is the best possible world is an absurdity, fit for ridicule. And I think too that the idea that the actual world, with all its horrendous evil, is the best (or one of the best) of all worlds creatable by an omnipotent, omniscient being is also deserving of ridicule. If theism leads to such a conclusion, we should abandon it, unless we have very strong grounds in support of theism. I should add here that the theist who agrees with me that such ideas are absurd, deserving of ridicule, has several options to consider before concluding that the theistic God does not exist. First, the theist might argue that there are incommensurable worlds with respect to value, worlds such that it simply isn't true that one is better than, worse than, or equal in value to another. If this were so, the idea that one world is better than, or as good as, all others would itself be incoherent. Second, and perhaps more plausibly, the theist may reject the idea that among creatable worlds there is a superlatively good world. As is well known, alongside the Leibnizian tradition stands the Thomistic tradition, according to which for any possible world there is a better possible world. If the Thomistic tradition is correct, it may well be impossible for the actual world to be a superlatively good world, with the result that our Leibnizian conclusion cannot be reached. Third, the theist may reject the

272

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 106 /AMERICAN PHILOSOPHICAL QUARTERLY

idea that if there is a best world that a perfect being can create, then that being necessarily creates the best world, arguing instead that the fact that one world is better than another need not be a determining reason for a perfect being to create it instead of the lesser world. And perhaps there are other objections that can be raised as well.11 The general point we must recognize here is that any formally valid argument from the existence of the theistic God to the absurd conclusion that the actual world is the best possible world, or the best creatable world, will have to appeal to additional premises, premises that may then be rejected by the theist, rather than the premise that the theistic God exists.

Ill Let's return now to the topic of God's perfections. According to theism, God's essence includes freedom with respect to creation: he was free to create or not to create, and he was free to create worlds other than the world he did create. If God were not free in creation, we could hardly make sense of thanking him or praising him for his creation. He would not be deserving of our gratitude and praise for the simple reason that he would create of necessity and not freely. So, along with omnipotence, omniscience, and perfect goodness, freedom with respect to creation is a fundamental attribute of the theistic God. But, as we shall see, God's perfect goodness places severe limitations on God's freedom with respect to creation. In spite of holding both that this is the best of all possible worlds and that God must choose to create the best, Leibniz argued that God's choice to create the best world was free. I'm convinced that he was mistaken in this view.12 But I want to consider here the view of another important 18th century figure, Samuel Clarke. For,

unlike Leibniz, Clarke was a libertarian about freedom, holding that a choice is free only if the agent could have chosen otherwise. Since many theists nowadays tend to hold a libertarian view of freedom not unlike Clarke's, it will be more germane to our inquiry to examine his attempt to reconcile God's freedom in creation with the necessity of God's choosing the best. What Clarke has to say about this matter is, I believe, representative of a major strand of thought within theism. IV

In contrast to Locke, who characterized freedom as the power to carry out the action that we choose (will) to do, leaving the choice (volition) itself to be causally necessitated by the agent's motives, 13 Clarke locates freedom squarely at the level of the choice to act or not act. "The Essence of Liberty consists in [a person's] having a continual Power of choosing, whether he shall Act, or whether he shall forbear Acting."14 The implication of Clarke's view is that freedom (liberty) would be impossible should a person's choices be causally necessitated by his motives or desires. For if a person's choice to act is causally necessitated by earlier states of his body or mind, then at the time of that choice it was not in the agent's power to choose to not act. It is for Clarke a secondary matter whether the agent is able to carry out his choice.15 Of course, since God is omnipotent, his power to carry out the action he chooses to do is unlimited. But our question is whether God has it in his power to choose to refrain from following what he knows to be the best course of action. Should he lack that power, it follows from Clarke's conception of freedom that God does not freely choose the best course of action. In fact, it would follow for Clarke that in this instance God is

DIVINE FREEDOM

273

EVIL AND GOD'S FREEDOM IN CREATION / 107

totally passive and not an agent at all. It would also follow for Clarke that it would make no sense to praise or thank God for choosing the best course of action. We must now see how Clarke endeavors to avoid the absolute necessity of God's choosing in accordance with his knowledge of what is the best course of action. Clarke's overall view is clear enough. He distinguishes between the intellect (understanding) and the will. It is the function of the understanding to determine what course of action to pursue. It is the function of the will (the power we have to will this or that) to initiate the action specified by the understanding. It is one thing, however, to arrive (through deliberation) at the judgment that doing a certain thing is best,16 and quite another thing to choose (will) to do that thing. Often enough, our motives and desires are sufficiently clear and strong to causally necessitate the judgment as to what to do. No other judgment is possible in the circumstances. In short, there may be no freedom at all with respect to the judgment as to what action to perform. On Clarke's view, freedom enters only when the will chooses to act or not act in accordance with the judgment of the understanding. Thus, when there is a best course of action for God to perform, his judgment that it is the best course to pursue is, Clarke tells us, absolutely necessary.11 But God's choice to act in accordance with what his understanding approves is completely free; he always has the power to choose otherwise. God always discerns and approves what is Just and Good, necessarily, and cannot do otherwise: But he always acts or does what is Just and Good freely; that is, having at the same Time a full natural or physical Power of acting differently.18

It is instructive to contrast Clarke's view of freedom with a stream of thought in

Christian theology, dating back at least to Augustine, according to which the saints in heaven are perfected to the degree that they not only do not sin, they no longer are able to sin, a perfection that is found in God and the angels. In our earthly state we have the freedom to turn from the good and do evil, but in the life to come we shall have a superior sort of freedom, a freedom that does not include the ability to do evil. Thus Augustine says: For the first freedom of will which man received when he was created upright consisted in an ability not to sin, but also in an ability to sin; whereas this last freedom of will shall be superior, inasmuch as it shall not be able to sin. This, indeed, shall not be a natural ability, but the gift of God. For it is one thing to be God, another thing to be a partaker of God. God by nature cannot sin, but the partaker of God receives this inability from God.19

In his book, A Philosophical Inquiry Concerning Human Liberty, Anthony Collins had appealed to this stream of thought in support of his view that freedom does not require any power to choose or do otherwise. Clarke wrote a rather devastating response to Collins's book.20 In the course of his response to Collins we find the following remark: Neither Saints, nor Angels, nor God himself, have in any Degree the less Liberty, upon Account of the Perfection of their Nature: Because, between the physical Power of Action, and the Perfection of Judgment which is not Action, (which two things this Author constantly confounds,) there is no Connection. God judges what is right, and approves what is good, by a physical Necessity of nature; in which physical Necessity, all Notion of Action is necessarily excluded: But doing what is good, is wholly owing to an Active Principle, in which is essentially included the Notion of Liberty.21

274

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 108 /AMERICAN PHILOSOPHICAL QUARTERLY

Clearly, Clarke rejects this stream of thought in Christian theology. He allows that the saints in heaven no longer have any desire to sin and take no delight in it. Indeed, it may be absolutely certain that with purified desires and a perfected judgment they will always freely do what is right.22 And this will be an enormous difference from life on earth where we are often tempted to sin by bad desires and faulty judgment. But what cannot be is that the saints or the angels, or God for that matter, cease to have the ability or power to choose to do other than what is right. For then they would not be free in choosing and doing what is right. To lose the power to choose otherwise is to lose the power to choose freely. And if one loses that power one ceases to be an agent at all. We can begin to get at the difficulty in Clarke's view of divine freedom by considering God's perfections and their implications for whether he can freely choose to do evil. Clarke readily sees that were a perfectly good, omniscient being to freely choose to do some evil deed, it would thereby cease to be perfectly good. And it would cease to be perfectly good even if, as could not happen in God's case, it were prevented from carrying out the evil deed it chose to do. For the free choice to do evil is itself inconsistent with continuing to be a perfectly good, omniscient being. A being who freely chooses to do what it knows to be an evil deed thereby ceases to be a perfectly good being. So, if God were to freely choose to do an evil deed, he would cease to be perfectly good. In short, it is not logically possible for God both to freely choose to do evil and to continue to be perfectly good. Now, since Clarke holds with Leibniz that God necessarily exists and necessarily is omnipotent, omniscient, and perfectly good,23 we can advance to the simpler conclusion that it

is not logically possible for God to freely choose to do evil. It is not logically possible because it is inconsistent with what is logically necessary: the existence of a being (God) who is necessarily omnipotent, omniscient, and perfectly good. Consider now the question: Does God ever freely choose not to do evil? I think we can see that Clarke's own views commit him to a negative answer to this question. For God chooses freely not to do something only if it is in his power to choose to do that thing—choosing freely, Clarke insists, logically requires the power to choose otherwise. But it cannot be in anyone's power to make a certain choice if it is logically impossible that the person make that choice.24 Therefore, since it is logically impossible for God to choose to do evil, it is not in God's power to choose to do evil. And since it is not in God's power to choose to do evil, it cannot be that God's choice not to do evil is a. free choice. If God chooses not to do evil, he so chooses of necessity, not freely. And this being so, it makes no sense for us to thank God, or to be grateful to him, for choosing not to do evil. He could not have chosen otherwise. Before turning to what I regard as two attempts by Clarke to avoid any limitations on the scope of divine liberty, we should consider whether it is in God's power to choose contrary to what he judges to be best. We've concluded that God cannot choose to do evil. But to choose contrary to what is judged to be best is evil or morally wrong only if choosing to do what is judged to be best is morally obligatory. To claim that it is morally obligatory ignores the real possibility that choosing what is best is supererogatory, beyond the call of duty. There are choices which are good to make but not required as our duty. It would be a mistake, therefore, to infer God's

DIVINE FREEDOM

275

EVIL AND GOD'S FREEDOM IN CREATION / 109

inability to choose to act contrary to what he judges to be best from his inability to choose to do evil. Nevertheless, it does seem to be logically impossible for perfect goodness to choose to act contrary to what is best. And this seems to be Clarke's own view of the matter. Thus he declares "that though God is a most perfectly free agent, yet he cannot but do always what is best and -wisest in the whole."25 To choose otherwise, he thinks, is to act contrary to perfect wisdom and goodness. How does Clarke endeavor to avoid the conclusion that God's perfect goodness precludes his being free in many of his choices? His general approach to this difficulty is to distinguish two sorts of necessities: moral and physical. If one state or event physically necessitates another state or event, then the second state or event cannot occur freely. Thus he would say that hanging a greater weight on the left end of an accurate balance physically necessitates the downward movement on the left side of the balance. Here, even if the balance were endowed with consciousness, there would be no possibility of the balance freely moving downward on the left. For the balance has no power to do other than move downward on the left side. To illustrate the other sort of necessity, he offers the example of God's promising that on a given day he will not destroy the world. The promise morally necessitates God's refraining from destroying the world on that particular day. But, says Clarke, it would be absurd to think that God therefore lacked the physical power on that day to destroy the world.26 God's refraining from destroying the world on that day is both morally necessary and free. For he both retains the physical power to destroy the world on that day and also cannot (morally speaking) break his promise.27

The trouble with this solution is that it doesn't focus on the particular act of choosing to break his promise. If we accept, as it seems we must in God's case, that it is logically impossible for God to choose to break his solemn promise, then it follows that it is not in God's power to break his solemn promise. Indeed, for God to choose to break his solemn promise is for God to divest himself of his absolute perfection. And, clearly, it is not logically possible for God to cease to be absolutely perfect. In a revealing passage Clarke appears to recognize that there are some choices that do not lie within God's power because they logically imply the destruction of his essential perfections. He begins the passage by noting that there are necessary relations among things, relations that God eternally knows. By this he means that some states of things are necessarily better than other states of things. (For example, there being innocent beings who do not suffer eternally is necessarily better than there being innocent beings who do suffer eternally.) By knowing these necessary relations, God knows the choices required by his perfect wisdom and goodness. Noting that God cannot but choose to act always according to this knowledge, he continues: It being as truly impossible for such a Free Agent, who is absolutely incapable of being Deceived or Depraved, to Choose, by acting contrary to these Laws, to destroy its own Perfections', as for Necessary Existence to be able to destroy its own Being.2*

He then draws the obvious conclusion that From hence it follows, that though God is both Perfectly Free, and also Infinitely Powerful, yet he cannot Possibly do any thing that is Evil. The Reason of this also, is Evident. Because, as 'tis manifest Infinite Power cannot extend to Natural Contradictions, which imply a Destruction of that very Power by which they must be supposed to be effected; so neither can it

276

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 110 /AMERICAN PHILOSOPHICAL QUARTERLY extend to Moral Contradictions, which imply a Destruction of some other Attributes, as necessarily belonging to the Divine Nature as Power. I have already shown that Justice, Goodness and Truth are necessarily in God; even as necessarily, as Power and Understanding, and Knowledge of the nature of Things. Tis therefore as Impossible and Contradictory, to suppose his Will should Choose to do any thing contrary, to Justice, Goodness or Truth; as that his Power should be able, to do any thing inconsistent with Power.29

The conclusion I draw from these remarks is that God's liberty is curtailed by his perfect goodness. If choosing to do something rules out his being perfectly good, then it is not in his power to choose to do that thing. He necessarily, not freely, chooses not to do that thing. This is the conclusion we argued for above. Clarke, however, rejects this conclusion, insisting instead that God's liberty is not in the least diminished. 'Tis is no Diminution of Power, not to be able to Do things which are no Object of Power: And 'tis in like manner no Diminution either of Power or Liberty, to have such a Perfect and Unalterable Rectitude of Will as never Possibly to Choose to do anything inconsistent with that Rectitude.30

Perhaps his reasoning here can be understood as follows. There are some things God cannot do. He cannot make a square circle. Nor can he choose to do evil. But in neither case is his inability due to his power being limited. For it isn't true that there is some degree of power he is lacking such that were he to have possessed that extra degree he could have done these things. There is no degree of power sufficient for a being to make a square circle. And there is no degree of power sufficient for an essentially perfect being to choose to do evil.31 So, the fact that God is unable to make a square circle or choose to do evil is consistent with his possession of perfect

power. Hence, that fact does not imply any diminution of power. And, by analogy, Clarke infers that it does not imply any diminution of liberty in God. Suppose we agree with Clarke that God's inability to choose to do evil is not a diminution of power. And suppose we grant him the inference that it is also not a diminution of freedom? We will agree, that is, that there is no degree of freedom that God lacks such that were he to possess it he would be free to choose to do evil. Will this solve the problem before us? No. Remember, on Clarke's account of the nature of freedom, the power to choose otherwise is necessary for a choice to be free. Therefore, since, as Clarke agrees, it is not in God's power to choose to do evil, God does not freely choose not to do evil. And if it is not in God's power to choose to act contrary to what is best, God does not freely choose to do what is best. So long as he lacks the power to choose to do evil, he lacks freedom in choosing not to do evil. And so long as he lacks the power to choose contrary to what is best, he lacks freedom in choosing to do what is best. It won't matter whether this lack of power results from a deficiency in his power or from the fact that his perfect power does not extend to such choices. For, since God's choosing to do what is best is absolutely necessary, it follows that God chooses to do what is best of necessity, not freely. And this means that we are left with no reason whatever to thank God or be grateful to him for choosing and acting in accordance with his knowledge of what is best. I conclude that Clarke's valiant effort to reconcile God's perfect liberty with his perfect goodness is unsuccessful. For his perfect goodness renders his choice of the best unfree and renders it senseless for us to thank him, or to be grateful to him, for choosing and doing what is best.32

DIVINE FREEDOM

277

EVIL AND GOD'S FREEDOM IN CREATION /111 V In this paper F ve discussed two problems in theism arising from the fact that God is necessarily perfectly good, omniscient, and omnipotent. The first problem is that we seem inevitably led to the absurd conclusion that this world, with all its unspeakable horrors, is either the best of all possible worlds or a world than which no better world is creatable. The second

problem is that God's perfect goodness results in the necessity of his always doing what he judges to be best, thus depriving him of the freedom not to do what is best an j depriving us of any grounds to praise him or thank him for so acting. Each of these problems, I've suggested, casts seriOus doubts on the plausibility of the theistic picture of God.

NOTES 1. Some may question whether theism permits the existence of possible worlds that are not good. I ignore that complication here. 2. "Theism" has both a broad and a narrow sense. In its broad sense, theism is covers any view which holds that there is a divine reality. In its narrow sense, theism is the view that the divine reality is a person that is perfectly good, all-powerful, and all-knowing. In this paper I use "theism" to express a refinement of its narrow sense. 3. In this paper I assume (i) that human freedom is inconsistent with causal determinism, and (ii) that God enjoys middle knowledge (the knowledge of counterfactuals of freedom). Both assumptions are controversial. The first assumption is favorable to theism. For if human freedom is consistent with causal determinism, God could have created a world in which humans are free to do good or evil but always do what is good. It is, however, very difficult not to suppose that some such world is better than our world. The second assumption is less favorable to theism. Although (ii) implies that there are some worlds God cannot create ("uncreatable" worlds), it deprives us of excusing God for creating the world he did on the ground that, lacking middle knowledge, he could not be sure just what world would result from his creative activity. On the other hand, if we deny (ii) we may leave open the possibility that God plays dice in creating a world. 4. Perhaps a world can be qualitatively better than another world even though it is not quantitatively better than that world. I will ignore this complication here. 5. "Must God Create the Best?" Philosophical Review 81 (1972): 317-332. 6. Ibid., p. 329. 7. For sake of simplicity I will ignore the possibility that what justifies God in permitting a certain evil is that he thereby is able to prevent an even greater evil. 8. By "the actual world would have been no better" I mean here that the world that would have been actual had God prevented that evil is no better than the world happens to be actual. 9. Essays on the Intellectual Powers, 8th edition, edited by Sir William Hamilton (Edinburgh: James Thin, 1895), Essay VI. 10.Ibid. 11. For example, one may reject the assumption of God's middle knowledge, as noted in note 4.

278

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

1127 AMERICAN PHILOSOPHICAL QUARTERLY 12. My argument to show that Leibniz was mistaken in this view is a common one and can be found in "Clarke and Leibniz on Divine Perfection and Freedom," Enlightenment and Dissent (Special issue on Samuel Clarke), No. 16, 1997, 60-82. 13. See An Essay Concerning Human Understanding, ed. Peter H. Nidditch (Oxford: Oxford University Press, 1979), book 2, chap. 21. 14. Works, II, 565-566. References to Clarke's writings other than his correspondence with Leibniz are to the reprinting of his Works (1738) in four volumes in British Philosophers and Theologians of the 17th and 18th Centuries (New York: Garland, 1978). 15. Clarke goes so far as to declare that a prisoner in chains is free to will to leave or will to stay. That he cannot successfully execute his choice doesn't rob him of the power to choose (Works, II, 565). (Of course, he would allow that one who knows he is in chains may well see the pointlessness of choosing to leave and, therefore, not exercise his power so to choose.) 16. Since such a judgment terminates the process of deliberation about what to do, Clarke and others referred to it as "the last judgment of the understanding." It is the judgment that terminates deliberation and is followed by the act of will to perform (or not perform) the action specified in the judgment. 17. Clarke uses several different expressions to designate the sort of necessity that precludes freedom to do otherwise. His favorite expressions are "physical necessity" and "natural necessity." But he also uses "absolute necessity" on occasion. And in his fifth letter in response to Leibniz he says: "Necessity, in philosophical questions, always signifies absolute necessity." The Leibniz-Clarke Correspondence, ed. H. G. Alexander (Manchester: Manchester University Press, 1956), p. 99. 18. In his answer to the third Letter from the "Gentleman from Cambridge," Jan. 12, 1716-17. See Works, IV, 717. 19. The City of God, Book XXII, chap. XXX, trans. M. Dods (New York: Random House, 1993), p. 865. 20. For a discussion of Collins's book and Clarke's critical response, see chap. 3 of my Thomas Reid on Freedom and Morality (Ithaca: Cornell University Press, 1991). 21. Dr. Clarke's Remarks upon a Book entitled A Philosophical Enquiry Concerning Human Liberty. See Clarke's Works, IV, 731. 22. See Works, II, 575 for his clear statement of this view. 23. An essential attribute of a being is an attribute that the being necessarily possesses. Clarke holds that the moral perfections of the deity are essential aspects of the divine nature. "Justice, Goodness, and all the other Moral Attributes of God are as Essential to the Divine Nature, as the Natural Attributes of Eternity, Infinity, and the like." Works, II, 574. 24. If there is no possible world in which a person makes a certain choice, it cannot be that the person, nevertheless, has it within his power to make that choice. 25. Works, IV, 574. 26. "God's performing his Promise is ALWAYS consequent upon his making it: Yet there is no Connection between them, as between Cause and Effect: For, not the Promise of God, but his Active Power, is the alone Physical or efficient Cause of the Performance." See Dr. Clarke's Remarks upon a Book entitled A Philosophical Enquiry Concerning Human Liberty, Works, IV, 723.

DIVINE FREEDOM

279

EVIL AND GOD'S FREEDOM IN CREATION / 113 27. Clarke may well be right in holding that God had the physical power on that day to destroy the world. But it won't follow that it was in God's power to destroy the world on that day. For having the physical power to do something is consistent with not being able to do that thing. I have the physical power to murder my grandchild tomorrow. But it is not in my power to murder her tomorrow, for I am unable to choose to murder her. Intentional acts can be done only if we are able to choose to do them and possess the physical strength or power to do them. And given that in the past God has solemnly promised not to destroy the world tomorrow, he does not now have it in his power to choose to destroy it tomorrow. Prior to his promising, he may have had it in his power to so arrange things that he chooses to and does destroy the world tomorrow. But ever since he promised, it has not been in his power to choose to destroy the world tomorrow. For he cannot change the past, he cannot bring it about that he did not so promise then. And since his choosing to destroy the world tomorrow entails that he did not promise in the past not to destroy it tomorrow, it is in his power to destroy it only if it is in his power to bring it about that he did not promise to destroy it. But the latter is a power God does not possess, since it involves making it the case that something that took place in the past did not take place. 28. Works, II, 574. 29. Ibid. 30. Works, II, 574-575. 31. It is worth noting that although no being has the power to make a square circle, many beings have the power to choose to do evil. It might be claimed that the latter fact implies that there is a possible power God lacks. Although this point is worthy of consideration, nothing I say depends on it. 32. Portions of Section IV in this paper are taken from my discussion of Clarke in "Clarke and Leibniz on Divine Perfection and Freedom," Enlightenment and Dissent (Special Issue on Samuel Clarke), No. 16, 1997, 60-82.

This page intentionally left blank

CHAPTER TWENTY

Can God Be Free?

The major conception of God in the West is that he necessarily exists and is necessarily all-powerful, all-knowing, and perfectly good. Can God, so conceived, be free with respect to the possible world he selects to create? I argue that if there is a best creatable world, God is not free to create any world other than the best world. I also argue that if for every creatable world there is a better creatable world, it cannot be that God exists and is the creator of a world.

The question, Can God Be Free?, is an important philosophical question. But before endeavoring to answer this question, there are two preliminary points that must be discussed. The first, and most important, is this: What conception of God is being presupposed when we ask whether or not God can be free? The conception of God presupposed here is the idea of God that has been dominant in the major religions of the West - Judaism, Christianity, and Islam. The dominant idea of God in these religious traditions is of a being who necessarily exists and is necessarily all-powerful, all-knowing, and perfectly good. Thus when we ask whether God exists, we here mean to be asking whether there is a being who necessarily exists and is necessarily all-powerful, all-knowing, and perfectly good. Of course, it is no easy matter to say in any precise way just what it is for a being to be all-powerful, allknowing, and perfectly good. Nevertheless, we will presuppose here that some plausible account can be given of these attributes. Second, when we ask whether God can be free, we need to first ask: "free with respect to what?" Clearly, in an important sense God is not free with respect to doing evil, for if he were free with respect to doing evil he would be free to cease to be perfectly good. But if there is such a being as God he is no more free to cease to be perfectly good than he is free to cease to exist, to cease to be allpowerful, or to cease to be all-knowing. God is necessarily perfectly good, all-powerful, and all-knowing. He is, therefore, not free to cease to possess these perfections. Hence, in our question "Can God be Free?" we are not asking whether God can be free with respect to his essential attributes. For the answer to that question is dearly negative. What then are we asking about when we ask whether God can be free? We are asking whether God is free with respect to creating a world. And one persistent theme in the great religious traditions of the West is that God does enjoy freedom with respect to creating a world. This freedom is thought to be twofold:

282

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 406

Faith and Philosophy

1. God was free to refrain from creating any world at all; and,

2. God was free to create other worlds instead of the world he did in fact create. So, supposing there is such a being as God, the question we shall explore here is whether God was free not to create at all, and free to have created other worlds than the one he has created. If we assume with Leibniz that among possible worlds there is one that is best, it is difficult to see how the best possible being (God) would be free to create some world other than the best. But before we pursue the question of whether God must create the best possible world, it will be helpful here to pause in our discussion and try to get clearer about the notion of a possible world and some related concepts. Often we think that although things are a certain way they didn't have to be that way, they could have been different. Suppose we were late getting to class today. We believe, however, that things could have turned out differently. Had we not stopped on the way to chat with a friend, for example, we almost certainly would have been on time. So, what we may call the actual state of affairs, our being late for dass, need not have been actual at all. Here then we make a distinction between two possible states of affairs - ways things might be - and note that although one is actual (our being late for class today), the other (our being on time for class) was possible, it could have been actual instead. The link between possible states of affairs that are not actual and our common ways of thinking about the world lies in our frequent belief that things could have been otherwise. Whenever we correctly think that things could have gone in a way different from the way they actually went, we are distinguishing between some possible state of affairs that is actual (the way things did go) and some possible state of affairs that didn't become actual (the way things could have gone but didn't). Every state of affairs that is actual is clearly a possible state of affairs, one that, logically speaking, could be actual. But, as we've seen in the example of our not being late for class, possible states of affairs may fail to be actualized. Perhaps, then, we should think of a possible state of affairs as one that could be actual and could fail to be actual. But this view overlooks a useful distinction philosophers draw between a state of affairs that is possible and a state of affairs that is contingent. A contingent state of affairs is a possible state of affairs that may be actual or fail to be actual. Since a possible state of affairs is one that could be actual, if it is also such that it could fail to be actual (like, for example, our being on time for dass today), then it is a contingent state of affairs. It can be actual and can fail to be actual. But some states of affairs are such that although they are possible, and therefore can be actual, they cannot fail to be actual, they must be actual. These possible states of affairs are necessary, not contingent. Consider the state of affairs consisting in the number three's being larger than the number two. Certainly, this state of affairs is possible - it's not like there being an object that is both square and round, an impossible state of affairs. So, it is a possible state of affairs. But is it contingent? Could it have

DIVINE FREEDOM CAN GOD BE FREE?

283 407

failed to be actual. No. The number three's being larger than the number two is not just possible, it is also necessary, it obtains in every possible world. So, while many possible states of affairs are contingent in that they obtain in some worlds but not in others,1 many possible states of affairs are necessary, they obtain in every possible world. The following diagram shows the way philosophers often distinguish among states of affairs. STATES OF AFFAIRS I

I POSSIBLE actual in at least some worlds I

NECESSARY actual in every world

I IMPOSSIBLE actual in no worlds

CONTINGENT actual in some worlds and not actual in other worlds

In order to grasp the idea of a possible world it is helpful to consider two important relations among states of affairs: inclusion and preclusion. A state of affairs S includes a state of affairs S* just in case it is impossible that S should obtain and S* not obtain. (For example, Gordie Howe's being the greatest hockey player of the 20th century includes someone's being the greatest hockey player of the 20th century.) S precludes S* just in case it is impossible that S obtain and S* obtain. (So, Gordie Howe's being the greatest hockey player of the 20th century precludes Wayne Gretsky's being the greatest hockey player of the 20th century.) Following Alvin Plantinga, we can now say what it is for a state of affairs to be maximal and, therefore, a possible world. "A State of affairs S is ... maximal if for every state of affairs S', S includes S' or S precludes S'. And a possible world is simply a possible state of affairs that is maximal/'2 Having seen that a possible world is a maximal state of affairs, we can now consider what it is for a possible world to be better than some other possible world. Some states of affairs may be said to be intrinsically better than other states of affairs. For example, following Samuel Clarke we may say that there being innocent beings who do not suffer eternally is necessarily better than there being innocent beings who do suffer eternally. Of the second of these two states of affairs we would say that it is a bad state of affairs, something that ought not to be. But the first state of affairs is not a bad state of affairs. The basic idea here is that some states of affairs possess intrinsic value. That is, they may be intrinsically good, intrinsically bad, or intrinsically neutral (neither good nor bad). They are intrinsically good by virtue of containing intrinsically good qualities such as happiness, love, enjoyment, beauty, good intentions, or the exercise of virtue.3 And states of affairs are intrinsically bad by virtue of containing intrinsically bad qualities such as unhappiness, hate, dissatisfaction, ugliness, bad intentions, or the exercise of vice. Still other states of affairs, may contain little or no intrinsic value. There being stones, for example, is a state of affairs that contains little if any intrinsic value. Such states of affairs are, we might say,

284

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 408

Faith and Philosophy

intrinsically neutral. But someone's being happy, for example, is an intrinsically good state of affairs, while someone's being unhappy is an intrinsically bad state of affairs. One might infer from the preceding paragraph that if God exists, the world he creates would not include any bad states of affairs.4 However, supposing it would be in God's power to create such a world, there are at least two reasons to question this inference. First, as theodicists have argued since the time of Augustine, freedom of the will, if not itself a great intrinsic good, appears to be indispensable for some of the very important goods we know of - freely given love, freely sacrificing for the well-being of others, freely chosen acts of charity, etc. Indeed, from the point of view of the creator it might well be uninteresting to create beings who are programmed from the start to worship God, to honor him, to do good to others. From the perspective of the creator it may well be better to have beings who can freely choose to love and worship or not to love and worship, for love and worship that is freely given is of much greater value than love and worship that is compelled. But if God does choose to create a world with creatures free to do good or evil, the world may include evil as a result of some of their free choices.5 Second, there is a principle, the Principle of Organic Unities,6 held by a number of philosophers from Leibniz to the present day. According to this principle, the intrinsic value of a whole may not be equal to the sum of the intrinsic value of each of its parts. Compare, for example, Jones's feeling happy upon contemplating torturing an innocent human being with Jones's feeling unhappy upon contemplating torturing an innocent human being. The difference between these two states of affairs is that the first contains an intrinsically good state (Jones's feeling happy) as a part, whereas the second contains an intrinsically bad state (Jones feeling unhappy) as a part. But surely the first state as a whole is a much worse state of affairs than the second. So, while a given part of a whole may be intrinsically good (Jones's feeling happy), the whole of which it is a part may be worse for the presence of the good part than it would be were a certain bad part (Jones's feeling unhappy) to be in its place. So, for all we know, the best world may include some intrinsically bad states of affairs. It hardly follows from this consideration that there may be tears in heaven, but it does suggest that we should hesitate to conclude too much from the mere presence of some tears on earth. For, as we've seen, a state of affairs that constitutes an organic unity may be better for the presence of a bad part than it would be were the bad part replaced by a good part. So, again, we must note that a possible world with some bad parts may be better than a possible world with no bad parts. We've seen that the good-making qualities (happiness, love, enjoyment, beauty, good intention, an exercise of virtue, etc.) figure in states of affairs (e.g., someone's being happy, someone's loving another, etc.) that are intrinsically good; whereas the bad-making qualities (unhappiness, hate, dissatisfaction, ugliness, bad intentions, or the exercise of vice) figure in states of affairs (e.g., someone's being unhappy, someone's hating another, etc.) that are intrinsically bad. It is important, however, to distinguish the intrinsic value of someone's being unhappy from the extrinsic value of someone's being unhappy. The intrinsic value of a state of affairs is inher-

DIVINE FREEDOM CAN GOD BE FREE?

285 409

ent in that state of affairs - it necessarily belongs to that state of affairs no matter what that state of affairs is a part of or what the circumstances are in which it occurs. But the extrinsic value of a state of affairs may change from one set of circumstances to another. Sometimes, for example, a person^ being unhappy is productive of good, in which case it may be a good thing (i.ev it may be extrinsicaUy good) for that person to be unhappy. But that doesn't affect the matter of the intrinsic value of someone's being unhappy. For it is a good thing that the person was unhappy only in the sense of what that person's unhappiness leads to, or is a necessary part of, not in terms of its own intrinsic value. Unhappiness, in itself, is always bad. In addition, we should not confuse the intrinsic value of a state of affairs with the intrinsic value of a state of affairs of which it is a part. As we've noted someone's being unhappy on contemplating the undeserved suffering of others is a better state of affairs than someone's being happy on contemplating the undeserved suffering of others. But that truth is entirely compatible with someone's being happy necessarily being intrinsically better than someone's being unhappy. For the intrinsic value of the part, someone's being unhappy, must not be confused with the intrinsic value of the whole (someone's being unhappy on contemplating the undeserved suffering of others) of which it is a part. Since a possible world just is a maximal state of affairs, its value will reflect the values of the states of affairs contained in it. So, possible worlds themselves will be intrinsically good, intrinsically bad, or intrinsically neutral. In addition one possible world will be intrinsically better than, equal to, or worse than another possible world. And, as Leibniz noted, it is by knowing the intrinsic values of the possible worlds that God is guided in his choice of a world to create. We are now in a position to consider seriously our question: Is God free with respect to creating a world? Assuming that God exists, this question falls into two parts: (1) Was God free to refrain from creating any world at all?; (2) Was God free to create other worlds instead of the world he did in fact create? Let's begin with our second question. We earlier noted that possible worlds can be ranked according to their value all the way from bad worlds, to neutral worlds, and then to good worlds. Suppose then that God chooses to display his goodness and power in creating a world. We can imagine God, as it were, surveying all these worlds and deciding which one to create. He considers all the bad worlds, the neutral worlds (neither good nor bad) and all the good worlds. Let's suppose, along with Leibniz and Samuel Clarke, that the series of increasingly good worlds culminates in the best possible world, a world than which no possible world is as good or better. Similarly, we can suppose that the series of increasingly bad worlds culminates in a world than which no possible world is as bad or worse.7 Faced with choosing from among these two series of worlds the world he shall create, it is obvious that an infinitely good being would not, indeed could not, create one of the bad worlds. Which good world would he then create? Again, it seems obvious that he would create the very best world, the best of all possible worlds. As Leibniz points out: since "to do less good than one could is to be lacking in wisdom or in goodness," the most perfect understanding "cannot fail to act in the most perfect way, and consequently to

286

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 420

Faith and Philosophy

choose the best/'8 In a well-known essay9 Robert Adams has argued on the basis of the doctrine of divine grace that God would not be morally obligated to create the best world that he can. But even if his argument is successful, it still may be necessary for God to create the best world he can. It just won't be his moral duty.10 In short, his creating the best world may be a supererogatory act, the morally best act he can do, even if his failure to do it would not be a violation of his moral duty. And it appears to be inconceivable that a supremely perfect being would act to bring about less good than he can. On the assumption that God (the supremely perfect being) exists and that there is a best, creatable world, we've reached the conclusion that God is neither free not to create a world nor free to create a world less than the best creatable world. Indeed, God would of necessity create the best of the creatable worlds, leaving us with no basis for thanking him, or praising him for creating the world he does. For given that God exists and that there is a best creatable world, God's nature as an omnipotent, omniscient, perfectly good being would require him to create that best world. Doing less than the best he can do - create the best creatable world - would be inconsistent with his being the perfect being he is. But what if there is no best world? What if, as Aquinas thought to be true, for each creatable world there is a better world that God can create instead?11 In short, there is no best world. Here, I believe, in supposing that God exists and creates a world when for every creatable world there is a better creatable world, we are supposing a state of affairs that is simply impossible. I'm not suggesting here that there is an impossibility in the idea that God exists. Nor am I suggesting that there is an impossibility in the idea that for every creatable world there is a better creatable world. I am suggesting that there is an impossibility in the idea both that God exists and creates a world and that for every creatable world there is a better creatable world. For whatever world God would create he would be doing less good than he can do. And it is impossible for God to do less good than he can. The underlying principle yielding the conclusion that there is an impossibility in the idea both that God creates a world and that for every creatable world there is a better creatable world is the following: If an omniscient being creates a world when there is a better world it could create, then it would be possible for there be a being morally better than it. Since God is a being than which it is not possible for there to be a morally better being, it is dear, given both the principle just cited and the no best world hypothesis, that God could not exist and be the creator of a world. For any being that exists and creates a world when there is a better world it could have created instead is, according to the principle cited above, a being than which a morally better being is possible, and, therefore, not the best possible being. So the issue now before us is whether this principle (if an omniscient being creates a world when there is a better world it could create, then it would be possible for there to be a being morally better than it) is indeed true. My own view is that the principle in question will appear to many to be plausible, if not self-evident For if an omniscient being creates a

DIVINE FREEDOM CAN GOD BE FREE?

287 411

world when it could have created a better world, then that being has done something less good than it could do (create a better world). But any being who knowingly does something (all things considered) less good than it could do falls short of being the best possible being. So, unless we find some reason to reject the principle stated above or a reason to reject the line of argument supporting it, we are at the very least within our rights to accept it and use it as a principle in our reasoning. But the result of using this principle in our reasoning about God and the world is just this: if the actual world is not the best world that an omnipotent, omniscient being could create, God does not exist. God does not exist because were he to exist and create a world when there is a better world he could have created instead, then he would be a being than which a better being is possible. For he himself would have been a better being had he created a better world. But since it is not possible for any being (including God) to be better than God (the best possible being) in fact is, the world God has created must be the very best world he could have created. Therefore, if God does exist and creates a world W, W is the very best among the worlds that God could have created. W is the best creatable world. Hence, we see the problem of no best creatable world. For if for every creatable world there is a better creatable world and our principle is true, God does not exist. What then can be said against the principle: if an omniscient being creates a world when there is a better world it could create, then it would be possible for there to be a being morally better than it? We may begin by considering the view set forth by Norman Kretzmann in his perceptive study of Aquinas's view of creation. In the course of his discussion of Aquinas, Kretzmann concludes with Aquinas that for any world God might create there is a better world he could create. (His disagreement with Aquinas concerns only whether God is free not to create at all.) Kretzmann's second conclusion - the one presently of interest to us - is that it is a mistake to think (as I do) that if God exists and cannot avoid choosing something less good than he could choose, then God cannot be essentially perfectly good. And he proceeds to explain why he thinks it is a mistake. Like Aquinas, I think that the logical truth that God's actions conform to the principle of noncontradiction entails no limit on his power. And if it would be a violation of the principle of noncontradiction for God to create a world better than any other world he could create, then a fortiori that logical truth which does not diminish his power also leaves his goodness undiminished. God's being that than which nothing better can be conceived of cannot entail his producing a world than which none better can be conceived of. No matter which possible world he actualizes, there must be infinitely many possible worlds better than the actual world in some respect or other.12 Kretzmann relies on what he takes to be an analogy or parallel between power and goodness. His idea is this. Since we agree that failure to bring about what is logically impossible does not imply any limit on God's power, we should also agree that failure to bring about what is logically impossible does not diminish God's goodness. Given that there is no best possible world, Krfctzmann points out that it is logically impossible for God

288

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 412

Faith and Philosophy

to create a world better than any other world he could create. So, the fact that God does not create such a world diminishes neither his power nor his goodness. And that being so, Kretzmann sees no difficulty in God's being perfectly good and creating a world less good that other creatable worlds. Perhaps we can view Kretzmann as appealing to the following principle: A. If S is a logically impossible state of affairs, then the fact that a being does not bring about S does not entail that the being in question lacks power or perfect goodness. This principle strikes me as self-evidentiy true. The fact that God fails to do what logically cannot be done is a bad reason to think that God is morally imperfect or lacking in power. On Aquinas's view it is logically impossible for God to create the best possible world. And since he cannot do that, the fact that he doesn't do it, as Kretzmann notes, implies no imperfection in God. I entirely agree with Kretzmann's point on this. But the fact that there is a bad reason to conclude that God is not perfectly good does not mean that there is no good reason to conclude that God is not perfectly good. And the fact that God fails to do what logically can be done may be a good reason to conclude that God is not perfectly good.13 The principle that provides this good reason is the principle we've already introduced and will now refer to as "Principle B." B. If an omniscient being creates a world when there is a better world that it could have created, then it is possible that there exists a being morally better than it.14 If B is true, as I think it is, and if it is also true that C If a being is essentially perfectly good then it is not possible that there exist a being morally better than it, then if it is true that for any creatable world there is another creatable world better than it, it is also true that no omnipotent, omniscient being who creates a world is essentially perfectly good. Moreover, if we add to this Kretzmann's first conclusion that a perfectly good, omnipotent, omniscient being must create, it will follow that there is no omnipotent, omniscient, perfectly good being. Suppose Aquinas and Kretzmann are right in believing that for any creatable world there is another creatable world that is better than it. Our second objection emerges when we consider what the theistic God is to do in this situation. If some creatable world is better than any world God alone inhabits, then, on my principle B (slightly extended) it appears that God must create some world. On the other hand, as we've just seen, on my principle B it also follows that he cannot create a world if some other creatable world is better. "So", the objector now concludes, "on your principle B it follows that God must create a world and also must not create a world. Surely, then, since your principle leads to a contradiction, however plausible principle B sounds, we must reject it."

DIVINE FREEDOM CAN GOD BE FREE?

289 413

My response to this objection is that on the supposition that for every creatable world there is another world that is better than it, principle B does not lead to a contradiction. What principle B leads to is the conclusion that there is no essentially omnipotent, omniscient, perfectly good being. But is principle B true? Daniel and Frances Howard-Snyder have endeavored to refute principle B by inviting us to consider three hypothetical world creators: Jove, Juno and Thor.15 They suppose Jove to be an omnipotent, omniscient being who is confronted with an infinite number of increasingly better possible worlds from which to select one to create. Jove, they suggest, decides to create one of these good worlds by using a randomizing device. Being good, Jove has no interest in creating a world that isn't good.16 Each of the infinite number of good worlds is assigned a positive natural number beginning with 'V for the least good world, '2' for a slightly better world, and so on. Jove uses the randomizing device to pick one of these good worlds, and, as a result, world no. 777 is created. Now, of course, Jove could have created a better world. But the HowardSnyders think that it does not follow from this fact that Jove is morally surpassable. That is, from the fact that Jove could have created a better world than the world he did create (no. 777), they think that it does not follow that it is logically possible for there to have existed a being with a degree of moral goodness in excess of Jove's.17 In a response to their article181 suggested the following: In support of their view the Howard-Snyder's invite us to consider other possible omnipotent, omniscient19 world creators, Juno and Thor, and argue that although they produce morally better worlds than Jove, they are not morally better creators. Juno does just what Jove did but her randomizing machine happens to select a better world, no. 999. Thor doesn't use a randomizing machine but selects world no. 888 over Jove's world no. 777 because he sees that it is better and prefers creating no. 888 to creating any lesser world. Even though Juno ends up producing a better world than Jove, the Howard-Snyder's are clearly right in viewing Jove and Juno as morally equivalent. For had her randomizing machine hit on world 777, rather than world 999, Juno would have created world 777. So, it was blind luck, not a higher standard of selection, that resulted in Juno's selection of world 999. But what of Thor? From their discussion it would seem that Thor is morally superior to Jove and Juno, for it looks as though Thor's degree of moral goodness is such that he is not prepared to settle for world no. 777 unless he is unable to create a better world. But the fact that Jove intentionally included worlds numbered 1 - 777 as possibilities for selection by his randomizing machine shows that Jove is morally prepared to settle for any of the worlds from 1 - 777 even though he is able to create a better world.20 So, it does appear that, other things being equal, Thor is a morally better being than Jove.21 In a subsequent article22 against principle B, the Howard-Snyders question my account of Thor, suggesting that it is incoherent. They wonder

290

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 414

Faith and Philosophy

what principle or reason Thor acts on. They say: For example, suppose Thor's reason is this: worlds numbered 888 and higher are better than worlds numbered 887 and lower. (This seems to be the reason that Rowe has Thor act on. See the quotation above.) This reason relies on the general principle that if world w is better than world w-1, then w-1 is unacceptable for creation. Any being who accepted an instance of this principle when it involved the world no. 888 but did not accept other instances of it would be irrational, and hence not essentially omniscient. Any being who accepted the principle in its full generality would be led never to create, given (as we are supposing) that for each world there is a better.23 This leads them to suggest that my account of Thor is incoherent. But I believe it is clear from my article that I do not have Thor act on the principle "don't create if there is a better creatable world/' for such a principle, given that for any world there is a better, can only result in Thor's not creating any world at all. Since Thor is omniscient and does in fact create world 888, it logically follows that he cannot act on the principle that they suggest I have him act on. Of course, while Thor, given his infinite intelligence, cannot act on such a principle, it doesn't follow that with finite intelligence I cannot make the mistake of attributing to Thor such a principle of action when he creates world 888. So, what principle do I have Thor act on? I believe that the principle on which Thor acts is very much like the principle on which Jove acts. Let's look again at Jove. Some worlds he sees as not good enough to be acceptable as candidates for creation. The worlds that are acceptable to him in terms of his own degree of goodness are then ordered in terms of increasing goodness, and one of them, world 777, is randomly selected for creation. Thor, as I have described him, does pretty much the same thing. The difference is that worlds 1 - 800 are insufficiently good to be acceptable to him as candidates for creation, given that there are better worlds he can create. The worlds that are acceptable to him in terms of his own degree of goodness are then ordered in terms of increasing goodness and one of them, world 888, is randomly selected for creation. I conclude that the description the Howard-Snyder's give of Jove is logically consistent with there being a being who is better than Jove. And the story we have told about Thor is consistent and, if true, gives us reason to believe that Thor is a better being than Jove. Can we state the principles on which both Jove and Thor act, and explain how it is that although they act on the same principles, they produce worlds that differ in their degree of goodness? I suggest that Jove and Thor may act on the following principles: PI: Do not create any world that is not a good world. P2. Do not create any good world whose goodness is less than what one judges as acceptable, given that one can create a better world. Clearly, both Thor and Jove act in accordance with PI and P2. Neither is

DIVINE FREEDOM CAN GOD BE FREE?

291 415

prepared to create a less than good world. And neither is prepared to create a good world whose degree of goodness is less than what he judges as acceptable in a world, given that he can create a better world. The difference between them is this. Jove's standard of goodness in world creating is such that he is prepared to settle for any good world even if there is a better that he can create. Thor, however, has a higher standard. He is not prepared to create any of the good worlds from Wl to W800 provided there is a better world that he can create. Of course, Thor's allegiance to P2 does not preclude him absolutely from creating, say, W777. It prevents him only on the condition that there is a better world he can create. Gala apples taste much better than Jonathan apples. I know that, and my standard of apple selection is never to come home with Jonathan apples when Gala are available. But that doesn't mean I won't or can't select Jonathan apples when Gala apples are not available. A good apple, even if it's a Jonathan, is better than no apple at all. In their article the Howard-Snyders suppose that it cannot be that Thor and Jove act on the same principle. They suppose that if my story about Thor being better than Jove is correct then Thor must be acting on a higher principle than the principle on which Jove acts. And, since there will be worlds better than the world Thor creates, they then conclude that "there is another principle which treats as unacceptable some of the worlds which were treated as acceptable by Thor's principle, and that other principle is such that there is a third principle which treats as unacceptable some of the worlds which were treated as acceptable by the second, and so on, ad infinitum." Lacking a proof of the impossibility of such an infinite array of world-creating principles, they say, It seems odd to say the least that there should be infinitely many such general principles. At least we see no reason to accept that there are.24 As I've tried to make clear above, the story I tell is quite consistent even if there is no such infinite progression of world-creating principles. Principles PI and P2 will suffice so long as for any being in the position of Jove or Thor there is another being whose degree of goodness is such that its application of PI and P2 results in the selection of a better world to create. And if we allow, as the Howard-Snyders do, an infinite number of possible worlds beginning with Jove's good world Wl, why not allow the possibility of an infinite series of good world creators each being better than the preceding one. Indeed, why not allow, if needed, an infinite number of different but related world-creating principles. But, as I've suggested, I don't see the necessity of supposing that my story is coherent only if there is an infinite number of distinct world-creating principles. But suppose an infinite number of such principles is required. Perhaps it is odd at that there should be infinitely many world-creating principles. But even if it is odd, we should note that oddness and impossibility are far different matters. Many extremely odd things are logically possible. And if it is logically possible that there is an absolutely infinite number of increasingly better worlds, why should it be impossible that there be an infinite number of principles of world creation? In any case, however, I see no reason to think that there

292

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 416

Faith and Philosophy

need be an infinite number of such principles in order for the story of Thor and Jove to be coherent. Indeed, I have suggested that Jove and Thor can act on the very same principles. Moreover, since it is possible that there be world creators whose degrees of goodness increasingly exceed Thor's, it is possible that the very same principles would result in increasingly better beings creating increasingly better worlds than Thor's world. The fundamental question at issue in the discussion concerning Jove and Thor is this: Is it logically possible both that for any creatable world there is a better creatable world and that there exists an omnipotent, omniscient, perfectly good being who creates one of these creatable worlds? My position is that it is not possible that both should be true. Why do I hold this position? I hold it because, as I've stated earlier, I think the following [Principle B] is necessarily true. B. If an omniscient being creates a world when it could have created a better world, then it is possible that there be a being morally better than it.25 By telling their story about Jove, the Howard-Snyder's hoped to cast doubt on Principle B. I believe that my alternative story about Thor undermines their attempt. Where does this leave us? I assert that B is necessarily true. Many theists assert the following [Principle A] to be true. A. It is logically possible both that for any creatable world there is a better creatable world and that there exists an omnipotent, omniscient, perfectly good being who creates one of these worlds. Both of us cannot be right. But how can we hope to settle the question of who has the more plausible position? Are we simply at a stalemate, a situation where neither can show the other's position to be implausible without employing as a premise one of the principles that is at issue in the debate? I believe that the Howard-Snyder's have endeavored to advance the debate in a way that does not beg the question. They suppose both that Jove is an omnipotent, omniscient creator of a good world (# 777) and that for every creatable world there is a better, but leave as an open question whether Jove's goodness can be unsurpassable. The question then is whether we have some good reason to think that an omnipotent, omniscient creator of a better world than # 777 may be better than Jove. As we've seen, the answer depends on the reason such a being has for creating a better world than # 777. If such a being (their Juno), given her degree of goodness, judges as acceptable for creation the same worlds as Jove, then the fact that her randomizer selects world # 999 for creation gives us no reason at all to think that Juno is a better being than Jove, even though she ends up creating a better world than does Jove. But if, like Thor, the being's degree of goodness is such that he judges that worlds of lesser value than # 800 are unacceptable candidates for creation, then the fact that its randomizer selects world # 800 or higher gives us reason to think that Thor is a better being than Jove. Of course, if we had simply concluded that Jove's goodness is surpassable because Jove could have created a world better than #

DIVINE FREEDOM CAN GOD BE FREE?

293 417

777, this would have been to beg the question at issue. For we would have been appealing to Principle B to rule out Jove being an unsurpassably good being. But no such appeal was made in reasoning to the conclusion that Thor is a better being than Jove. Principle B, if true, does not refute theism. But if both principle B and theism are true, then the world we live in is an unsurpassably good world - no possible world that an omnipotent being could have created would be better than the actual world. I suspect that part of the motivation for the theist to accept the view that there is no best creatable world is that the alternative seems (1) to limit severely God's freedom in creating, and (2) to leave the theist with the burden of defending the Leibnizian thesis that this world, with all its evil, is a world than which a better creatable world is not even a logical possibility. Thomas Morris, like the Howard-Snyder's and Kretzmann, thinks that among the worlds creatable by God there is no best world. He notes two difficulties in the liebnitzian idea that there is a best possible world. First, he points out that some philosophers are doubtful that there is a single scale on which all creaturely values can be weighted so as to determine what world possesses the maximum amount of value. "Some world A might be better than rival world B in some respects, but with B surpassing A in others, and the relevant values not such that they could be summed over and compared overall."26 In short, if some valuable states of affairs are incommensurable with other valuable states of affairs, it may be impossible to rank the states of affairs in terms of one being better than, worse than, or equal to the other. And if that should be so, we could have two worlds such that neither is better than the other, worse than the other, or equal in value with the other. Second, Morris notes that a number of philosophers have thought that for any world containing "a certain number of goods, n, there is always conceivable a greater world with n + 1 goods, or good creatures. So, on the simplest grounds of additive value alone, it seems impossible there could be a single best possible world. And without this, of course, the Leibnizian demand collapses/'27 But Morris's main concern lies elsewhere. Like the Howard-Snyders and Kretzmann, Morris wants to show that there is no incoherence in the idea of a perfectly good creator creating a world when there is no best world for that being to create. For just as it seems initially very natural to suppose that a superlatively good, wise, and powerful being will produce only an unsurpassable perfect creation, so likewise it can seem every bit as natural to suppose that an incoherence or impossibility discovered in the latter notion indicates an incoherence or impossibility buried within the former.28 Since it is just that incoherence we have been arguing for, it is important to note Morris's efforts to show that the incoherence in question is imagined, and not real. Noting that Quinn holds that in the no best world scenario a creator of a world is such that "it is possible that there is an agent morally better than he is, namely an omnipotent moral agent who actualizes one of those morally better worlds," and Rowe holds "if a being were to create a

294

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 418

Faith and Philosophy

world when there is a morally better world it could create, then it would be possible for there to be a being morally better than it/'Morris states that these views are "absolutely unacceptable to traditional theists, for whom both perfection and creation are important ideas/'29 In developing his objection to the views expressed by Quinn and Rowe, Morris introduces a useful thesis - the Expression Thesis: The goodness of an agent's actions is expressive of the agent's goodness. I'm inclined to take something like this thesis as underlying the daim expressed in principle B: If an omniscient being creates a world when it could have created a better world, then it is possible that there be a being morally better than it. Of course, the expression thesis depends on what Morris may well have supposed: that the agent's motive for performing the good action is to bring about a good state of affairs. Without supposing that motive we have no reason to think that the goodness of an agent's action - measured in terms of the quality of its result - is expressive of the agent's goodness. But what are we to say of a being who performs an action that he knows will bring about less good, all things considered, than he could have brought about by performing a slightly different action? In this case, applying the expression thesis, we should conclude that the agent's degree of goodness is something less that it could be. For the agent has acted to bring about less good than he knew would have been brought about by his performing a slightly different action. But clearly, if an agent knowingly acts to bring about less good overall than he could have brought about by performing a slightly different action that was in his power to perform, that agent's degree of moral goodness is somewhat less than it could be. And it is precisely this point that underlies the judgment that in the no best world scenario it is impossible for the creator to be perfectly good. For, as we saw in the discussion of the Howard-Snyder's story about Jove, Juno, and Thor, when a being creates a world that is less good than another world it could have created, the world it creates will satisfy its standard of world-creating, even given that it could create a better world. But then it is possible that there should be a being whose degree of goodness is such that it will not create that less good world given that it is able to create a better world. So, again I conclude that if a being creates a world when it could have created a better world, then it is possible that there should be a being morally better than it. And from this it follows that if for every creatable world there is a better creatable world, there is no absolutely perfect being who creates a world. And since it is better to create a good world rather than not create any world at all, on the no best creatable world scenario there is no maximally perfect being. Morris's basic mistake, I believe, is his view, shared by Kretzmann, that to hold, as I do, that if there is no best world for a being to create then no being can create a world and be a being than which a better creator is impossible, just is to hold God accountable for not doing what is logically impossible to be done - creating the best world. Thus Morris writes: If you and I do less well than we're capable of doing, then those around us may conclude, and may sometimes justifiably conclude, that we are not at the level of goodness that could be exemplified. But

DIVINE FREEDOM CAN GOD BE FREE?

295 419

failing to do the best you can is a flaw or manifests an incompleteness in moral character in this way only if doing the best you can is at least a logical possibility. If doing the best he can in creating a world is for God an impossibility,... then not doing his best in creating cannot be seen as a flaw or as manifesting an incompleteness in the character of God. The notion of a perfect expression of an unsurpassable character would then itself be an incoherence.30 Of course, if it is logically impossible for there to be a best world, then God's not creating the best possible world does not count against his perfect goodness. Nowhere do I suggest that it does. What counts against God's perfect goodness (specifically, his moral perfection) is his creating a world when he could have created a world better than it. The charge is not that a being who fails to do what is impossible to be done (create the best world when there is no best world to be created) is lacking in perfect goodness. The charge is that a being who creates a world when it could have created a better world is less than supremely perfect. And the plain fact is that if there is no best creatable world then God, if he creates a world, will create a world than which he could have created a better world.31 Morris simply fails to address the issue at stake here. It is important to distinguish three different principles: a. Failing to do the best one can is a defect only if doing the best one can is possible for one to do. b. Failing to do better than one did is a defect only if doing better than one did is possible for one to do. c. Failing to do better than one did is a defect only if doing the best one can is possible for one to do. Both (a) and (b) are true. But (c) is not true. And it is (c) that Morris needs to make his argument work. Suppose, for the moment, that you are an omnipotent, omniscient being and are contemplating the infinite series of numbers: 1,2,3,4, etc., etc. You are also contemplating the infinite series of creatable worlds containing creatures that are overall good worlds, as opposed both to bad worlds and neutral worlds - worlds that are neither good nor bad. You let each of the numbers represent the overall degree of good that a possible world possesses, where T represents the least good world - a world with no pain perhaps, and just one momentary experience of pleasure on the part of some lower animal. '2' represents the possible world that is one degree better than the world 1, '3' represents the possible world that is one degree better than the world 2, etc., etc. Being omniscient you see that there is no best possible world for you to create. Just as the series of natural numbers increases infinitely so does the series of increasingly better worlds from which you will select one to create. Seeing that there is no best possible world to create, you realize that no matter how good a world you create there will be better worlds you could have chosen to create instead of it.

296

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 420

Faith and Philosophy

Glancing at world 1, and comparing it with world 1000, you see that world 1000 is significantly better than world 1, just as you see that world 1000,000 is significantly better than world 1000. Nevertheless, in spite of noticing the enormous disparity between the least good world and the goodness of some worlds numerically much greater, you decide that you will create the least good world and proceed to actualize world 1. Isn't it obvious that in deliberately choosing to create the least good of the infinite series of increasingly better possible, creatable worlds you display a degree of goodness in world-creating that is inconsistent with perfect goodness? "Wait!" you will say. "You judge me unfairly. I see that if I could have created a maximally good world I might be subject to some criticism here for creating a world so limited in value as world 1. But there is no maximally good world. So clearly I'm perfectly justified in creating the poorest in the infinite series of increasingly better worlds. You should not have any doubts at all about my being perfectly good\" Surely this defense of one's "perfect goodness" is woefully inadequate. A perfectly good being cannot, consistent with its perfect goodness, consciously elect to create the least good world when there is an infinite number of increasingly better worlds as available for creation as the least good world. But it is just this conclusion that Morris's position would require us to accept. Since the conclusion is clearly false, if not absurd, we should reject it. Instead, we should say that the degree of goodness an omniscient being possesses is reflected in the degree of goodness in the world it creates. And what this reasoning leads us to is the conclusion Leibniz reached: An unsurpassably good, omnipotent, omniscient creator will create an unsurpassably good world. Indeed, unsurpassable goodness in an omnipotent, omniscient world-creator is consistent only with the creation of an unsurpassably good world. For there is an impossibility in the idea both that there exists an infinite series of increasingly better creatable worlds and that there also exists an unsurpassably good, omnipotent, omniscient being who creates one of these worlds. The conclusion we've just reached points to an incompatibility between the necessary existence of the theistic God and the possibility Morris embraces: that the series of increasingly better creatable worlds goes on to infinity. But how could a mere possibility be inconsistent with the existence of the theistic God? After all, isn't it one thing to conclude that God's infinite perfection precludes his actualizing a bad possible world, and quite another thing to conclude that God's infinite perfection precludes there even being such a thing as a bad possible world? And if God's existence doesn't rule out bad possible worlds, why should it rule out an infinite series of increasingly better possible worlds? The answer to these questions consists in seeing that if God necessarily exists and is necessarily such that whatever world is actual can be so only by virtue of his creating it, then since it is impossible for God (an absolutely perfect being) to create a bad world, there cannot be any bad worlds. In short, given God's necessary perfections and necessary existence, the only possible, non-actual worlds are worlds God can create. And once we see that given God's

DIVINE FREEDOM CAN GOD BE FREE?

297 421

necessary existence and necessary perfections no world creatable by God can be a bad world, we are well on our way to seeing that it is likewise impossible that there should be an unending series of increasingly better creatable worlds. Morris nicely captures the essence of the view I've just described by noting that such a God "is a delimiter of possibilities." If there is a being who exists necessarily, and is necessarily omnipotent, omniscient, and good, then many states of affairs which otherwise would represent genuine possibilities, and by all non-theistic tests of logic and semantics do represent possibilities, are strictly impossible in the strongest sense. In particular, worlds containing certain sorts or amounts of disvalue or evil are metaphysically ruled out by the nature of God, divinely precluded from tike realm of real possibility.32 Return now to our earlier contention that possible worlds include very good worlds, neutral worlds, and very bad worlds. In addition we suggested that just as for every good world there is a better possible world, so too for every bad world there is a possible world whose degree of badness is greater. Morris will allow that such worlds are "conceivable/7 But since he holds that God is a delimiter of possibilities and that it is impossible for God to create a world that is a bad world, the bad worlds we conceive of are not, at least for the theist, genuinely possible. For the only way such worlds could be genuinely possible is for it to be possible for God to create them. But God's perfect nature necessarily precludes him from creating such worlds. Therefore, such worlds aren't really possible all things considered. Morris's general point here strikes me as sound. If p is necessarily true and q is inconsistent with p, then, even though we can conceive of q and q seems to us to be paradigm case of a genuine possibility, q isn't really possible at all. So, if there is a necessarily perfect being who necessarily exists, then even though we can conceive of a bad world, that bad world is really not a possible world provided that for a world to be actual it must be actualized by the necessarily perfect being. Consider, for example, a world in which nearly all the sentient beings have lives so full of suffering that it would be better had they never existed. Such a world is a bad world. Is this world, so understood, a possible world? It certainly seems to be. But given that a possible world can be actual only if it is created by a necessarily perfect being, and such a being necessarily exists, then that world really isn't a possible world, it only seems to be possible. Of course, what is sauce for the goose (Morris) is sauce for the gander (Rowe). If this bad world, which certainly seems to be possible, really is a possible world, then it is simply impossible that there is a necessarily perfect being who is necessarily the creator of any world that is actual. Which then are we more sure of: that some bad world is genuinely possible or that there necessarily exists a being who is necessarily omnipotent, omniscient, and perfectly good? The former is a "delimiter of necessities" just as the latter is a "delimiter of possibilities." Just as what is necessary precludes certain "possibilities," so does what is possible preclude certain "necessities." The theist begins with the neces-

298

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 422

Faith and Philosophy

sary existence of a being who is essentially perfect and concludes that a bad world isn't even a possibility. The non-theist begins with the possibility of there being a bad world and concludes that there is no essentially perfect being who necessarily exists. Suppose we accept Morris's view about God as a delimiter of possibilities. Suppose, that is, that we agree with him that if there is a being who exists necessarily, and is necessarily omnipotent, omniscient, and good, then many states of affairs which otherwise would have been possible are strictly impossible. If so, then if there is such a being we should agree that there are no possible worlds that are overall bad worlds. For such a world is possible only if it is possible for God to actualize that world. But God's necessary perfections preclude him from actualizing it. Therefore, such a world is not really a possible world. Having adopted Morris's view that God is a delimiter of possibilities, suppose we now return to the stalemate between the view I've argued for: It is impossible for God to exist and create an inferior world when he could have created a better world', and the view my opponent maintains: God is free to create some good world even though there is an unlimited number of better worlds any one of which he could have created instead. My opponent may well agree with me that given God's absolute perfection God could not create a world less than the best world. In short, if there is a best world all things considered, we may both agree with Leibniz and Clarke that God will necessarily create that world. (Of course, there remains the problem of explaining the precise sense in which God could be free in creating the best possible world.) It is only when we come to the no best world scenario that our views clash so profoundly. But Morris, perhaps unwittingly, has shown us a way to resolve the problem. God is the ultimate delimiter of possibilities. Thus, if God exists the series of increasingly good possible worlds has a limit - the best possible world. A creator that is necessarily good could not possibly create a less than good world. So, given that this being is a delimiter of possibilities, there are no po.ssible worlds that are not good worlds. Furthermore, a necessarily perfect being could not possibly create a world that is less good than some other world it could create. So, given that this being creates a world and is a delimiter of possibilities, the world he creates cannot be one than which there is a better creatable world. Thus, following the path that Morris has pointed out, we condude that God's necessary existence and necessary perfections would rule out two seeming possibilities: (1) there being possible worlds that are bad; (2) there being no best possible world. If God exists, his necessary existence and necessary perfections rule out the apparent existence of possible worlds that are bad as well as the apparent possibility that for any world God can create there is a better world he could create. What then should we conclude about the actual world? We should conclude that if the theistic God does exist, the actual world is the best possible world.33 And therein lies the

DIVINE FREEDOM CAN GOD BE FREE?

299 423

seeds of another argument against the existence of the God of traditional theism. For however much we may succeed in trying to fit the terrible evils in our world into some rational plan, few are prepared to think with Leibniz that this world is as good as any world could possibly be.

NOTES 1. In saying that a state of affairs obtains (or is actual) in a given possible world we mean that that state of affairs would be actual were mat world the actual world. 2. Alvin Plantinga, The Nature of Necessity (Oxford: Oxford University Press, 1974), p. 45. 3. See Roderick M. Chisholm, "The Defeat of Good and Evil/' in The Problem of Evil, ed. by Marilyn McCord Adams and Robert Merrihew Adams, (Oxford: Oxford University Press, 1990). 4. Actually, since possible worlds necessarily exist, God doesn't create them. But from the fact that a possible world exists, it doesn't follow that it is actual Only one world can be the actual world. And what God does is create particular things - stones, human beings, etc.- and enable them to be arranged in such a way that a particular possible world is actualized. So, it is not in the literal sense of 'creates' that God creates a world. With this understood, we will continued to refer to some possible world as being 'created' by God. 5. But surely there would be possible worlds in which creatures are free to do good or evil and, as it happens, always use their freedom to do good. Wouldn't God create one of those worlds? For an impressive argument as to why it might not be in God's power to create such a world see Alvin Plantinga, The Nature of Necessity (Oxford University Press, 1974), Ch. DC. 6. See G. E. Moore, Principia Ethica (Cambridge: Cambridge University Press, 1903), pp. 187ff. 7. In the service of simplicity we will set aside the neutral worlds. 8. See Leibniz's Theodicy, trans. E. M. Huggard, ed. Austin Farrer (LaSalle, HI: Open Court 1985), section 201. (Emphasis mine) 9. "Must God Create the Best?" Philosophical Review 81 (July 1972): 317-32. 10. I advance this point in "The Problem of Divine Perfection and Freedom," Reasoned Faith, ed. Eleonore Stump, Cornel University Press, 1993, 223-233. 11. Aquinas didn't understand a world to be a maximal state of affairs. But were he to have thought of a world in this way he would have asserted that there is no best world. 12. "A Particular Problem of Creation," 238. 13. In this discussion of Kretzmann I suppose, for effect, that it is possible for God to be less than perfectly good. What is true is that any being that fails to do what is the best it can do is not perfectly good and, therefore, not God. 14. As we noted earlier, a being may be perfectly morally correct in the sense of never failing in its obligations and still be such that it could be morally better by virtue of the performance of some supererogatory act. 15. Daniel and Frances Howard-Snyder, "How and Unsurpassable Being Can Create a Surpassable World," Faith and Philosophy, April, 1994. 16. In order not to beg the question at issue, the Howard-Snyder's do not assume that Jove is morally unsurpassable.

300

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 424

Faith and Philosophy

17. In the context of this discussion, a being is morally unsurpassable only if it is logically impossible for there to be a morally better being. 18. "The Problem of No Best World," Faith and Philosophy, April, 1994. 19. For some reason the Howard-Snyder's neglect to attribute omniscience to Juno and Thor. I assume this to be a slip. Clearly, if we want to compare their goodness to Jove's, we should attribute to them the infinite power and knowledge that was attributed to Jove. I'll return to this point in discussing Trier's degree of goodness. 20. And the same is true of Juno, even though she accidentally ends up with world no. 999. 21. It is important to note that to say one being is morally better than another is not to imply that the second being has done anything morally wrong or violated any moral obligation. 22. "The Real Problem of No Best World/' Faith and Philosophy, July 1996. 23. Mi., p. 423. 24. Ibid., p. 424. 25. Indeed, it is possible for that very being to have been better than it in fact is. 26. "Perfection and Creation," Reasoned Faith, ed. Eleonore Stump, Cornel University Press, 1993,223-233. 27. M*.,p.237. 28. Ibid. 29. Ibid., 239. 30. Ibid.,2M. 31. This is technically incorrect. For if there is no best world, but several worlds equally good and none better, then a being could freely chose one of these worlds to create. The reader is to understand that in supposing that there is no best world we are thinking only of the circumstance in which for every world there is a better world. 32. Anselmian Explorations, (University of Notre Dame Press, 1987), p. 48. 33. If we countenance libertarian freedom among the creatures in the best possible world, it could be that the best possible world is not creatable by God. For if the creatures are free to do right or not do right, then it won't be entirely up to God whether the best world he can create is the best possible world. I ignore these complications here. But clearly they are important to a thorough discussion of the problem of divine freedom.

PARTY THE COSMOLOGICAL ARGUMENT

This page intentionally left blank

CHAPTER TWENTY-ONE

Two Criticisms of the Cosmological Argument In this paper I wish to consider two major criticisms that have been advanced against the Cosmological Argument for the existence of God, criticisms which many philosophers regard as constituting a decisive refutation of that argument. Before stating and examining these objections it will be helpful to have before us a version of the Cosmological Argument. The Cosmological Argument has two distinct parts. The first part is an argument to establish the existence of a necessary being. The second part is an argument to establish that this necessary being is God. The two objections I shall consider are directed against the first part of the Cosmological Argument. Using the expression 'dependent being1 to mean 'a being that has the reason for its existence in the causal efficacy or nature of some other being*, and the expression 'independent being1 to mean 'a being that has the reason for its existence within its own nature', we may state the argument for the existence of a necessary being as follows: 1. Every being is either a dependent being or an independent being; therefore, 2. Either there exists an independent being or every being is dependent; 3. It is false that every being is dependent; therefore, 4. There exists an independent being; therefore, 5. There exists a necessary being. This argument consists of two premises—propositions (1) and (3) -and three inferences. The first inference is from (1) to (2), the second from (2) and (3) to (4), and the third inference is from (4) to (5). Of the premises neither is obviously true, and of the inferences only the first and second are above suspicion. Before discussing the main subject of this paper—the reasoning in support of proposition (3) and the two major objections that have been advanced against that reasoning—I want to say something about the other questionable parts of the argument; namely, proposition (1) and the inference from (4) to (5).

304

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 442

WILLIAM L. ROWE

Proposition (1) expresses what we may call the strong form of the Principle of Sufficient Reason. It insists not only that those beings which begin to exist must have a cause or explanation (the weak form of the Principle of Sufficient Reason) but that absolutely every being must have an explanation of its existing rather than not existing—the explanation lying either within the causal efficacy of some other being or within the thing's own nature. In an earlier paper I examined this Principle in some detail.1 The objections I wish to consider in this paper are, I believe, independent of the Principle of Sufficient Reason. That is, these objections are meant to refute the argument even if the first premise is true. This being so it will facilitate our examination of these two objections if we take proposition (1) as an unquestioned premise throughout our discussion. Accordingly, in this paper proposition (1) will function as an axiom in our reasoning. This, of course, should not be taken as implying that I think the first premise of the argument is true. The inference from proposition (4) to proposition (5) is not considered in this paper. Indeed, for purposes of this paper we could have ended the statement of the argument with proposition (4). I have included the inference from (4) to (5) simply because it is an important element in the first part of the Cosmological Argument. Proposition (4) asserts the existence of a being that has the reason or explanation of its existence within its own nature. Proposition (5) asserts the existence of a necessary being. By 'a necessary being1 is meant a being whose nonexistence is a logical impossibility.2 Many philosophers have argued that it is logically impossible iSee "The Cosmological Argument and the Principle of Sufficient Reason," Man and World, 1, No. 2. 2 Not all versions of the Cosmological Argument employ the notion of a logically necessary being. It seems likely, for example, that in Aquinas' 3rd Way the expression 'necessary being* is not used to mean a logically necessary being. (See P. Brown, "St. Thomas' Doctrine of Necessary Being," Philosophical Review. 73 (1964), 76-90.) But in the version we are considering it is clear that by 'necessary being* is meant a being whose existence is logically necessary. Thus Samuel Clarke, from whose work our version has been adapted, remarks: "... the only true idea of a self-existent or necessarily existing being, is the idea of a being the supposition of whose not-existing is an express contradiction." (See Samuel Clarke, A Demonstration of the Being and Attributes of God, 9th edition, p. 17.) David Hume,also understands the notion of a necessary being this way. Thus in his statement of the argument, which he adapted from Clarke, he

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

305 443

for there to be a necessary being in this sense of 'necessary being9. Hence, even if the two objections I shall examine in this paper can be met, the defender of the Cosmological Argument must still face objections not only to the inference from (4) to (5) but to (5) itself. But again, this is a matter that I shall not pursue in this paper. Unlike proposition (1), however, which I treat as an unquestioned assumption, neither proposition (5) nor the inference from (4) to (5) will be appealed to in this paper. In what follows we may simply ignore that part of the argument. Indeed, our attention will be focused entirely on proposition (3), the reasoning that supports it, and the two major criticisms that have been advanced against that reasoning. Proposition (3) asserts that it is false that every being is dependent. For what reasons? Well, if every being that exists (or ever existed) is dependent then the whole of existing things, it would seem, consists of a collection of dependent beings, that is, a collection of beings each member of which exists by reason of the causal efficacy of some other being. This collection would have to contain an infinite number of members. For suppose it contained a finite number, let us say three, a, b, and c. Now if in Scotus's phrase "a circle of causes is inadmissible" then if c is caused by b and b by 0, a would exist without a cause, there being no other member of the collection that could be its cause. But in that case a would not be what by supposition it is, namely a dependent being. Hence, if we grant that a circle of causes is inadmissible it is impossible that the whole of existing things should consist of a collection of dependent beings finite in number. Suppose, then, that the dependent beings making up the collection are infinite in number. Why is it impossible that the whole of existing things should consist of such a collection? The proponent of the Cosmological Argument answers as follows.3 The infinite collection itself, he argues, requires an explanation of its existence. has Demea conclude, "We must, therefore, have recourse to a necessarily existent being, who carries the reason of his existence in himself, and who cannot be supposed not to exist, without an express contradiction." (Dialogues Concerning Natural Religion, Part IX.) 3 See, for example, Samuel Clarke's discussion of Propositions II and III in his Demonstration. This discussion is summarized by Hume in Part IX of his Dialogues.

306

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 444

WILLIAM L. ROWE

For since it is true of each member of the collection that it might not have existed, it is true of the whole infinite collection that it might not have existed. But if the entire infinite collection might not have existed there must be some explanation of why it exists rather than not. The explanation cannot lie in the causal efficacy of some being outside of the collection since by supposition the collection includes every being that is or ever was. Nor can the explanation of why there is an infinite collection be found within the collection itself, for since no member of the collection is independent, has the reason of its existence within itself, the collection as a whole cannot have the reason of its existence within itself. Thus the conception of an infinite collection of dependent beings is the conception of something whose existence has no explanation whatever. But since premise (1) tells us that whatever exists has an explanation for its existence, either within itself or in the causal efficacy of some other being, it cannot be that the whole of existing things consists of an infinite collection of dependent beings. The reasoning developed here is exhibited as follows: 1. If every being is dependent then the whole of existing things consists of an infinite collection of dependent beings; 2. If the whole of existing things consists of an infinite collection of dependent beings then the infinite collection itself must have an explanation of its existence; 3. If the existence of the infinite collection of dependent beings has an explanation then the explanation must lie either in the causal efficacy of some being outside the collection or it must lie within the infinite collection itself; 4. The explanation of the existence of the infinite collection of dependent beings cannot lie in the causal efficacy of some being outside the collection; 5. The explanation of the existence of the infinite collection of dependent beings cannot lie within the collection itself; therefore, 6. There is no explanation of the infinite collection of dependent beings; (from 3,4, and 5), therefore, 7. It is false that the whole of existing things consists of an infinite collection of dependent beings; (from 2 and 6), therefore, 8. It is false that every being is dependent, (from 1 and 7).

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

307 445

Perhaps every premise in this argument is open to criticism. I propose here, however, to consider what I regard as the two major criticisms that have been advanced against this reasoning in support of proposition (3) of the main argument. The first of these criticisms may be construed as directed against premise (2) of the above argument. According to this criticism it makes no sense to apply the notion of cause or explanation to the totality of things, and the arguments used to show that the whole of existing things must have a cause or explanation are fallacious. Thus in his B.B.C. debate with Father Copleston, Bertrand Russell took the view that the concept of cause is inapplicable to the universe conceived of as the total collection of things. When pressed by Copleston as to how he could rule out "the legitimacy of asking the question how the total, or anything at all comes to be there," Russell responded: "I can illustrate what seems to me your fallacy. Every man who exists has a mother, and it seems to me your argument is that therefore the human race must have a mother, but obviously the human race hasn't a mother—that's a different logical sphere."4 The second major criticism is directed at premise (5). According to this criticism it is intelligible to ask for an explanation of the existence of the infinite collection of dependent beings. But the answer to this question, so the criticism goes, is provided once we learn that each member of the infinite collection has an explanation of its existence. Thus Hume remarks: "Did I show you the particular causes of each individual in a collection of twenty particles of matter, I should think it very unreasonable, should you afterwards ask me, what was the cause of the whole twenty. This is sufficiently explained in explaining the cause of the parts." 5 These two criticisms express the major reasons philosophers have given for rejecting what undoubtedly is the most important part of the Cosmological Argument; namely, that portion of the argument which seeks to establish that not every being can be a dependent being. In this paper my aim is to defend the Cosmological Argu4 "The Existence of God, A Debate between Bertrand Russell and Father F. C, Gopleston," in John Hick (ed.)f The Existence of God (New York: The Macmillan Company, 1964), p. 175. The debate was originally broadcast by the British Broadcasting Corporation in 1948. References are to the debate as reprinted in The Existence of God. * Dialogues, Part IX.

308

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 446

WILLIAM L. ROWE

ment against both of these criticisms. I shall endeavor to show that each of these criticisms rests on a philosophical mistake. The first criticism draws attention to what appears to be a fatal flaw in the Cosmological Argument. It seems that the proponent of the argument (i) ascribes to the infinite collection itself a property (having a cause or explanation) which is applicable only to the members of that collection, and (ii) does so by means of a fallacious inference from a proposition about the members of the collection to a proposition about the collection itself. There are, then, two alleged mistakes committed here. The first error is, perhaps, a category mistake—the ascription to the collection of a property applicable only to the members of the collection. As Russell would say, the collection, in comparison with its members, belongs to a "different logical sphere." The second error is apparently what leads the proponent of the Cosmological Argument to make the first error. He ascribes the property of having an explanation to the infinite collection because he infers that the infinite collection must have a cause or explanation from the premise that each of its members has a cause. But to infer this, Russell suggests, is as fallacious as to infer that the human race must have a mother because each member of the human race has a mother. That the proponent of the Cosmological Argument ascribes the property of having a cause or explanation to the infinite collection of dependent beings is certainly true. That to do so is a category mistake is, I think, questionable. But before pursuing this point I want to deal with the second charge. The main question we must consider in connection with the second charge is whether the Cosmological Argument involves the inference: Every member of the infinite collection has an explanation of its existence; therefore, the infinite collection itself has an explanation of its existence. As we have seen, Russell thinks Copleston has employed this inference in coming to the conclusion that there must be an explanation for the totality of things, and not simply for each of the things making up that totality. Perhaps some proponents of the Cosmological Argument have used the argument that Russell regards as fallacious. But not all of them have.6 Moreover, there is no need to employ such an inference 6 Samuel Clarke did not. Nor do we find Hume appealing to this inference in the course of presenting the Cosmological Argument in Part IX of the Dialogues.

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

309 447

since in its first premise the Cosmological Argument has available a principle from which it follows that the infinite collection of dependent beings must have an explanation of its existence. Thus one famous exponent of the argument—Samuel Clarke—reasons that the infinite collection of beings must have an explanation of its existence by appealing to the strong form of the Principle of Sufficient Reason. The principle assures us that whatever exists has an explanation of its existence. But if there exists an infinite succession or collection of dependent beings then that collection or succession, Clarke reasons, must have an explanation of its existence. Hence, we can, I think, safely dismiss the charge that the Cosmological Argument involves an erroneous inference from the premise that the members of a collection have a certain property to the conclusion that the collection itself must have that property. We must now deal with the question whether it makes sense to ascribe the property of having an explanation or cause to the infinite collection of dependent beings. Clearly only if it does make sense is the reasoning in support of proposition (3) of the main argument acceptable. Our question, then, is whether it makes sense to ask for a cause or explanation of the entire universe, conceiving the universe as an infinite collection of dependent beings. One recent critic of the Cosmological Argument, Ronald Hepburn, has stated our problem as follows: When we are seriously speaking of absolutely everything there is, are we speaking of something that requires a cause, in the way that events in the universe may require causes? What indeed can be safely said at all about the totality of things? For a great many remarks that one can make with perfect propriety about limited things quite obviously cannot be made about the cosmos itself. It cannot, for instance, be said meaningfully to be 'above* or 'below' anything, although things-in-the-universe can be so related to one another. Whatever we might daim to be 'below the universe* would turn out to be just some more universe. We should have been relating part to part, instead of relating the whole to something not-the-universe. The same applies to 'outside the universe*. We can readily imagine a boundary, a garden wall, shall we say, round something that we want to call the universe. But if we imagine ourselves boring a hole through that wall and pushing a stick out beyond it into a nameless zone 'outside*, we should still not in fact have given meaning to the phrase 'outside the universe*. For the place into which the stick was intruding would deserve to be called a part of the universe (even if con-

310

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 448

WILLIAM L. ROWE

sisting of empty space, no matter) just as much as the area within the walls. We should have demonstrated not that the universe has an outside, but that what we took to be the whole universe was not really the whole. Our problem is this. Supposing we could draw up a list of questions that can be asked about objects in the universe, but cannot be asked about the whole universe: would the question, 'Has it a cause?' be on that list? One thing is dear. Whether or not this question is on the proscribed list, we are not entitled to argue as the Cosmological Argument does that because things in the world have causes, therefore the sum of things must also have its cause. No more (as we have just seen) can we argue from the fact that things in the world have tops and bottoms, insides and outsides, and are related to other things, to the belief that the universe has its top and bottom, inside and outside, and is related to a supra-cosmical something.7

In this passage Hepburn (i) points out that some properties (e.g., *abovef, 'below', etc.) of things in the universe cannot properly be ascribed to the total universe, (ii) raises the question whether 'having a cause* is such a property, and (iii) concludes that " . . . we are not entitled to argue as the Cosmological Argument does that because things in the world have causes, therefore, the sum of things must also have its cause. We noted earlier that the Cosmological Argument (i.e., the version we are examining) does not argue that the sum of things (the infinite collection of dependent beings) must have a cause because each being in the collection has a cause. Thus we may safely ignore Hepburn's main objection. However, his other two points are well taken. There certainly are properties which it makes sense to apply to things within a collection but which it makes no sense to apply to the collection itself. What assurance do we have that 'having a cause1 is not such a property? Suppose we are holding in our hands a collection of ten marbles. Not only would each marble have a definite weight but the collection itself would have a weight. Indeed, from the premise that each marble weighs more than one ounce we could infer validly that the collection itself weighs more than an ounce. This example shows that it is not always fallacious to infer that a collection has a certain property from the premise that each member of the collection has T Ronald W. Hepburn, Christianity and Paradox (London: Watts, 1958), pp. 167-68.

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

311 449

that property. 8 But the collection in this example is, we might say, concrete rather than abstract. That is, we are here considering the collection as itself a physical entity, an aggregate of marbles. This, of course, is not a collection in the sense of a class or set of things. Holding several marbles in my hands I can consider the set whose members are those marbles. The set itself, being an abstract entity, rather than a physical heap, has no weight. Just as the set of human beings has no mother, so the set whose members are marbles in my hand has no weight. Therefore, in considering whether it makes sense to speak of the infinite collection of dependent beings as having a cause or explanation of its existence it is important to decide whether we are speaking of a collection as a concrete entity—for example, a physical whole or aggregate—or an abstract entity. Suppose we view the infinite collection of dependent beings as itself a concrete entity. So far as the Cosmological Argument is concerned one advantage of so viewing it is that it is understandable why it might have the property of having a cause or explanation of its existence. For concrete entities—physical objects, events, physical heaps—can be caused. Thus if the infinite collection is a concrete entity it may well make sense to ascribe to it the property of having a cause or explanation. But such a view of the infinite collection is implausible, if not plainly incorrect. Many collections of physical things cannot possibly be themselves concrete entities. Think, for example, of the collection whose members are the largest prehistoric beast, Socrates, and the Empire State Building. By any stretch of the imagination can we view this collection as itself a concrete thing? Clearly we cannot. Such a collection must be construed as an abstract entity, a class or set.9 But if there are many collections of beings which cannot be concrete entities, what grounds have we for thinking that 8 For a consideration of inferences of this sort in connection with the fallacy of composition see my paper, "The Fallacy of Composition," Mind, January 1962. For some needed corrections of my paper see Bar-Hillel, "More on the Fallacy of Composition." Mind, January 1964. *> Of course, the three members of this collection, unlike the members of the collection of dependent beings, presumably are causally unrelated. But it is equally easy to think of collections which cannot possibly be concrete entities whose members are causally related—e.g., the collection whose members are the ancestors of a given man.

312

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 450

WILLIAM L. ROWE

on the supposition that every being that is or ever was is dependent the collection of those beings would itself be a concrete thing such as a physical heap? At any rate our knowledge of the things (both past and present) comprising the universe and their interrelations would have to be much greater than it presently is before we would be entitled to view the sum of concrete things, past and present, as itself something concrete. But if the infinite collection of dependent beings is to be understood as an abstract entity, say the set whose members include all the beings that are or ever were, haven't we conceded the point to Russell? A set or class conceived of as an abstract entity has no weight, isn't below or above anything, and can't be thought of as being caused or brought into being. Thus if the infinite collection is a set, an abstract entity, isn't Russell right in charging that it makes no more sense to ascribe the property of having a cause or an explanation to the infinite collection than it does to ascribe the property of having a mother to the human race? Suppose that every being that is or ever was is dependent. Suppose further that the number of such beings is infinite. Let A be the set consisting of these beings. Thus no being exists or ever existed that is not a member of A. Does it make sense to ask for an explanation of A's existence? We do, of course, ask questions about sets which are equivalent to questions about their members. For example, 'Is set X included in set Y?' is equivalent to the question 'Is every member of X a member of Y?' I suggest that the question 'Why does A exist?1 be taken to mean 'Why does A have the members that it does rather than some other members or none at all?' Consider, for example, the set of men. Let M be this set. The question 'Why does M exist?' is perhaps odd if we understand it as a request for an explanation of the existence of an abstract entity. But the question 'Why does M exist?' may be taken to mean 'Why does M have the members it has rather than some other members or none at all'? So understood the form of words 'Why does M exist?' does, I think, ask an intelligible question. It is a contingent fact that Hitler existed. Indeed, it is a contingent fact that any men exist at all. One of Leibniz' logically possible worlds is a world that includes some members of M, for example Socrates and Plato, but not others, say Hitler and Stalin. Another is a world in which the set of men is entirely empty and therefore identical with the null set. Why is it then, that M exists? That is, why does M have just the

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

313 451

members it has rather than some other members or none at all? Not only is this question intelligible but we seem to have some idea of what its answer is. Presumably, the theory of evolution might be a part of the explanation of why M is not equivalent to the null set and why its members have certain properties rather than others. But if the question 'Why does M exist?' makes sense, why should not the question 'Why does A exist?' also make sense? A is the set of dependent beings. In asking why A exists we are not asking for an explanation of the existence of an abstract entity, we are asking why A has the members it has rather than some other members or none at all. I submit that this question.does make sense. Moreover, I think it is precisely this question that the proponents of the Cosmological Argument were asking when they asked for an explanation of the existence of the infinite collection or succession of dependent beings.10 Of course, it is one thing for a question to make sense and another thing for there to be an answer to it. The interpretation I have given to the question 'Why does A exist?' is somewhat complex. For according to this interpretation what is being asked is not simply why does A have members rather than having none, but also why does A have just the members it does rather than having some other members. Although the proponents of the Cosmological Argument do seem to interpret the question in this way, it will facilitate our discussion if we simplify the interpretation somewhat by focusing our attention solely on the question why A has the members it has rather than having none. Hence, for purposes of simplification, in what follows I shall take the question 'Why does A exist?' to mean 'Why does A have the members it has rather than not having any?' For any being to be a member of A it is necessary and sufficient that it have the reason of its existence in the causal efficacy of some other being. Imagine the following state of affairs. A has exactly three members: a^ a2, and a3. a3 exists by reason of the causal efficacy of a2 and a2 exists by reason of the causal efficacy of a x . There exists an eternal being b that does not exist by reason of 10 Thus in speaking of the infinite succession, Hume has Demea say: "... and yet it is evident that it requires a cause or reason, as much as any particular object which begins to exist in time. The question is still reasonable, why this particular succession of causes existed from eternity, and not any other succession, or no succession at all'9 (Dialogues, Part IX, italics mine.)

314

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 452

WILLIAM L. ROWE

the causal efficacy of any other being. Since b is not a dependent being, b is not a member of A. At a certain time ax came into existence by reason of the causal efficacy of b. Clearly the question 'Why does A exist?1 when taken to mean 'Why does A have the members it has rather than none at all?', makes sense when asked within the context of this imagined state of affairs. Indeed, part of the answer to the question would involve reference to 6 and itscausal efficacy in bringing about the existence of one of the members of A, namely a^ What this case shows is that the question 'Why does A exist?' is not always (i.e., in every context) meaningless. If Russell holds that the question is meaningless in die framework of the Cosmological Argument it must be because of some special assumption about A that forms part of the context of the Cosmological Argument. The assumption in question undoubtedly is that absolutely every being is dependent. On this assumption every being that is or ever was has membership in A and A has an infinite number of members. Perhaps Russell's view is that within the context of the assumption that every being is dependent it makes no sense to ask why A has the members it has rather than none at all. It makes no sense, he might argue, for two reasons. First, on the assumption that every being is dependent there could not be such a thing as the set A whose members are all dependent beings. For the set A is, although abstract, presumably a being. But if every being is dependent then A would have to be dependent and therefore a member of itself. But apart from whatever difficulties arise when a set is said to be a member of itself, it would seem to make little sense to think of an abstract entity, such as a set, as being caused, as having the reason of its existence within the causal efficacy of some other being. Second, Russell might argue that the assumption that every being is dependent and therefore a member of A rules out the possibility of any answer to the question why A has the members it has rather than none at all. For on that assumption our question about A is in effect a question about the totality of things. And, as Russell observes, "I see no reason whatsoever to suppose that the total has any cause whatsoever." * 1 Neither of these reasons suffices to show that our question about A is meaningless. The first reason does, however, point up the 11 Debate, p. 175.

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

315 453

necessity of introducing some restriction on the assumption 'Every being is dependent' in order that abstract entities like numbers and sets not fall within the scope of the expression 'Every being*. Such a restriction will obviate the difficulty that A is said to be both a member of itself and dependent. I propose the following rough restriction. In speaking of beings we shall restrict ourselves to beings that could be caused to exist by some other being or could be causes of the existence of other beings. God (if he exists), a man, the sun, a stone are beings of this sort. Presumably, numbers, sets and the like are not. The assumption that every being is dependent is to be understood under this restriction. That is, we are here assuming that every being of the sort described by the restriction is in fact a being that exists by reason of the causal efficacy of some other being. The second reason given confuses the issue of whether a question makes sense, is meaningful, with the issue of whether a question has an answer. Of course, given the assumption that every being is a member of A we cannot expect to find the cause or reason of A's existence in some being that is not a member of A. If the explanation for A's existence cannot be found within A itself then we must conclude that there can be no explanation for the infinite collection of dependent beings. But this is to say only that on our assumption that every being is dependent there is no answer to the question "Why does A exist?" It is one thing for a question not to have an answer and quite another thing for the question to be meaningless. We have been examining the first of the two major criticisms philosophers have directed at the reasoning the Cosmological Argument provides in support of the proposition that not every being is dependent. The heart of this criticism is that it makes no sense to ascribe the property of having a cause or explanation to the infinite collection of dependent beings. This criticism, I think, has been shown to be correct in one way, but incorrect in another. If we construe the infinite collection of dependent beings as an abstract entity, a set, it perhaps does not make sense to claim that something caused the existence of this abstract entity. But the question 'Why does A exist?' may be interpreted to mean 'Why does A have the members it does rather than none at all?1 I have argued that taken in this way the question 'Why does A exist?' is a meaningful question. According to the Principle of Sufficient Reason there must be an

316

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 454

WILLIAM L. ROWE

answer to the question 'Why does A exist?1, an explanation of the existence of the infinite collection of dependent beings. Moreover, the explanation must lie either in the causal efficacy of some being outside of the collection or it must lie within the collection itself. But since by supposition every being is dependent—and therefore in the collection—there is no being outside the collection whose causal efficacy might explain the existence of the collection. Therefore, either the collection has the explanation of its existence within itself or there can be no explanation of its existence. If the first alternative is rejected then since the Principle of Sufficient Reason requires that everything has an explanation of its existence we must reject the supposition that every being is dependent. For on that supposition there is no explanation for why there is an infinite collection of dependent beings. The second major criticism argues that the proponent of the Cosmological Argument is mistaken in thinking that the explanation of the existence of the infinite collection cannot be found within the collection itself. The explanation of the existence of the collection is provided, so the criticism goes, once we learn what the explanation is of each of the members of the collection. As we noted earlier, this criticism was succinctly expressed by Hume in his remark: "Did I show you the particular causes of each individual in a collection of twenty particles of matter, I should think it very unreasonable, should you afterwards ask me, what was the cause of the whole twenty. This is sufficiently explained in explaining the cause of the parts." Applying this objection to the infinite collection of dependent beings, we obtain the result that to explain the existence of the infinite collection, A, amounts to no more than explaining the existence of each of its members. Now, of course, A is unlike Hume's collection of twenty particles in that we cannot give individual explanations for each of the members of A. For since A has an infinite number of members we would have to give an infinite number of explanations. But our inability to give a particular explanation for each of the members of A does not imply that there is any member of A for whose existence there is no explanation. Indeed, from the fact that each member of A is dependent (i.e., has the reason of its existence in the causal efficacy of some other being) we know that every member of A has an explanation of its existence, and from the assumption that every being is a member of A we know that for each member of A the explanation lies in the

THE COSMOLOGICAL ARGUMENT CRITICISMS

OF THE COSMOLOGICAL ARGUMENT

317 455

causal efficacy of some other member of A. But, so the criticism goes, if every member of A has an explanation of its existence then the existence of A has been sufficiently explained. For to explain why a certain collection of things exists it is sufficient to explain the existence of each of its members. Hence, since we know that the existence of every one of A's members is explained we know that the existence of the collection A is explained. This forceful criticism, originally advanced by Hume, has gained wide acceptance in contemporary philosophy. Indeed, the only remaining problem seems to be to explain why the proponents of the Cosmological Argument failed to see that to explain the existence of all the members of a collection is to explain the existence of the collection. In restating Hume's criticism, Paul Edwards suggests that perhaps they may have been misled by gram-

mar. The demand to find the cause of the series as a whole rests on the erroneous assumption that the series is something over and above the members of which it is composed. It is tempting to suppose this, at least by implication, because the word 'series' is a noun like 'dog* or 'man*. Like the expression 'this dog' or 'this man' the phrase 'this series' is easily taken to designate an individual object. But reflection shows this to be an error. If we have explained the individual members there is nothing additional left to be explained. Suppose I see a group of five Eskimos standing on the corner of Sixth Avenue and 50th Street and I wish to explain why the group came to New York. Investigation reveals the following stories: Eskimo No. 1 did not enjoy the extreme cold in the polar region and decided to move to a warmer climate. No. 2 is the husband of Eskimo No. 1. He loves her dearly and did not wish to live without her. No. 3 is the son of Eskimos 1 and 2. He is too small and too weak to oppose his parents. No. 4 saw an advertisement in the New York Times for an Eskimo to appear on television. No. 5 is a private detective engaged by the Pinkerton Agency to keep an eye on Eskimo No. 4. Let us assume that we have now explained in the case of each of the five Eskimos why he or she is in New York. Somebody then asks: "All right, but what about the group as a whole; why is it in New York?" This would plainly be an absurd question. There is no group over and above the five members, and if we have explained why each

318

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 456

WILLIAM L. ROWE

of the five members is in New York we have ipso facto explained why the group is there. It is just as absurd to ask for the cause of the series as a whole as distinct from asking for the causes of the individual members.12 The principle underlying the Hume-Edwards criticism may be stated as follows: // the existence of every member of a set is explained the existence of that set is thereby explained. This principle seems to be a corollary of our interpretation of the question 'Why does this set exist?1 For on our interpretation once it is explained why the set has the members it has rather than none at all it is thereby explained why the set exists. And it would seem that if a set A has, say, three members, a x , a2, and a3, then if we explain the existence of a lf a2, and a3 we have explained why A has the members it has rather than none at all. Thus the principle that underlies the second major criticism seems to be implied by our conception of what is involved in explaining the existence of a set. The principle underlying the Hume-Edwards criticism seems plausible enough when restricted to finite sets, i.e., sets with a finite number of members. But the principle is false, I believe, when extended to infinite sets in which the explanation of each member's existence is found in the causal efficacy of some other member. Consider M, the set of men. Suppose M consists of an infinite number of members, each member owing its existence to some other member which generated it. Suppose further that to explain the existence of a given man it is sufficient to note that he was begotten by some other man. That is, where x and y are men and x begat y we allow that the existence of y is explained by the causal efficacy of x. On these suppositions it is clear that the antecedent of the principle is satisfied with respect to M. For every member of M has an explanation of its existence. But does it follow that the existence of M has an explanation? I think not. We do not have an explanation of the existence of M until we have an explanation of why M has the members it has rather than none at all. But clearly if all we know is that there always have been men and that every man's 12 Paul Edwards, "The Cosmological Argument," in Donald R. Burrill (ed.), The Cosmological Arguments (New York: Doubleday & Company, Inc., 1967), pp. 113-14. Edwards' paper was originally published in The Rationalist Annual for the Year 1959.

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

319 457

existence is explained by the causal efficacy of some other man, we do not know why there always have been men rather than none at all. If I ask why M has the members it does rather than none, it is no answer to say that M always had members. We may, I suppose, answer the question "Why does M have the presently existing members it has?" by saying that M always had members and there were men who generated the presently existing men. But in asking why M has the members it does rather than none at all we are not asking why M has the presently existing members it has. To make this clear, we may rephrase our question as follows: "Why is it that M has now and always had members rather than never having had any members at all?" Surely we have not learned the answer to this question when we have learned that there always have been members of M and that each member's existence is explained by the causal efficacy of some other member. What we have just seen is that from the fact that the existence of each member of a collection is explained it does not follow that the existence of the collection is thereby explained. It does not follow because when the collection (set) has an infinite number of members, each member's existence having its explanation in the causal efficacy of some other member, it is true that the existence of every member has an explanation and yet still an open question whether the existence of the set has an explanation. To explain the existence of a set we must explain why it has the members it has rather than none. But clearly if every member's existence is explained by some other member then although the existence of every member has an explanation it is still unexplained why the set has the members it has, rather than none at all. Put somewhat differently, we have seen that the fact (assuming for the moment that it is a fact) that there always have been men, each man's existence brought about by some other man, is insufficient to explain why it is a fact that there always have been men rather than a fact that there never have been any men. If someone asks us to explain why there always have been men rather than never having been any it would not suffice for us to observe that there always have been men and each man has been brought into existence by some other man. I have argued that the second major criticism rests on a false principle, namely, that if the existence of every member of a set is explained then the existence of that set is thereby explained. This

320

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 458

WILLIAM L. ROWE

principle, so far as I can determine, is true when restricted to sets with a finite number of members. For example, if a set A has two members, aj and a2, and if we explain a2 by ax and ax by some being 6 that caused a^ then, I think, we have explained the existence of A. In any case we have explained why A has members rather than none at all. Thus I am not claiming that the principle underlying Hume's objection is always false. Indeed, as I've just indicated, it is easy to provide an example of a finite set of which the principle is true. And perhaps it is just this feature of the principle—i.e., its plausibility when applied to finite sets such as Hume's collection of twenty particles and Edwards' five Eskimos— that has led Hume and many philosophers since Hume to reject the Cosmological Argument's thesis that even if every member of the infinite succession of dependent beings has an explanation the infinite succession itself is not thereby explained. If so, then the mistake Hume and his successors have made is to assume that a principle which is true of all finite sets also is true of all infinite sets. We know, for example, that if we have a set B consisting of five members and a set C consisting of three of the members of B, the members of C cannot be put in one-to-one correspondence with those of B. In reflecting on this fact, it is tempting to conclude that for any two sets X and Y, if all the members of X are members of Y but some members of Y are not members of X then the members of X cannot be put in one-to-one correspondence with those of Y. Indeed, so long as X and Y are restricted to finite sets the principle just stated is true. But if we let X be the set of even natural numbers—2 , 4, 6,.. .—and Y be the set of natural numbers—1, 2, 3, .. . —the principle is shown to be false. For although all the members of X are members of Y and some members of Y—the odd integers—are not members of X, it is not true that the members of X cannot be put in one-to-one correspondence with those of Y. What this example illustrates is that a principle which holds of all finite sets may not hold of all infinite sets. The principle underlying the second major criticism is, I have argued, such a principle. One final point concerning my reply to the second major criticism needs to be made clear. In rejecting the principle on which the criticism rests I have contended that when a set has an infinite number of members, every one of which has an explanation of its existence, it does not follow that the existence of the set is thereby

THE COSMOLOGICAL ARGUMENT CRITICISMS OF THE COSMOLOGICAL ARGUMENT

321 459

explained. In saying this I do not mean to imply that in explaining the existence of every member of an infinite set we never thereby explain the existence of the set, only that we sometimes do not. Specifically, we do not, I think, when we explain the existence of each member of the set by some other member of that set. Recall our example of M, the set of men. If we think of the members of this set forming a temporal series stretching infinitely back in time, each member's existence explained by the causal efficacy of the preceding member, we have an example, I think, in which an explanation of the existence of each member of M does not constitute an explanation of the existence of M. Let us suppose that each man is produced not by another man but by some superior being, say a god. What we are supposing is that M is described as before except that instead of every member having the explanation of its existence in some preceding member of M the explanation is found in the causal efficacy of some member of the set of gods. From eternity, then, gods have been producing men. There have always been members of M and every member has an explanation of its existence. Here it does seem true to say that in explaining the existence of every member of M we have thereby explained the existence of M. If someone asks why there now are and always have been men rather than never having been any, we can say in response that there always have been men because there always have been gods producing them. This, if true, would explain why M has always had members. In this paper I have examined two criticisms that have been advanced against that part of the Cosmological Argument which seeks to establish that not every being can be a dependent being. I have argued that each of these criticisms is mistaken and, therefore, fails as a refutation of the Cosmological Argument. If my arguments are correct, it does not follow, of course, that the Cosmological Argument is a good argument for its conclusion. But it does follow that those philosophers who have rejected the argument for either of the two criticisms discussed in this paper need to reexamine the argument and, if they continue to reject it, provide some good reasons for doing so.

This page intentionally left blank

CHAPTER TWENTY-TWO

The Cosmological Argument

The Cosmological Argument began with Plato and Aristotle, flourished in the writings of Aquinas, Duns Scotus, Leibniz, and Samuel Clarke, and was laid to rest by Hume and Kant. I think its death premature, if not unjustified. But my purpose here is not to resurrect it. My aim in this paper is to uncover, clarify, and examine some of the philosophical concepts and theses essential to the reasoning exhibited in the argument. There are, I believe, two high-watermarks in the history of the Cosmological Argument. The first occurred in the 13th century and is represented by the first three of Aquinas* five ways. The second occurred in the 18th century and is represented by the form the argument takes in the writings of Leibniz and Samuel Clarke. The focus of attention in this paper is on the second of these two high-watermarks; namely, the form the argument takes in the 18th century.1 The Cosmological Argument is an argument for the existence * I have benefited from discussions of the ideas in this paper. In particular I wish to thank Bill Forgie, Bill Gustason, Norman Kretzmann, George Nakhnikian, Kermit Scott, Ted Ulrich, and James Zartman. 1 Although the argument we shall examine is only one version of the Cosmological Argument, for convenience I shall use the phrase "The Cosmological Argument" to denote this argument

324

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

50

NO£S

of God. As such, the argument has two distinct parts. The first part is an argument to establish the existence of a necessary being. The second part is an argument to establish that this necessary being is God. In what follows I shall be concerned solely with the concepts and theses essential to the reasoning exhibited in the first part of the Cosmological Argument, the argument for the existence of a necessary being. Using the expression 'dependent being* to mean 'a being that has the reason for its existence in the causal efficacy of some other being*, and the expression 'independent being* to mean 'a being that has the reason for its existence within its own nature*, we may state the argument for the existence of a necessary being as follows: 1. Every being is either a dependent being or an independent being; 2. It is false that every being is dependent; therefore, 3. There exists an independent being; therefore, 4. There exists a necessary being. This argument contains two premises, propositions (1) and (2), and two inferences, the inference from (1) and (2) to (3), and the inference from (3) to (4). Neither premise is obviously true, and only the first of the two inferences is above suspicion. Let's begin our examination of this argument by considering the second premise. The second premise asserts that not every being is such that its existence is accounted for by the causal efficacy of some other being. Now unless some restriction is placed on die scope of the term *being* what this premise asserts is surely true. For consider such "beings** as numbers or sets. The number two, for example, exists and its existence is not accounted for by the causal efficacy of some other being. Hence, it*s true that not every being is a dependent being. This interpretation and justification of the second premise, however, is of no help to the proponent of the Cosmological Argument. We must, then, in his behalf, place some restriction on the scope of the term *being* as it is used in the second premise. I propose the following rough restriction. In speaking of beings in the second premise we shall restrict ourselves to beings that could be caused to exist by some other being or could be causes of the existence of other beings. God (if He exists), a man, a star, a stone are beings of this sort. Presumably, numbers, sets and the like are not. What premise (2) asserts, then, is that not every being of the

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

325 51

sort described by this restriction is a being that exists by reason of the causal efficacy of some other being. Granted this restriction, why should we accept the second premise of the Cosmological Argument? Well, if every being that exists (or ever existed) is dependent then the whole of existing things (i.e., things that could be caused or could be causes), it would seem, consists of a collection of dependent beings, that is, a collection of beings each member of which exists by reason of the causal efficacy of some other being. This collection would have to contain an infinite number of members. For suppose it contained a finite number, let us say three, 0, &, and c. Now if in Scotus* phrase "a circle of causes is inadmissible" then if c is caused by b and b by a, a would exist without a cause. But in that case a would not be what by supposition it is, namely, a dependent being. Hence, if we grant that a circle of causes is inadmissible, it is false that the whole of existing things consists of a collection of dependent beings finite in number. Suppose, then, that the dependent beings making up the collection are infinite in number. Why is it false that the whole of existing things consists of such a collection? The proponent of the Cosmological Argument answers as follows.2 The infinite collection itself, he argues, requires an explanation of its existence. For since it is true of each member of the collection that it might not have existed, it is true of the whole infinite collection that it might not have existed. What then accounts for the fact that there is an infinite collection of dependent beings, rather than no dependent beings at all? The explanation cannot lie in the causal efficacy of some being outside of the collection, since by supposition the collection itself includes every being capable of causal efficacy. Nor, it seems, can the explanation of the existence of the collection be found within the collection itself. Hence, if we assume that every being is dependent we must conclude that there exists an infinite collection of dependent beings for whose existence there is no explanation whatever. But since premise (1) tells us that whatever exists has an explanation of its existence, it cannot be that there exists an infinite collection of dependent beings for whose existence there is no explanation. The assumption which led to this conclu2

See, for example, Samuel Clarke's discussion of Propositions II and III in his A Demonstration of the Being and Attributes of God (9th Edition, London, 1738). This discussion is summarized by Hume in Fart IX of his Dialogues Concerning Natural Religion.

326

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

52

NO£S

sion must, therefore, be false. Since the assumption in question is that every being is dependent, its denial, premise (2), must be true. Critical discussion of the reasoning just given has centered around two questions: I. Is it true that there must be an explanation of the existence of the collection of dependent beings? II. Is it true that if every being is dependent there will be no explanation of the existence of the collection of dependent beings? The 18th century proponents of the Cosmological Argument contend that reason requires affirmative answers to these questions. But many philosophers from Hume to Russell either have argued that these questions (one or both) must be answered negatively or have argued that there are no good reasons for the view that each requires an affirmative answer. How are we to settle the disagreements concerning the answers to these two questions? G. E. Moore once observed that philosophical disagreements are mainly due to the attempt to answer questions, without first discovering precisely what question it is which you desire to answer. The philosophical disagreements concerning the answers to our two questions illustrate, I believe, Moore's observation. For we cannot answer either question without first discovering what it is for there to be an explanation of the existence of the collection of dependent beings. I suggest, then, that we begin with the prior question: III. What are the necessary and sufficient conditions for there to be an explanation of the existence of the collection of dependent beings? There is, unfortunately, no direct answer to this question in the writings of the 18th century proponents of the Cosmological Argument. But it is clear, I think, that the following two conditions are, in their view, necessary, if not sufficient, for there to be an explanation of the existence of the collection of dependent beings. Cl. There is an explanation of the existence of each of the members of the collection of dependent beings. C2. There is an explanation of why there are any dependent beings.

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

327 53

In what follows I shall assume that the existence of the collection of dependent beings will have an explanation only if Cl and C2 obtain. Are the proponents of the Cosmological Argument right in giving an affirmative answer to the second of our two questions? Are they right, that is, in contending that on the supposition that every being is dependent there will be no explanation of the existence of the infinite collection of dependent beings? It is, of course, true that Cl will obtain. That is, each member of the collection will have an explanation of its existence, the explanation being found in the causal efficacy of some other dependent being. When, therefore, the proponents of the Cosmological Argument contend that on the supposition that every being is dependent there will be no explanation of the infinite collection of dependent beings they are not contending that some member of that collection will lack an explanation, they are not contending that Cl will not obtain. What they are contending is that C2 will not obtain. They are contending that if the whole of existing things consists of a collection of dependent beings there will be no explanation of why there are (or have ever been) any dependent beings at all. For by supposition there is no being outside of the collection of dependent beings whose causal efficacy might provide the explanation. Nor can the explanation be found within the collection itself. For since there might have been no dependent beings at all we cannot explain why there are dependent beings by appealing to the existence and causal efficacy of some member of the collection of dependent beings—any more than we can explain why there are men by appealing to the existence and causal efficacy of Adam. Nor can we explain why there are (or have ever been) dependent beings by appealing to all the members of the infinite collection of dependent beings. For if the question to be answered is why there are (or have ever been) any dependent beings at all, we cannot answer that question by noting that there always have been dependent beings, each one accounting for the existence of some other dependent being. Thus on the supposition that every being is dependent there will be no explanation of why there are dependent beings. C2 will not obtain. Therefore, on the supposition that every being is dependent there will be no explanation of the existence of the collection of dependent beings. Although I think the proponents of the Cosmological Argument are right in their contention that the supposition that every

328

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

54

Nofts

being is dependent leaves us with no explanation of the existence of the collection of dependent beings, there is a difficulty in their reasoning to this conclusion. For they contend that there might (logically) have been no dependent beings at all. If this were not so, if, that is, it were logically necessary that dependent beings exist, I think the proponents of the argument would concede either that no explanation of why there are dependent beings is required or that if an explanation is required it might lie within the collection of dependent beings—for the nature of the collection is such, it might be argued, that it is logically necessary for there to exist members of it. Thus it seems that the proponents of the Cosmological Argument must claim that it is possible for there to be no dependent beings at all. But how can it be established that it is logically possible for there to be no dependent beings? In considering the supposition th$t every being is dependent, Samuel Clarke remarks: According to this . . . supposition, there is nothing in the universe self-existent or necessarily-existing. And if so, then it was originally equally possible that from eternity there should never have existed any thing at all; as that there should from eternity have existed a succession of changeable and dependent beings.3

Apparently, Clarke tries to establish the proposition that it is possible for there to be no dependent beings by inferring it from the proposition that no dependent being necessarily exists. The inference he makes may be expressed as follows: A. If for every dependent being it is possible that it doesn't exist then it is possible that every dependent being doesn't exist. The difficulty is that this inference does not appear to be sanctioned by any valid rule of modal logic. We may take A as an instance of B. B. If for every being it is possible that it has P then it is possible that every being, has P. But B is an invalid principle. Of course, from the fact that A is an instance of an invalid principle it doesn't follow that A isn't a necessary truth. A might (for aU we know) be a necessary truth. But if A isn't a necessary truth then the proponents of the Cosmological * Demonstration, (9th Edition), p. 14.

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

329 55

Argument are mistaken in thinking that the proposition that it is possible for there to be no dependent beings follows from the proposition that no dependent being is such that it necessarily exists. Furthermore, even if A is a necessary truth, since we do not seem to know that it is, appealing to it cannot enable us to know that it is possible for there to be no dependent beings. Hence, it appears that a vital portion of the reasoning in the Cosmological Argument rests on the unproved premise that there might (logically) have been no dependent beings. In spite of this difficulty it is reasonable, I think, to concede to the proponent of the Cosmological Argument the premise that it is possible that no dependent beings exist. Indeed, some critics hold views which, I think, entail the truth of that premise. In any case, if we grant that premise I think we must conclude that the proponents of the Cosmological Argument are right in their contention that reason requires an affirmative answer to the second of the two critical questions noted earlier. They are right in contending that if every being is dependent there will be no explanation of the existence of the collection of dependent beings—for if every being is dependent there will be no explanation of why there are any dependent beings at all. Turning now to the first of our two questions: Is it true that there must be an explanation of the existence of the collection of dependent beings? Many critics are prepared to grant that on the supposition that every being is dependent there will be no explanation of the collection of dependent beings. What they deny is the need for an explanation. Why, then, do the proponents think there must be an explanation of the existence of the collection of dependent beings? There is, I believe, a rather widespread misconception concerning this matter. Many critics think that the proponents of the Cosmological Argument infer that the collection of dependent beings has an explanation from the premise that every member of that collection has an explanation.4 Rejecting this inference as fallacious, the critics conclude that the proponents' contention that there must be an explanation of the collection of dependent beings is unwarranted. Perhaps some proponents of the Cosmological Argument have used the inference which the critics reject as fallacious. But many 4 See, for example, Ronald W. Hepburn, Christianity and Paradox (London: Watts, 1958), pp. 167-168.

330

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 56

Nods

of them have not. Samuel Clarke, for example, reasons that the collection of dependent beings must have an explanation of its existence by appealing to the principle expressed in the first premise of the Cosmological Argument—namely, the Principle of Sufficient Reason. That principle assures us that whatever exists has an explanation of its existence. But if there exists an infinite collection of dependent beings then, Clarke reasons, there must be an explanation of the existence of that collection. Is Clarke right in thinking that the Principle of Sufficient Reason, as expressed in the first premise, entails that the collection of dependent beings has an explanation of its existence? Is he right, that is, in accepting the following inference? PSRi: Every existing being has a reason for its existence either within itself or in the causal efficacy of some other being; therefore, 5. The infinite collection of dependent beings has an explanation of its existence. I doubt that Clarke is right in accepting this inference. For, as we saw earlier, to explain the existence of the collection of dependent beings is not to explain the existence of some particular being. It is, in part, to explain why there are any dependent beings at all. And it does not appear to be inconsistent for it to be true that every being has an explanation of its existence and yet false that there is an explanation of why there are any dependent beings at all. The question I'm raising about the inference of (5) from PSRi is one that apparently did not occur to the 18th century proponents of the Cosmological Argument. The reason for this, I think, is that the proponents of the argument never saw clearly that the collection of dependent beings is not itself a dependent being. They tended to confuse the question of why the collection of dependent beings has members (rather than not having any) with the altogether different question of why a certain being exists. Thus in considering the view that the infinite succession of dependent beings might itself not have an explanation, Clarke remarks: "Tis in reality, and in point of argument, the very same supposition as it would be to suppose one continued being, of beginningless and endless duration, neither self-existent and necessary

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

331 57

in itself, nor having its existence founded in any self-existent cause."5 Although the inference of (5) from PSRi is dubious, we must not dismiss too quickly the contention that (5) is entailed by the Principle of Sufficient Reason. For that principle is a good deal more general than PSRi. PSRi implies only that every existing thing has a reason for its existence. Thus if we come upon a man in a room, PSRi implies that there must be a reason why that particular man exists. But, as stated, PSRi doesn't imply that there must be a reason why the man in question is in the room he's in, rather than somewhere else. As stated, PSRi doesn't imply that there must be a reason why the man in question is in poor health, say, or a reason why he is at the moment thinking of Paris rather than, say, Lafayette, Indiana. But surely the Principle of Sufficient Reason is meant to imply that there must be a reason not only for the existence of the man in question but also for the fact that he's in a particular room, in poor health, and thinking of Paris. Leibniz' statement of the principle makes this clear. "The principle in question is the principle of the want of a sufficient reason in order to anything's existing, in order to any event's happening, in order to any truth's taking place."6 Clarke also appears to understand the Principle of Sufficient Reason in this general way. Thus he remarks: "Nothing can be more absurd than to suppose that any thing (or any circumstance of any thing) is, and yet that there be absolutely no reason why it is, rather than not."7 Perhaps, then, when expressed in its more general form, the Principle of Sufficient Reason does entail that there is an explanation of the existence of the collection of dependent beings. Suppose we take the Principle of Sufficient Reason in its general form to be the following: PSR2: Every actual state of affairs has a reason either within itself or in some other state of affairs. PSR2, I think, does entail that there is an explanation of the existence of the collection of dependent beings. But it suffers from the defect that there are reasons for rejecting it as false. For consider the following argument. & Demonstration, (9th Edition): 13-14. 6 Leibniz-Clarke Correspondence, 5th Letter. 7 "Letter to a Critic", published as an appendix in the 9th Edition of his Demonstration, p. 490.

332

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

58

Note 1. There are positive, contingent states of affairs; 2. If there are positive, contingent states of affairs then there is some state of affairs for which there is no reason;

therefore, 3. PSR2 is false. The reasoning in support of premise (2) in this argument proceeds as follows. Consider the state of affairs S expressed by the proposition There are positive, contingent states of affairs'. It's fair to assume that S is itself a contingent state of affairs. There must, then, according to PSR2 be some state of affairs which is the reason for S, which accounts for the fact that there are positive, contingent states of affairs rather than not. Suppose someone says that the state of affairs constituted by the exitsence of God is what accounts for S, giving as his reason that God caused there to be positive, contingent states of affairs. We then ask whether the existence of God is a contingent state of affairs. If the existence of God is a contingent state of affairs then, since it clearly is a positive state of affairs, it cannot account for S, it cannot explain why there are positive, contingent states of affairs rather than not—no more than citing the existence of Adam and his act of generating children can explain why there are any men rather than none. Hence, the defender of PSR2 must say that the state of affairs constituted by the existence of God is a necessary state of affairs. We then ask whether the state of affairs recorded by the proposition 'God caused there to be positive, contingent states of affairs' is a contingent or a necessary state of affairs. If Gods causing there to be positive, contingent states of affairs is itself a contingent state of affairs, then, since it is clearly a positive state of affairs, it cannot account for S, it cannot explain why there are positive, contingent states of affairs rather than not.8 That is, if the question to be answered is "Why are there any positive, contingent states of affairs rather than noner^ we cannot answer it by appealing to some positive, contingent state of affairs. Consequently, the defender of PSR2 must say that God's causing there to be positive, contingent states of affairs is itself a necessary state of affairs. But if the existence of God is necessary and God's causing there to be positive, contingent states of affairs 8 I'm supposing throughout this argument that any state of affairs which explains the fact that there are positive, contingent states of affairs must itself be a positive state of affairs or entail some positive state of affairs.

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

333 59

is also necessary, it follows that it is necessary that there are positive, contingent states of affairs. But, as we noted at the outset, it is contingent, riot necessary, that there are positive, contingent states of affairs. Consequently, since PSR2 implies that there is a reason for S, implies that there is an explanation of the fact that there are positive, contingent states of affairs, PSR2 is false.9 It seems, then, that neither the general form of the Principle of Sufficient Reason (PSR2) nor its more restricted form (PSRi) can be used to justify the contention that the existence of the collection of dependent beings has an explanation. For PSRi, although perhaps true, does not entail proposition (5). And PSR2, although it entails proposition (5), does not appear to be true. Perhaps the best move for the proponent of the Cosmological Argument to make at this juncture is to formulate a version of the Principle of Sufficient Reason which will yield the desired result concerning the collection of dependent beings but which will not fall victim to the argument against PSR2. He might, for example, appeal to the following principle: PSR3: For every collection whose members are existing beings (which can be caused to exist or which can cause the existence of other beings) there is an explanation of the fact that it has members. PSR3 does not imply that every positive, contingent state of affairs has an explanation; it does not fall prey to the argument that refutes PSR2. PSR3, however, does imply that there is an explanation of the existence of the collection of dependent beings. Hence, if PSR3 is true the proponents of the Cosmological Argument are right in giving an affirmative answer to the first of the two main questions noted earlier; they are right, that is, in contending that there must be an explanation of the existence of the collection of dependent beings. But is PSR3 true? Indeed, is even PSRi, the first premise of the Cosmological Argument, true? These are important questions. Any final judgment of the Cosmological Argument depends on how they are answered. Some philosophers thought that the truth of PSRi could be 9 For an argument which is similar to the one just presented see James F. Ross, Philosophical Theology (Indianapolis: The Bobbs-Merrill Company, Inc., 1969): 295-304. Although similar, Ross's argument is logically independent of the one given above.

334

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

60

Note

proved. But their arguments to this end are notoriously bad.10 Of course, if PSRi cannot be proved, it does not follow that it cannot be known to be true. For if we know any propositions to be true there must be some propositions which we can know to be true without having to prove them, without having to derive them from other propositions we know to be true. But it is doubtful that any of us knows that PSRi is true. Other philosophers have claimed that although PSRi is not known to be true it is, nevertheless, a presupposition of reason, a metaphysical assumption that all men make, whether or not they reflect sufficiently to become aware of the assumption.11 If true, what bearing would this view have on the Cosmological Argument? It would not, of course, show that it is a sound argument. For PSRi could still be false. The fact, if it is a fact, that all of us presuppose that every existing being has an explanation of its existence does not imply that no being exists without an explanation of its existence. Nature is not bound to satisfy our presuppositions. As James has remarked in another connection: In 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/' However, if we do make such a presupposition we could not consistently reject the Cosmological Argument solely because it contains PSRi as a premise. But it is doubtful that all men do make this metaphysical assumption. Perhaps the truth we can know about the Cosmological Argument is this. Although it may be a perfectly sound argument, it cannot reasonably be maintained to be a proof of its conclusion. For although its premises may be true, we are not in the position of knowing that they are true. And since to claim of an argument that it is a proof of its conclusion is to imply that its premises are known to be true, we are not entitled to claim that the Cosmological Argument is a proof of the existence of a necessary being. I am aware that this conclusion will satisfy neither the proponents of the argument nor its critics. The critics wish to show that the argument is invalid, contains false premises, or premises arrived at by fallacious reasoning. And if our judgment is correct 10 For a discussion of some of these arguments see my paper "The Cosmological Argument and the Principle of Sufficient Reason", Man and World 2 (1968): 278-292. 11 See, for example, Richard Taylor, Metaphysics (Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1963): 86-87.

THE COSMOLOGICAL ARGUMENT THE COSMOLOGICAL ARGUMENT

335 61

this wish must go unfulfilled. The proponents wish to show that the argument provides a good and conclusive reason for accepting its conclusion. This wish too must remain unfulfilled. Like most important philosophical arguments, it appears that the Cosmological Argument is neither as good as its supporters have claimed nor as bad as its critics have believed.

This page intentionally left blank

CHAPTER TWENTY-THREE

Cosmological Arguments

Within philosophy of religion, a cosmological argument is understood to be an argument from the existence of the world to the existence of God. Typically, such arguments proceed in two steps. The first step argues from the existence of the world to the existence of a first cause or necessary being that accounts for the existence of the world. The second step argues that such a first cause or necessary being has, or would very likely have, the properties associated with the idea of God. Cosmological arguments appeared in Plato and Aristotle, played a prominent role in Jewish, Christian, and Islamic thought during the medieval period, and were forcefully presented in the eighteenth century by the German philosopher Leibniz and the English theologian Samuel Clarke. In the modern period these arguments, particularly as presented by Aquinas, Leibniz, and Clarke, have been severely criticized by Hume, Kant, and others. Since the second half of the twentieth century, however, there has been a revival of interest in cosmological arguments, and several challenges to the major criticisms of these arguments have appeared. Cosmological arguments may be divided into two broad types: those that depend on a premise denying an infinite regress of causes and those that do not depend on such a premise. Among the former are contained the first "three ways" presented by Aquinas, as well as an interesting argument, developed by Islamic thinkers, that the world cannot be infinitely old and, therefore, must have come into existence by the creative will of God. An important difference between the arguments represented by Aquinas's first "three ways" and the Islamic argument is that while both reject an infinite regress of causes, only the latter bases the objection on the alleged impossibility of an infinite temporal regress. Unlike Bonaventure, who adopted the Islamic argument, Aquinas did not think that philosophy could show that the world had a temporal beginning. But whether the world had a temporal beginning or has always existed, he insisted that the world requires a cause. For, according to Aquinas, things that now exist require a presently existing cause that sustains them in existence, and the series of sustaining

338

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe causes (an essentially ordered series of causes), he argued, does terminate in a first cause. He therefore rejected an infinite regress of essentially ordered causes (a non-temporal causal series), identifying God as the first cause in such a nontemporal series. Leibniz and Clarke, however, allowed an infinite regress of causes, arguing only that there must be a sufficient reason for the existence of such a series of causes. Thus, unlike the cosmological arguments of both Bonaventure and Aquinas, the eighteenth-century arguments of Leibniz and Clarke do not depend on rejecting an infinite regress of causes. Appealing to the principle of sufficient reason, Leibniz and Clarke insist only that such a series could not be self-explanatory and, therefore, would require an explanation in the causal activity of some being outside the series. Cosmological arguments relying on philosophical objections to an infinite temporal series of causes typically proceed as follows. (1) Whatever begins to exist has a cause. (2) The world began to exist. (3) Therefore, the world has a cause of its existence. The philosophical argument for premise (2) is based on the alleged impossibility of an infinite series of past events. Why is such a series thought to be impossible? If we begin with some present event and consider further events proceeding endlessly into the future, such a series is potentially infinite. For at any future event in the series there will have actually occurred only a finite number of events between that event and the present event. But if we think of events preceding endlessly into the past from the present, we would be thinking of an infinite series that has actually occurred, a series that is actually infinite. The claim is that while a series of events can be potentially infinite, it cannot be actually infinite. So, the world could not have always existed. Arguments in support of this view are presented in William Craig (1979). It must be admitted that it is difficult to imagine an absolutely infinite number of temporally discrete events having already occurred. But just what is the philosophical objection to it? It is sometimes suggested that if the series of events prior to the present is actually infinite, then there must be events in the past that are separated from the present by an infinite number of events. But this suggestion seems mistaken. As Quentin Smith (1987) argues, no particular past event is separated from the present by an infinite number of events. It is sometimes suggested that if the past is actually infinite then new events cannot be added to the series, for the series thus added to would be the same size as the series before the addition was made. The response given to this objection is that one can add to an infinite collection even though the number of entities in the collection before the addition will be the same as the number of entities in the collection after the addition. The fact that this is so does not prevent the old collection from being a proper subset of the collection composed of the old collection and the new member. For reasons such as these, many philosophers who have studied 104

THE COSMOLOGICAL ARGUMENT Cosmological Arguments these matters remain unconvinced that an actually infinite series of past events is impossible. In addition to the philosophical argument against the possibility that the world has always existed, some proponents endeavor to support premise (2) by appealing to recent scientific theories that imply that the world had a beginning. For example, they appeal to the Big Bang theory, according to which the universe probably began to exist some 10 to 20 billion years ago. After considering the more strictly philosophical versions of the cosmological argument we will examine the Big Bang theory and the version of the cosmological argument based on that theory. A good example of a cosmological argument based on a rejection of a nontemporal infinite regress of causes is Aquinas's second way. This argument may be summarized as follows: (1) Some things exist and their existence is caused. (2) Whatever is caused to exist is caused to exist by something else. (3) An infinite regress of (non-temporal) causes resulting in the existence of anything is impossible. ' (4) Therefore, there is a first cause of existence. There are two major difficulties in assessing the third premise of this argument. First, there is the difficulty of understanding exactly what a non-temporal causal series is. Second, there is the difficulty of determining exactly why such a series cannot proceed to infinity. To resolve the first difficulty we must distinguish the earlier cause that brought some presently existing object into existence from whatever presently existing things are causally responsible for its existence at this very moment. The basic idea is that if A (a human being, say) now exists, A is right now being caused to exist by something else B, which may itself be simultaneously caused by C to be causing A to exist. Although A would not exist now had it not been brought into existence by something else that existed temporally prior to A (a temporal causal series), it is also true, so Aquinas thought, that A would not now exist were it not now being caused to exist by something else B (a non-temporal causal series). In such a non-temporal series of causes of A's present existence, Aquinas held that the cause of any member in the series either is the first cause in the series or is itself being caused to cause that member by some non-temporally prior cause in the series. Although Aquinas allowed that it is theoretically possible for a temporal series of causes to proceed backwards to infinity, he thought it obvious that a nontemporal causal series must terminate in a first member, itself uncaused. But why is this supposed to be obvious? Presumably, the idea is that it is obvious that if B is right now causing A to exist, and C is right now causing B to be causing A to exist, then if C and every prior member in the series were to have the same status as B, no causing would be occurring at all. Or, to put it differently, if there were no first cause in this series it would be simply inexplicable that such a series of 105

339

340

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe causings is actually occurring. But once the argument is put in this fashion it invites the skeptical challenge that the fact that such causing goes on may simply be inexplicable. Thus, understanding the third premise of this argument and determining • exactly why it must be true has proved to be difficult. And, of course, it would be question-begging to simply define a non-temporal causal series as one that terminates in a first cause. As a result, many philosophers find the argument unconvincing. As noted above, the cosmological arguments developed by Leibniz and Clarke do not depend on a premise that rejects an infinite regress of causes. What they do depend on is a rather strong explanatory principle according to which there must be a determining reason for the existence of any being whatever. If we think of a dependent being as a being whose determining reason lies in the causal activity of other beings, and think of a self-existent being as a being whose determining reason lies within its own nature, the first step of Clarke's cosmological argument can be put as follows. (1) Every being (that exists or ever did exist) is either a dependent being or a self-existent being. (2) Not every being can be a dependent being. (3) Therefore, there exists a self-existent being. While the principle that there must be a determining reason for the existence of any being whatever immediately yields premise (1), it is difficult to see how it establishes premise (2). For if we allow for an infinite regress of dependent beings, each having the reason for its existence in some preceding member of the series, it is difficult to see how any being exists that lacks a reason or explanation of its existence. Of course, if we view the infinite series of dependent beings as itself a dependent being, we might argue that unless there is a self-existent being there would be no determining reason for the existence of the series itself. But it does not seem right to view the succession or series of dependent beings as still another dependent being. So, as strong as the principle we are considering appears to be, it does not appear to be strong enough to do away with the supposition that every being that exists or ever did exist is a dependent being. To carry out this task the cosmological arguments of Leibniz and Clarke required a stronger principle, the principle of sufficient reason (PSR). The explanatory principle we've been considering is restricted to requiring an explanation for the existence of individual beings. PSR is a principle concerning facts, including facts consisting in the existence of individual beings. But PSR also requires an explanation for facts about individual beings, for example, the fact that John is happy. In addition, PSR requires an explanation for general facts such as the fact that someone is happy or the fact that there are dependent beings. Leibniz expresses PSR as the principle "that no fact can be real or existent, no statement true, unless there be a sufficient reason why it is so and not otherwise" (Leibniz, 1951, paragraph 32). And Clarke asserts: "Undoubtedly 106

THE COSMOLOGICAL ARGUMENT Cosmological Arguments nothing is, without a sufficient reason why it is, rather than not; and why it is thus, rather than otherwise" (Clarke, 1956, third reply). If we understand a contingent fact to be a fact that possibly might not have been a fact at all, it is clear that Leibniz held that every contingent fact has a sufficient reason or explanation. And so long as we restrict ourselves to contingent facts concerning the existence of things, it is clear that Clarke held that all such facts must have a sufficient reason. If either view should be correct, it does seem that Clarke's premise (2) must be true. For if every being were dependent it does seem that there would be a contingent fact without any explanation - the fact that there are dependent beings. But if PSR is true, the fact that there are dependent beings must have an explanation or sufficient reason. So, given Clarke's convictions about PSR, it is understandable why he should hold that not every being can be a dependent being. For if every being that exists or ever did exist is a dependent being, what could possibly be the sufficient reason for the fact that there are dependent beings? It won't do to point to some particular dependent being and observe that it produced other dependent beings. The question why there are any dependent beings cannot be answered by appealing to the causal activity of some particular dependent being, any more than the question why there are any human beings can be answered by appealing to Adam and Eve and their causal activity in producing other human beings. Nor will it do to observe that there always have been dependent beings engaged in causing other dependent beings. The question why there are any dependent beings cannot be answered by noting that there always have been dependent beings any more than the question why there are any elephants can be answered simply by observing that there always have been elephants. To note that there always have been elephants may explain how long elephants have been in existence, but it won't explain why there are elephants at all. Should we conclude that the premises in Clarke's cosmological argument are true? No. For at best all we have seen is that the premises are true z/PSR is true. But what of PSR itself? Is it true? In its unrestricted form PSR holds that every fact has an explanation; in its restricted form it holds that every contingent fact has an explanation. Even if we take PSR in its restricted form, there are serious objections to it. One objection to PSR is that it cannot avoid the dark night of Spinozism, a night in which all facts appear to be necessary. This difficulty was particularly acute for Leibniz. He explained God's creation of this world by this world's being the best and God's choosing to create the best. But what accounts for God's choosing to create the best, rather than some inferior world or none at all? God chooses the best because of his absolute perfection - being absolutely perfect he naturally chooses to create the best. The difficulty is that God's being perfect is, for Leibniz, a necessary fact. It seems, then, that God's choice to create the best must also be necessary and, consequently, the existence of this world is necessary. If we avoid this conclusion by saying that God's being perfect is not the sufficient reason of his choice to create the best we run into an infinite regress 107

341

342

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe of explanations of his choice to create the best. For suppose we say that it is God's perfection in conjunction with his choice to exercise his goodness that constitutes the sufficient reason for his choice to create the best. What then of his choice to exercise his goodness? A similar problem would arise in providing a sufficient reason for it. And we seem to be off to the races, each reason determining a choice only by virtue of a prior choice to act in accordance with that reason. The problem of God's creation being necessary can be avoided by denying that there is a best possible world. For, if for every possible world there is a better possible world, it would seem to suffice that God creates, perhaps freely, a very good world, even though for any such world there is a better he could have created instead. But there would still remain a problem for PSR. PSR requires that God have a sufficient reason to create the particular world he chooses to create. But it is doubtful on the no best world hypothesis that God would have a sufficient reason for creating the world he does, rather than some equally good or better world. A second objection to PSR holds that it is impossible for every contingent fact to have an explanation. For consider the huge conjunctive fact whose conjuncts are all the other contingent facts that there are. This huge conjunctive fact must itself be a contingent fact, otherwise its conjuncts would not all be contingent. Now what can be the sufficient reason for this huge conjunctive fact? It cannot be some necessary fact. For the sufficient reason for a fact is another fact that entails it; and whatever is entailed by a necessary fact is itself necessary. The huge conjunctive fact cannot be its own sufficient reason since only a necessary fact could be self-explanatory. So, the sufficient reason for the huge conjunctive fact would have to be one of the contingent facts that is a conjunct of it. But then that conjunct would have to be a sufficient reason for itself, since whatever is a sufficient reason for a conjunctive fact must be a sufficient reason for each of its conjuncts. It follows, then, that the huge conjunctive fact cannot have an explanation. It thus appears that PSR is false. I am inclined to think this argument against the truth of PSR has merit. However, Vallicella (1997) has criticized it on the grounds that a simple list of truths - say, p, q, r, etc. - may suffice to explain the contingent, conjunctive fact that (p & q & r & etc.). And if this is so, then the fact that each conjunct in the infinite conjunction is true may explain the fact that the infinite conjunction is true. Whereas, it would be odd to suggest the reverse: that each conjunct's being true is explained by the fact that the infinite conjunction is itself true. In the above argument it is important not to confuse the huge conjunctive fact constituted by'every other contingent fact with the general fact that there are contingent facts. The latter fact - that there are contingent facts - is not itself a contingent fact. It is a necessary fact. For every possible world contains some contingent fact or other. Consider the contingent fact that there are elephants. That there are elephants is a fact in the actual world. But if some possible world in which there are no elephants were to be actual, it would be a fact that there are no elephants. So, no matter what possible world is actual, either that there are 108

THE COSMOLOGICAL ARGUMENT Cosmological Arguments elephants will be a fact or that there are no elephants will be a fact. Thus, that there are contingent facts is itself a necessary fact. But the huge conjunctive fact described above is itself a contingent fact. Had some other possible world been actual, the huge conjunctive fact described above would not have been a fact. A third, somewhat more complicated, argument against PSR has been advanced by Rowe (1975). Consider the idea of a positive, contingent state of aftairs, where x is a positive contingent state of affairs if and only if from the fact that x obtains it follows that at least one contingent being exists. Given this account, it is clear, for example, that there being elephants is a positive contingent state of affairs. For, since elephants are contingent beings, from the fact that there being elephants obtains, it follows that at least one contingent being exists. There being no unicorns, however, is not a positive contingent state of affairs. Consider now the following state of affairs: (T)

There being positive contingent states of affairs.

(T) itself also is a positive contingent state of affairs. For if (T) obtains then it must also be true that at least one contingent being exists. But could there be a sufficient reason for the fact that (T) obtains? No. That it is impossible for there to be a sufficient reason for the fact that (T) obtains follows from two considerations. First, any sufficient reason (full explanation) for the fact that (T) obtains would itself be a positive contingent state of affairs. For from the fact that (T) obtains it follows that at least one contingent being exists. Therefore, since any sufficient reason for (T) would entail (T), and therefore entail whatever is entailed by (T), any sufficient reason for (T) would entail that at least one contingent being exists, and thus would itself be a positive contingent state of affairs. This first consideration establishes that any sufficient reason for (T) must itself be a positive contingent state of affairs. Now we come to the second consideration. Any sufficient reason for the fact that (T) (there being positive contingent states of affairs) must constitute a full explanation for why there are positive contingent states of affairs. But surely, nothing that itself is a positive contingent state of affairs can be an explanation for why there are positive contingent states of affairs. For such a proposed explanation is simply circular. Rowe suggests that we can see the circularity involved by considering the following example. Suppose we try to explain why there are positive contingent states of affairs by citing the fact, let us suppose, that God willed that positive contingent states of affairs be actual - just as, for example, we might explain why there are men by citing the (supposed) fact that God willed that men should exist. The fact, then, consisting of God's willing that positive contingent states of affairs be actual is what explains why there are positive contingent states of affairs. But now let us consider the fact of God's willing that positive contingent states of affairs be actual. If that fact does explain why there are positive contingent states of aftairs it must entail that some positive contingent states of affairs are actual. And if this is so, then the fact that God willed that there be positive contingent states of affairs entails that 109

343

344

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe at least one contingent being exists. We then ask whether the fact in question is contingent or necessary. It cannot be necessary, for then it would be necessary that at least one contingent being exists - and, as we have seen, it seems to be a contingent matter that contingent beings exists. What follows, then, is that the fact consisting of God's willing that positive contingent states of affairs be actual is itself*, positive contingent state of affairs; for it is contingent and, from the fact that it obtains, it follows that at least one contingent being exists. But clearly, the fact that accounts for why there are positive contingent states of affairs cannot itself be a positive contingent state of affairs. As with the previous two objections to PSR, this objection also has been criticized in the literature on the cosmological argument. Quentin Smith (1995) argues that it is a mistake to think that there can be no sufficient reason for the fact that there are positive contingent states of affairs. All that needs to be explained is that there obtain positive contingent states of affairs, which is logically equivalent to the state of affairs that there are contingent concrete objects. Now this state of affairs does appear to have an explanation, namely, by the state of affairs: (s) God wills that there are contingent concrete objects. (Smith, 1995, p. 240)

What Smith says here seems right. The fact (supposing it is a fact) that God wills that there be contingent concrete objects does explain the fact that there are contingent concrete objects. But the issue is whether God's willing that there be contingent concrete objects can explain the fact that there are positive contingent states of affairs. To arrive at this further point, Smith appeals to what he calls "the principle of explanatory equivalence." According to this principle if p explains q, and r is "relevantly equivalent" to q, p also explains r. To understand this principle it is important to distinguish "strict equivalence" from "relevance equivalence." The proposition An isosceles triangle has three angles, is strictly equivalent to All red things are red.

since any possible world in which the one is true is a world in which the other is true. If the principle of explanatory equivalence were stated in terms of strict equivalence it would clearly fall prey to obvious counterexamples. For an explanation of why an isosceles triangle has three angles need not be an explanation of why all red things are red. For two propositions to be relevantly equivalent, each must relevantly imply the other. And one proposition relevantly implies another if and only if it strictly implies it by virtue of its meaning. Smith then points out that the proposition 110

THE COSMOLOGICAL ARGUMENT Cosmological Arguments There are positive contingent truths, is relevantly equivalent to There are contingent concrete objects, and concludes that since God wills that there are contingent concrete objects. explains the fact that there are contingent concrete objects, it also explains the fact that there are positive contingent truths. The difficulty with Smith's argument is that the principle of explanatory equivalence is false. And, as argued in Rowe (1997), it is not just false when equivalence is understood as strict equivalence, it is also false when equivalence is understood as relevance equivalence. Consider, for example, the propositions John is angry at t. and

John exists at t and John is angry at t. These two propositions are relevantly equivalent in Smith's sense. But an explanation of the former need not be an explanation of the latter. It is one thing to explain why John is angry at t and another thing to explain why he exists at t. Given the attacks on PSR and the debate over the cogency of those attacks, it's clear that PSR is subject to significant objections, even if these objections themselves are not conclusive. In light of this, perhaps the best option for the proponent of the version of the cosmological argument set forth by Samuel Clarke is to employ a principle that is weaker than PSR, not subject to the objections to PSR that we've discussed, but still strong enough to support the premises of Clarke's argument. A principle like the following may satisfy these conditions. (A) For every kind of being such that beings of that kind can be caused to exist or can cause the existence of other beings, there must be a sufficient reason for the existence of each being of that kind and for the general fact that there exist beings of that kind. This principle is at least as initially plausible as PSR What distinguishes it from PSR is that it does not require that every fact, or even every contingent fact, has an explanation. Also, it does not imply that every positive contingent state of affairs has an explanation. But since it is a fact that there are dependent beings, principle (A) requires that there be a sufficient reason (full explanation) for the

111

345

346

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe fact that there are dependent beings. So principle (A), rather than the much stronger PSR, is all we need in order to justify the second premise of Clarke's cosmological argument: Not every being can be a dependent being. For as we've seen, if every being were dependent any proposed explanation of why there are dependent beings would be viciously circular. Thus, if every being (that can be caused or can cause other things to exist) were dependent, there would be a kind of being (dependent) such that the fact that there are beings of that kind would have no explanation. Also, so long as the first premise of Clarke's argument is restricted to beings that either can be caused or can cause the existence of other beings, principle (A) will justify the first premise: Every being is either a dependent being or a self-existent being. We should also note that principle (A), unlike PSR, does not raise problems for free acts of will. While there may be a determining cause of an individual being free (in the incompatibilist's sense) to will or not will, there can be no determining cause of the agent's freely causing one volition rather than another. It was this issue that made Clarke hesitant to fully endorse Leibniz's statement of the principle of sufficient reason. Principle (A) does not conflict with the existence of free acts of will. God's freely choosing to create Adam may constitute a determining cause of the existence of Adam. And this may be true even though there is no determining cause of God's freely choosing to create Adam. What we've seen is that there is a weaker principle than PSR that supports the two premises of Clarke's cosmological argument equally well but is not subject to the objections raised against PSR. Should we then declare victory for Clarke's argument and conclude that it does prove there exists a being whose existence is accounted for by its own nature (a self-existent being)? This would be premature unless we are in the position of knowing that principle (A) is true. And it is doubtful that we do know that principle (A) is true. Short of a proof of principle (A), perhaps the best that can be said is that when we reflect on our cognitive expectations of the world we discover ourselves operating with something like principle (A) as a presupposition of inquiry. And if we do conclude that this is so and we come to see that Clarke's premises are supported by principle (A), we would be inconsistent in rejecting the premises of Clarke's cosmological argument. Earlier we noted that in addition to the philosophical argument against the possibility that the world has always existed, we may appeal to a scientific argument to support premise (2) of the following argument. (1) Whatever begins to exist has a cause. (2) The world began to exist. (3) Therefore, the world has a cause of its existence. The scientific argument involves the Big Bang theory according to which the world began to exist some 10 to 20 billion years ago. The standard Big Bang theory holds that the present universe, if continually traced backwards in time, would continually diminish both spatially and temporally until finally reaching 112

THE COSMOLOGICAL ARGUMENT Cosmological Arguments - between 10 and 20 billion years ago - what is called a singularity^ which for practical purposes can be understood to be a point at which the universe came into existence in an explosion of such force that since then the universe has been continually expanding outwards. An important part of this scientific theory is that space and time also came to exist with the Big Bang, making it difficult to intelligibly ask "What, if anything, transpired before the Big Bang>", or "Just where in space did the Big Bang occur?" Nevertheless, since science tells us that the universe is finitely old, and since, according to our first premise the universe would then require a cause, it appears that we are only left to determine whether that cause would possess the properties of the theistic God. But before turning to that question, it should be noted that critical questions have been raised about the first premise of this argument. Since on the current scientific theory of the origin of the universe there is no time prior to the Big Bang, some philosophers have concluded that there cannot be a cause of the Big Bang since, on their view, a cause must be temporally prior to its effect. In addition, it is argued that since all we have evidence for concerning the coming into existence of physical things is evidence for things being produced out of earlier existing things^ we have no reason to think that the universe was caused to come into existence - for that would be something being caused to come into existence out of nothing at all. And even if there is no impossibility in the idea of something being produced out of absolutely nothing, we have no evidence at all that it ever occurs. To these objections defenders of the first premise of the argument have argued that simultaneous causation is possible, and have contended that the first premise of the argument (Whatever begins to exist has a cause) is not grounded on our experience of how things come to exist from pre-existing things. Instead, they hold that the first premise, while not true by virtue of the meaning of its terms, is, nevertheless, metaphysically necessary in the way in which the proposition expressed by "An effect never precedes its cause" is thought to be metaphysically necessary, even though it is not necessary simply by virtue of the meanings of its terms. So they may admit that in our experience things that are caused to come into existence are produced out of earlier existing things, but argue that the first premise of the argument (Whatever begins to exist has a cause) is metaphysically necessary, and therefore not dependent for confirmation on the evidence derived from our experience of how things come to exist in the universe. And there is another possible basis for support to which the defender of the first premise may appeal. For it does seem that many, if not all, of us are so constituted that we cannot help but believe, or at least expect, that if something comes into existence then there is a cause or explanation of its coming into existence. It is as though we are hard-wired to view our world in this way. Perhaps we are mistaken in having this belief or expectation. After all, the world doesn't have to be the way we are bound to expect it to be. But if we are so constituted as to expect or believe that things can't just pop into existence without a cause or explanation, then we would be inconsistent to also believe that there is no cause or explanation for the coming to be of the universe in the Big Bang. 113

347

348

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe Supposing that it is reasonable to infer a cause of the Big Bang, is there any good reason to identify that cause with the God of traditional theism, an infinitely powerful, all-knowing, eternal, perfecdy good being? Here the proponent of the cosmological argument is confronted with the rather difficult task of justifying the view that whatever caused the universe has the properties of the theistic God. And while it is clear that whatever caused the Big Bang would possess enormous power, it is quite difficult to see how a successful argument can be made for the cause of the universe being a perfecdy good being. There is some reason, perhaps, to view the eternal cause of the universe as a person. Craig, for example, argues as follows. Moreover, the personhood of the cause of the universe is implied by its timelessness and immateriality, since the only entities we know of which can possess such properties are either minds or abstract objects, and abstract objects do not stand in causal relations. Therefore, the transcendent cause of the origin of the universe must be of the order of mind. (Craig, 1999, pp. 734-5)

Since space, time, and matter come to exist in the Big Bang, Craig reasons that on the premise that the universe has a cause of its coming to exist, its cause must be both timeless and immaterial. But how do we get from the timelessness and immateriality of the cause of the universe to the conclusion that the cause of the universe must be a person? Given that a mind is a person, Craig proposes the following argument. (1) The cause of the universe is timeless and immaterial. (2) The only entities we know of which can be timeless and immaterial are minds or abstract objects. (3) Abstract objects cannot cause something to.come into existence. Therefore, (4) The cause of the universe is a mind (i.e., a person).

As it stands, there are two difficulties in Craig's argument. First, the argument is invalid. Given its premises, for the argument to be valid the conclusion has to be (4a) The only entity we know of which can be the cause of the universe is a mind. But (4a), due to the limitations of our knowledge, leaves it open that the cause of the universe is not a mind at all. For it may be both that among the entities we know of only a mind could cause the Big Bang and that, owing to the paucity of our knowledge, the cause of the universe is not a mind. And if we render the argument valid by replacing premise (2) with (2a) The only entities which can be timeless and immaterial are minds or abstract objects. 114

THE COSMOLOGICAL ARGUMENT Cosmological Arguments we face the formidable difficulty of establishing the truth of (2a). Why suppose that human knowledge has a grasp of all the possible kinds of timeless and immaterial entities there may be? Or, even supposing we do have such a grasp, why suppose we know them all well enough to know that among them only minds cause something to come into existence? So, when we render Craig's argument valid by replacing (4) with (4a), the conclusion doesn't give us what Craig wants to establish. And if we render his argument valid by replacing (2) with (2a), we will then have a valid argument for the conclusion Craig wants at the cost of introducing a premise that we don't know to be true. Even granting that the cause of the Big Bang is a mind, is it clear that it is a single mind rather than a multiplicity of minds who collaborated on the project of producing the Big Bang? Or, supposing it is shown that it is likely that the cause of the Big Bang is a single, immensely powerful mind, why suppose that this immensely powerful mind is God, an eternal being of such immense goodness, knowledge, and power that it is not possible for there to have been a better, wiser, or more powerful being than it? And what of the eternal existence of this mind who caused the universe to come into existence? Is it just an accident that it eternally exists, as opposed to not existing at all? If so, it would not satisfy one standard conception of the theistic God. But why isn't it just an accident? What philosophical principle justifies us in taking its eternal existence to be metaphysically necessary? And what proof can we give of that principle? Perhaps, however, God need not be taken to be a metaphysically necessary being. Perhaps by virtue of the simplicity of his nature, God provides a more plausible explanation for the temporally finite existence of the world than we would have if there were no God. If so, then the cosmological argument might justify belief in God without our having to suppose that God's existence is metaphysically necessary. But until we have convincing answers to these and other skeptical questions, the version of the cosmological argument based on the Big Bang theory will fall short of a successful argument for the existence of the God of traditional theism. Nevertheless, it must be acknowledged that the emergence of the Big Bang theory of the origin of the universe has given new weight to an argument for the existence of some sort of creator. References Clarke, Samuel, and Leibniz, Gottfried (1956) The Leibniz-Clarke Correspondence^ ed. H. G. Alexander. Manchester: Manchester University Press. Originally published in 1717. Craig, William Lane (1979) The Kalam Cosmological Argument. London: Macmilian. Craig, William Lane (1999) The Ultimate Question of Origins: God and the Beginning of the Universe. Astrophysics and Space Science^ 269-70, 723-40. Leibniz, Gottfried (1951) Monadology. In P. P. Wiener (ed.), Leibniz Selections. New York: Charles Scribner's Sons. Originally published in 1714. Rowe, William L. (1975) The Cosmological Argument. Princeton: Princeton University Press. Reprinted with a new foreword by Fordham University Press, 1998. 115

349

350

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION William L. Rowe Rowe, William L. (1997) Circular Explanations, Cosmological Arguments, and The Principle of Sufficient Reason. Midwest Studies in Philosophy, 21, 188-201. Smith, Quentin (1987) Infinity and the Past. Philosophy of Science, 54, 63-75. Smith, Quentin (1995) Explanatory Rationalism and Contingent Truths. Religious Studies, 31, 237-42. Vallicella, William (1997) On an Insufficient Argument Against Sufficient Reason. Ratio, 11,76-81.

Suggested Further Reading Craig, William Lane (1980) The Cosmological Argument from Plato to Lebiniz. London: Macmillan. Swinburne, Richard (1979) The Existence of God. Oxford: Oxford University Press. Van Inwagen, Peter (1993) Metaphysics. San Francisco: Westview Press. Chapter 6.

116

PART VI THE ONTOLOGICAL ARGUMENT

This page intentionally left blank

CHAPTER TWENTY-FOUR

The Ontological Argument

Arguments for the existence of God are commonly divided into a posteriori and a priori arguments. An a posteriori argument depends on a principle or premise that can be known only by means of our experience of the world. An a priori argument, on the other hand, purports to rest on principles which can be known independently of our experience of the world, just by reflecting on and understanding them. Of the three major arguments for the existence of God—the Cosmological, Teleological, and Ontological—only the last is entirely a priori. In the Cosmological argument one starts from some simple fact about the world, such as the fact that it contains things which are caused to exist by other things. In the Teleological argument a somewhat more complicated fact about the world serves as a starting point: the fact that the world exhibits order and design. In the Ontological argument, however, one begins simply with a concept of God. It is perhaps best to think of the Ontological argument as a family of arguments, each member of which begins with a concept of God, and by appealing only to a priori principles, endeavors to establish that God actually exists. Within this family of arguments the most important historically is the argument set forth by Anselm in the second chapter of his Proslogium (A Discourse).1 Indeed, the Ontological argument begins with chapter II of Anselm's Proslogium. In an earlier work, Monologium (A Soliloquy), Anselm had endeavored to establish the existence and nature of God by weaving together several versions of *Copyright © William L. Rowe, 1974. This essay was commissioned by the editor expressly for this volume. It has not been previously published.

the Cosmological argument. In the Preface to Proslogium Anselm remarks that after the publication of Monologium he began to search for a single argument which alone would establish the existence and nature of God. After much strenuous but unsuccessful effort, he reports that he sought to put the project out of his mind in order to turn to more fruitful tasks. The idea, however, continued to haunt him until one day the proof he had so strenuously sought became clear to his mind. Anselm sets forth this proof in the second chapter of Proslogium. Before discussing Anselm's argument in stepby-step fashion, there are certain concepts that will help us understand some of the central ideas of the argument. Suppose we draw a vertical line in our imagination and agree that on the left side of our line are all the things which exist, while on the right side of the line are all the things which don't exist. We might then begin to make a list of some of the things on both sides of our imaginary line, as follows: THINGS WHICH EXIST The Empire State Building Dogs The planet Mars

THINGS WHICH DON'T EXIST The Fountain of Youth Unicorns The Abominable Snowman

Now each of the things (or sorts of things) listed thus far has (have) the following feature: it (they) logically might have been on the other side of the line. The Fountain of Youth, for example, is on the right side of the line, but logically there is no absurdity in the idea that it might have been on the left side of the line. Similarly, although dogs do exist, we surely can imagine without logical absurdity that they might not have existed, that they might have been on the right side of the line.

354

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION WILLIAM L. ROWE

Let us then record this feature of the things thus far listed by introducing the idea of a contingent thing as a thing that logically might have been on the other side of the line from the side it actually is on. The planet Mars and the Abominable Snowman are contingent things, even though the former happens to exist and the latter does not. Suppose we add to our list the phrase "the object which is completely round and completely square at the same time" on the right side of our line. The round square, however, unlike the other things thus far listed on the right side of our line, is something that logically could not have been on the left side of the line. Noting this, let us introduce the idea of an impossible thing as a thing that is on the right side of the line and logically could not have been on the left side of the line. Looking again at our list, we wonder if there is anything on the left side of our imaginary line which, unlike the things thus far listed on the left side, logically could not have been on the right side of the line. At this point we don't have to answer this question, but it is useful to have a concept to apply to any such things, should there be any. Accordingly, let us say that a necessary thing is a thing on the left side of our imaginary line and logically could not have been on the right side of the line. Finally, a possible thing is any thing that is either on the left side of our imaginary line or logically might have been on the left side of the line. Possible things, then, will be all those things that are not impossible things—that is, all those things that are either contingent or necessary. If there are no necessary things, then all possible things will be contingent and all contingent things will be possible. If there is a necessary thing, however, then there will be a possible thing which is not contingent. Armed with these concepts, we can clarify certain important distinctions and ideas in Anselm's thought. The first of these is his distinction between existence in the understanding and existence in reality. Anselm's notion of existence in reality is the same as our notion of existence; that is, being on the left side of our imaginary line. Since the Fountain of Youth is on the right side of the line, it does not exist in reality. The things which exist are, to use Anselm's phrase, the things which exist in reality. Anselm's notion of

p

existence in the understanding, however, is not the same as any idea we normally employ. When we think of a certain thing, say the Fountain of Youth, then that thing, on Anselm's view, exists in the understanding. Also, when we think of an existing thing like the Empire State Building, it, too, exists in the understanding. So some of the things on both sides of our imaginary line exist in the understanding, but only those on the left side of our line exist in reality. Are there any things that don't exist in the understanding? Undoubtedly there are, for there are things, both existing and non-existing, of which we have not really thought. Now suppose I assert that the Fountain of Youth does not exist. Since to meaningfully deny the existence of something I have to have that thing in mind, I have to think of it, it follows on Anselm's view that whenever someone asserts that some thing does not exist, that thing does exist in the understanding.2 So in asserting that the Fountain of Youth does not exist, I imply that the Fountain of Youth does exist in the understanding. And in asserting that it does not exist I have asserted (on Anselm's view) that it does not exist in reality. This means that my simple assertion amounts to the somewhat more complex claim that the Fountain of Youth exists in the understanding but does not exist in reality— in short, that the Fountain of Youth exists only in the understanding. We can now understand why Anselm insists that anyone who hears of God, thinks about God, or even denies the existence of God is, nevertheless, committed to the view that God exists in the understanding. Also, we can understand why Anselm treats what he calls "the fool's claim" that God does not exist as the claim that God exists only in the understanding—that is, that God exists in the understanding but does not exist in reality. In Monologium Anselm sought to prove that among those beings which do exist there is one which is the greatest, highest, and the best. But in Proslogium he undertakes to prove that among those beings which exist there is one which is not just the greatest among existing beings, but is such that no conceivable being is greater. We need to distinguish these two ideas: (1) a being than which no existing being is greater, and (2) a being than which no conceivable being is greater.

THE ONTOLOGICAL ARGUMENT 10

355

REASON AND RELIGIOUS BELIEF

If the only things in existence were a stone, a frog, and a man, the last of these would satisfy our first idea but not our second—for we can conceive of a being (an angel or God) greater than a man. Anselm's idea of God, as he expresses it in Proslogium II, is the same as (2) above; it is the idea of "a being than which nothing greater can be conceived.'* It will facilitate our understanding of Anselm's argument if we make two slight changes in the way he has expressed his idea of God. For his phrase I shall substitute the following: 'the being than which none greater is possible. '3 This idea says that if a certain being is God then no possible being can be greater than it, or conversely, if a certain being is such that it is even possible for there to be a being greater than it, then that being is not God. What Anselm proposes to prove, then, is that the being than which none greater is possible exists in reality. If he proves this he will have proved that God, as he conceives of Him, exists in reality. But what does Anselm mean by "greatness"? Is a building, for example, greater than a man? In Monologium, chapter II, Anselm remarks: "But I do not mean physically great, as a material object is great, but that which, the greater it is, is the better or the more worthy—wisdom, for instance." Contrast wisdom with size. Anselm is saying that wisdom is something that contributes to the greatness of a thing. If a thing comes to have more wisdom than it did before then (given that its other characteristics remain the same), that thing has become a greater, better, more worthy thing than it was. Wisdom, Anselm is saying, is a great-making quality. However, the mere fact that something increases in size (physical greatness) does not make that thing a better thing than it was before, so size is not a greatmaking quality. By "greater than" Anselm means "better than," "superior to," or "more worthy than," and he believes that some characteristics, like wisdom and moral goodness, are great-making characteristics in that anything which has them is a better thing than it would be (other characteristics of it remaining the same) were it to lack them. We come now to what we may call the key idea in Anselm's Ontological argument. Anselm believes that existence in reality is a great-making quality. Does Anselm mean that anything that

exists is a greater thing than anything that doesn't? Although he does not ask or answer the question, it is perhaps reasonable to believe that Anselm did not mean this. When he discusses wisdom as a great-making quality he is careful not to say that any wise thing is better than any unwise thing—for he recognizes that a just but unwise man might be a better being than a wise but unjust man.4 I suggest that what Anselm means is that anything that doesn't exist but might have existed (is on the right side of our line but might have been on the left) would have been a greater thing if it had existed (if it had been on the left side of our line). He is not comparing two different things (one existing and one not existing) and saying that the first is therefore greater than the second. Rather, he is talking about one thing and pointing out that if it does not exist but might have existed, then it would have been a greater thing if it had existed. Using Anselm's distinction between existence in the understanding and existence in reality, we may express the key idea in Anselm's reasoning as follows: If something exists only in the understanding but might have existed in reality, then it might have been greater than it is. Since the Fountain of Youth, for example, exists only in the understanding but (unlike the round square) might have existed in reality, it follows by Anselm's principle that the Fountain of Youth might have been a greater thing than it is. II We can now consider the step-by-step development of Anselm's Ontological argument. I shall use the term "God" in place of the longer phrase "the being than which none greater is possible" —wherever the term "God" appears we are to think of it as simply an abbreviation of the longer phrase.

1. God exists in the understanding. As we have noted, anyone who hears of the being than which none greater is possible is, on Anselm's view, committed to premise (1). 2. God might have existed in reality (God is a possible being). Anselm, I think, assumes the truth of (2) without making it explicit in his reasoning. By asserting

356

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION WILLIAM L. ROWE

(2) I do not mean to imply that God does not exist in reality, but that, unlike the round square, God is a possible being. 3. If something exists only in the understanding and might have existed in reality, then it might have been greater than it is. As we noted, this is the key idea in Anselm's Ontological argument. It is intended as a general principle, true of anything whatever. Steps (l)-(3) constitute the basic premises of Anselm's Ontological argument. From these three items, Anselm believes, it follows that God exists in reality. But how does Anselm propose to convince us that if we accept (l)-(3) we are committed by the rules of logic to accept his conclusion that God exists in reality? Anselm's procedure is to offer what is called a reductio ad absurdum proof of his conclusion. Instead of showing directly that the existence of God follows from steps (l)-(3), Anselm invites us to suppose that God does not exist (i.e., that the conclusion he wants to establish is false) and then shows how this supposition, when conjoined with steps (1M3), leads to an absurd result, a result that couldn't possible be true because it is contradictory. Since the supposition that God does not exist leads to an absurdity, that supposition must be rejected in favor of the conclusion that God does exist. Does Anselm succeed in reducing the "fool's belief that God does not exist to an absurdity? The best way to answer this question is to follow the steps of his argument. 4. Suppose God exists only in the understanding. This supposition, as we saw earlier, is Anselm's way of expressing the belief that God does not exist. 5. God might have been greater than He is. (2, 4, and 3)5 Step (5) follows from steps (2), (4), and (3). Since (3), if true, is true of anything whatever, it will be true of God. Therefore, (3) implies that if God exists only in the understanding and might have existed in reality, then God might have been greater than He is. If so, then given (2) and (4), (5) must be true. For what (3) says when applied to God is that given (2) and (4), it follows that (5).

11

6. God is a being than which a greater is possible. (5) Surely if God is such that He logically might have been greater, then He is such than which a greater is possible. We can now appreciate Anselm's reductio argument. He has shown that if we accept steps (l)-(4), we must accept step (6). But (6) is unacceptable; it is the absurdity Anselm was after. By replacing "God" in (6) with the longer phrase it abbreviates, we see that (6) amounts to the absurd assertion: 7. The being than which none greater is possible is a being than which a greater is possible. Now since steps (l)-(4) have led us to an obviously false conclusion, and if we accept Anselm's basic premises (l)-(3) as true, then (4), the supposition that God exists only in the understanding, must be rejected as false. Thus we have shown that: 8. It is false that God exists only in the understanding. But since premise (1) tells us that God does exist in the understanding and (8) tells us that God does not exist only there, we may infer that 9. God exists in reality as well as in the understanding. (1, 8) HI

Most of the philosophers who have considered this argument have rejected it because of a basic conviction that from the logical analysis of a certain idea or concept we can never determine that there exists in reality anything answering to that idea or concept. We may examine and analyse, for example, the idea of an elephant or the idea of a unicorn, but it is only by our experience of the world that we can determine that there exist things answering to our first idea and not to the second. Anselm, however, believes that the concept of God is utterly unique—from an analysis of this concept he believes that it can be determined that there exists in reality a being which answers to it. Moreover, he presents us with an argument to show that it can be done in the case of the idea of God. We can, of course, simply reject Anselm's argument on the grounds that it

THE ONTOLOGICAL ARGUMENT 12

357

REASON AND RELIGIOUS BELIEF

violates the basic conviction noted above. Many critics, however, have sought to prove more directly that it is a bad argument and to point out the particular step that is mistaken. Next we shall examine the three major objections that have been advanced by the argument's critics. The first criticism was advanced by a contemporary of Anselm's, a monk named "Gaunilo," who wrote a response to Anselm entitled, "On Behalf of the Fool.*'6 Gaunilo sought to prove that Anselm's reasoning is mistaken by applying it to things other than God, things which we know don't exist. He took as his example the island than which none greater is possible. No such island really exists. But, argues Gaunilo, if Anselm's reasoning were correct we could show that such an island really does exist. For since it is greater to exist than not to exist, if the island than which none greater is possible doesn't exist then it is an island than which a greater is possible. But it is impossible for the island than which none greater is possible to be an island than which a greater is possible. Therefore, the island than which none greater is possible must exist. About this argument Gaunilo remarks: If a man should try to prove to me by such reasoning that this island truly exists, and that its existence should no longer be doubted, either I should believe that he was jesting, or I know not which I ought to regard as the greater fool: myself, supposing I should allow this proof; or him, if he should suppose that he had established with any certainty the existence of this island.7

Gaunilo's strategy is clear: by using the very same reasoning Anselm employs in his argument, we can prove the existence of things we know don't exist. Therefore, Anselm's reasoning in his proof of the existence of God must be mistaken. In reply to Gaunilo, Anselm insisted that his reasoning applies only to God and cannot be used to establish the existence of things other than God. Unfortunately, Anselm did not explain just why his reasoning cannot be applied to things like Gaunilo's island. In defense of Anselm against Gaunilo's objection, there are two difficulties in applying Anselm's reasoning to things like Gaunilo's island. The first derives from the fact that Anselm's principle that existence is a great-making quality was

taken to mean that if something does not exist then it is not as great a thing (being) as it would have been had it existed. Now if we use precisely this principle in Gaunilo's argument, all we will prove is that if Gaunilo's island does not exist then the island than which none greater is possible is an island than which a greater thing is possible. But this statement is not an absurdity. For the island than which no greater island is possible can be something than which a greater thing is possible—an unsurpassable island may be a surpassable thing. (A perfect man might be a greater thing than a perfect island.) Consequently, if we follow Anselm's reasoning exactly, it does not appear that we can derive an absurdity from the supposition that the island than which none greater is possible does not exist. A second difficulty in applying Anselm's reasoning to Gaunilo's island is that we must accept the premise that Gaunilo's island is a possible thing. But this seems to require us to believe that some finite, limited thing (an island) might have unlimited perfections. It is not at all clear that this is possible. Try to think, for example, of a hockey player than which none greater is possible. How fast would he have to skate? How many goals would he have to score in a game? How fast would he have to shoot the puck? Could he ever fall down, be checked, or receive a penalty? Although the phrase, "the hockey player than which none greater is possible," seems meaningful, as soon as we try to get a clear idea of what such a being would be like we discover that we can't form a coherent idea of it at all. For we are being invited to think of some limited, finite thing —a hockey player or an island—and then to think of it as exhibiting unlimited, infinite perfections. Perhaps, then, since Anselm's reasoning applies only to possible things, Anselm can reject its application to Gaunilo's island on the grounds that the island than which none greater is possible is, like the round square, an impossible thing. By far the most famous objection to the Ontological argument was set forth by Immanuel Kant in the eighteenth century. According to this objection the mistake in the argument is its claim, implicit in premise (3), that existence is a quality or predicate that adds to the greatness of a thing. There are two parts to this claim: (1) existence is a quality or predicate, and (2) existence, like wis-

358

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION WILLIAM L. ROWE

dom and unlike physical size, is a great-making quality or predicate. Someone might accept (1) but object to (2); the objection made famous by Kant, however, is directed at (1). According to this objection, existence is not a predicate at all. Therefore, since in its second premise Anselm's argument implies that existence is a predicate, the argument must be rejected. The central point in the philosophical doctrine that existence is not a predicate concerns what we do when we ascribe a certain quality or predicate to something: for example, when we say of a man next door that he is intelligent, six feet tall, or fat. In each case we seem to assert or presuppose that there exists a man next door and then go on to ascribe to him a certain predicate—"intelligent," "six feet tall," or "fat." And many proponents of the doctrine that existence is not a predicate claim that this is a generalfeature of predication. They hold that when we ascribe a quality or predicate to anything we assert or presuppose that the thing exists and then ascribe the predicate to it. Now if this is so, then it is clear that existence cannot be a predicate which we may ascribe to or deny of something. For if it were a predicate, then when we assert of some thing (things) that it (they) exists (exist) we would be asserting or presupposing that it (they) exists (exist) and then going on to predicate existence of it (them). For example, if existence were a predicate, then in asserting "tigers exist" we would be asserting or presupposing that tigers exist and then going on to predicate existence of them. Furthermore, in asserting "dragons do not exist" we would be asserting or presupposing, if existence were a predicate, that dragons do exist and then going on to deny that existence attaches to them. In short, if existence were a predicate, the affirmative existential statement "tigers exist" would be a redundancy and the negative existential statement "dragons do not exist" would be contradictory. But clearly "tigers exist" is not a redundancy; and "dragons do not exist" is true and, therefore, not contradictory. What this shows, according to the proponents of Kant's objection, is that existence is not a genuine predicate. According to the proponents of the above objection, when we assert that tigers exist and that dragons do not we are not saying that certain things (tigers) have and certain other things (dragons) do not have a peculiar predicate, exis-

13

fence; rather, we are saying something about the concept of a tiger and the concept of a dragon. We are saying that the concept of a tiger applies to something in the world and that the concept of a dragon does not apply to anything in the world. Although this objection to the Ontological argument has been widely accepted, it is doubtful that it provides us with a conclusive refutation of the argument. It may be true that existence is not a predicate, that in asserting the existence of something we are not ascribing a certain predicate or attribute to that thing. But the arguments presented for this view seem to rest on mistaken or incomplete claims about the nature of predication. For example, the argument which we stated earlier rests on the claim that when we ascribe a predicate to anything we assert or presuppose that that thing exists. But this claim appears to be mistaken. In asserting that Dr. Doolittle is an animal lover I seem to be ascribing the predicate "animal lover" to Dr. Doolittle, but in doing so I certainly am not asserting or presupposing that Dr. Doolittle actually exists. Dr. Doolittle doesn't exist, but it is nevertheless true that he is an animal lover. The plain fact is that we can talk about and ascribe predicates to many things which do not exist and never did. Merlin, for example, no less than Houdini, was a magician, although Houdini existed but Merlin did not. If, as these examples suggest, the claim that whenever we ascribe a predicate to something we assert or presuppose that the thing exists is a false claim, then we will need a better argument for the doctrine that existence is not a predicate. There is some question, however, whether anyone has succeeded in giving a really conclusive argument for this doctrine.8 A third objection against the Ontological argument calls into question the premise that God might have existed in reality (God is a possible being). As we saw, this premise claims that the being than which none greater is possible is not an impossible object. But is this true? Consider the series of positive integers: 1, 2, 3, 4, etc. We know that any integer in this series, no matter how large, is such that a larger integer than it is possible. Therefore, the positive integer than which none larger is possible is an impossible object. Perhaps this is also true of the being than which none greater is possible. That is, perhaps no matter how great a being may be, it is possible

THE ONTOLOGICAL ARGUMENT 14

359

REASON AND RELIGIOUS BELIEF

for there to be a being greater than it. If this were so, then, like the integer than which none larger is possible, Anselm's God would not be a possible object. The mere fact that there are degrees of greatness, however, does not entitle us to conclude that Anselm's God is like the integer than which none larger is possible. There are, for example, degrees of size in angles—one angle is larger than another—but it is not true that no matter how large an angle is it is possible for there to be an angle larger than it. It is logically impossible for an angle to exceed four right angles. The notion of an angle, unlike the notion of a positive integer, implies a degree of size beyond which it is impossible to go. Is Anselm's God like a largest integer, and therefore impossible, or like a largest angle, and therefore possible? Some philosophers have argued that Anselm's God is impossible,9 but the arguments for this conclusion are not very compelling. Perhaps, then, this objection is best construed not as proving that Anselm's God is impossible, but as raising the question whether any of us is in a position to know that the being than which none greater is possible is a possible object. For Anselm's argument cannot be a successful proof of the existence of God unless its premises are not just true but are really known to be true. Therefore, if we do not know that Anselm's God is a possible object, then his argument cannot prove the existence of God to us, cannot enable us to know that God exists.

IV Finally, I want to present a somewhat different critique of Anselm's argument, a critique suggested by the basic conviction noted earlier; namely that from the mere logical analysis of a certain idea or concept we can never determine that there exists in reality anything answering to that idea or concept. Suppose someone comes to us and says: I propose to define the term "God** as an existing, wholly perfect being. Now since it can't be true that an existing, wholly perfect being does not exist, it can't be true that God, as I've defined Him, does not exist. Therefore, God must exist.

His argument appears to be a very simple Ontological argument. It begins with a particular idea or concept of God and ends by concluding that God, so conceived, must exist. What can we say

in response? We might start by objecting to his definition, claiming: (1) that only predicates can be used to define a term, and (2) that existence is not a predicate. But suppose he is not impressed by this response—either because he thinks that no one has fully explained what a predicate is or proved that existence isn't one, or because he thinks that anyone can define a word in whatever way he pleases. Can we allow him to define the word "God" in any way he pleases and still hope to convince him that it will not follow from that definition that there actually exists something to which his concept of God applies? I think we can. Let us first invite him, however, to consider some concepts other than his peculiar concept of God. Earlier we noted that the term "magician" may be applied both to Houdini and Merlin, even though the former existed and the latter did not. Noting that our friend has used "existing" as part of his definition of "God," suppose we agree with him that we can define a word in any way we please, and, accordingly, introduce the following definitions: A "magican" is defined as an existing magician. A "magico" is defined as a non-existing magician. Here we have introduced two words and used "existing" or "non-existing" in their definitions. Now something of interest follows from the fact that "existing" is part of our definition of a "magican." For while it is true that Merlin was a magician, it is not true that Merlin was a magican. And something of interest follows from our including "non-existing" in the definition of a "magico"—it is true that Houdini was a magician, but it is not true that Houdini was a magico. Houdini was a magician and a magican, but not a magico\ Merlin was a magician u*id a magico, but not a magican. We have just seen that introducing "existing" or "non-existing" into the definition of a concept has a very important implication. If we introduce "existing" into the definition of a concept, it follows that no non-existing thing can exemplify that concept. And if we introduce "non-existing" into the definition of a concept, it follows that no existing thing can exemplify that concept. No non-existing thing can be a magican, and no existing thing can be a magico.

360

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION WILLIAM L. ROWE

But must some existing thing exemplify the concept "magican"? No! From the fact that "existing" is included in the definition of "magican" it does not follow that some existing thing is a magican—all that follows is that no non-existing thing is a magican. If there were no magicians in existence there would be nothing to which the term "magican" would apply. This being so, it clearly does not follow merely from our definition of "magican" that some existing thing is a magican. Only if magicians exist will it be true that some existing thing is a magican. We are now in a position to help our friend see that from the mere fact that "God" is defined as an existing, wholly perfect being it will not follow that some existing being is God. Something of interest does follow from his definition; namely that no non-existing being can be God. But whether some existing thing is God will depend entirely on whether some existing thing is a wholly perfect being. If no wholly perfect being exists there will be nothing to which his concept of God can apply. This being so, it clearly does not follow merely from his definition of "God" that some existing thing is God. Only if a wholly perfect being exists will it be true that God, as he conceives of Him, exists. The implications of these considerations for Anselm's ingenious argument can now be traced. Anselm conceives of God as a being than which none greater is possible. He then claims that existence is a great-making quality and something that has it is greater than it would have been had it lacked existence. Clearly then, no non-existing thing can exemplify Anselm's concept of God. For if we suppose that some non-existing thing exemplifies Anselm's concept of God and also suppose that that non-existing thing might have existed in reality (is a possible thing) then we are supposing that that non-existing thing (1) might have been a greater thing, and (2) is, nevertheless, a thing than which a greater is not possible. Thus far Anselm's reasoning is, I believe, impeccable. But what follows from it? All that follows from it is that no non-existing thing can be God (as Anselm conceives of God). All that follows is that given Anselm's concept of God, the proposition, "Some non-existing thing is God," cannot be true. But, as we saw earlier, this is also the case with the proposition, "Some non-existing thing is a magican." What remains to be shown is that

1J

some existing thing exemplifies Anselm's concept of God. What really does follow from his reasoning is that the only thing that logically could exemplify his concept of God is something which actually exists. And this conclusion is not without interest. But from the mere fact that nothing but an existing thing could exemplify Anselm's concept of God, it does not follow that some existing thing actually does exemplify his concept of God—no more than it follows from the mere fact that no non-existing thing can be a magican that some existing thing is a magican.10 There is, however, one major difficulty in this critique of Anselm's argument. This difficulty arises when we take into account Anselm's implicit claim that God is a possible thing. To see just what this difficulty is, let us return to the idea of a possible thing, which is any thing that either is on the left side of our imaginary line or logically might have been on the left side of the line. Possible things, then, will be all those things that, unlike the round square, are not impossible things. Suppose we concede to Anselm that God, as he conceives of Him, is a possible thing. Now, of course, the mere knowledge that something is a possible thing does not enable us to conclude that that thing is an existing thing. Many possible things, like the Fountain of Youth, do not exist. But if something is a possible thing then it is either an existing thing or a non-existing thing. The set of possible things can be exhaustively divided into those possible things which actually exist and those possible things which do not exist. Therefore, if Anselm's God is a possible thing it is either an existing thing or a non-existing thing. We have concluded, however, that no non-existing thing can be Anselm's God; therefore, it seems we must conclude with Anselm that some actually existing thing does exemplify his concept of God. To see the solution to this major difficulty we need to return to an earlier example. Let us consider again the idea of a "magican," an existing magician. It so happens that some magicians have existed—Houdini, the Great Blackstone, etc. But, of course, it might have been otherwise. Suppose, for the moment that no magicians have ever existed. The concept "magician" would still have application, for it would still be true that Merlin was a magician. But would any possible object be picked out by the concept of a "magi-

THE ONTOLOGICAL ARGUMENT l6

361

REASON AND RELIGIOUS BELIEF

can?" No, for no non-existing thing could exemplify the concept "magican." And on the supposition that no magicians ever existed, no existing thing would exemplify the concept "magican.'*11 We then would have a coherent concept "magican" which would not be exemplified by any possible object at all. For if all the possible objects which are magicians are non-existing things, none of them would be a magican and, since no possible objects which exist are magicians, none of them would be a magican. Put in this way, our result seems paradoxical. We are inclined to think that only contradictory concepts like "the round square" are not exemplified by any possible things. The truth is, however, that when "existing" is included in or implied by a certain concept, it may be the case that no possible object does in fact exemplify that concept. For no possible object that doesn't exist will exemplify a concept like "magican" in which "existing" is included; and if there are no existing things which exemplify the other features included in the concept—for example, "being a magician" in the case of the concept "magican"—then no possible object that exists will exemplify the concept. Put in its simplest terms, if we ask whether any possible thing is a magican the answer will depend entirely on whether any existing thing is a magician. If no existing things are magicians then no possible things are magicans. Some possible object is a magican just in the case some actually existing thing is a magician. Applying these considerations to Anselm's argument, we can find the solution to our major difficulty. Given Anselm's concept of God and his principle that existence is a great-making quality, it really does follow that the only thing that logically could exemplify his concept of God is something which actually exists. But, we argued, it doesn't follow from these considerations alone that God actually exists, that some existing thing exemplifies Anselm's concept of God. The difficulty we fell into, however, is that when we add the premise that God is a possible thing, that some possible object exemplifies his concept of God, it really does follow that God actually exists, that some actually existing thing exemplifies Anselm s concept of God. For if some possible object exemplifies his concept of God, that object is either an existing thing or a non-existing thing. But since no non-existing thing could exemplify

Anselm's concept of God, it follows that the possible object which exemplifies his concept of God must be a possible object that actually exists. Therefore, given (1) Anselm's concept of God, (2) his principle that existence is a great-making quality, and (3) the premise that God, as conceived by Anselm, is a possible thing, it really does follow that Anselm's God actually exists. But we now can see that in granting Anselm the premise that God is a possible thing we have granted far more than we intended. All we thought we were conceding is that Anselm's concept of God, unlike the concept of a round square, is not contradictory or incoherent. But without realizing it we were in fact granting much more than this, as became apparent when we considered the idea of a magican. There is nothing contradictory in the idea of a magican, an existing magician. But in asserting that a magican is a possible thing we are, as we saw, directly implying that some existing thing is a magician. For if no existing thing is a magician, the concept of a magican will apply to no possible object whatever. The same point holds with respect to Anselm's God. Since Anselm's concept of God logically cannot apply to some non-existing thing, the only possible objects to which it could apply are possible objects which actually exist. Therefore, in granting that Anselm's God is a possible thing we are conceding far more than that his idea of God isn't incoherent or contradictory. Suppose, for example, that every existing being has some defect which it might not have had. Without realizing it we were denying this when we granted that Anselm's God is a possible being. If every existing being has a defect it might not have had, then every existing being might have been greater. But if every existing being might have been greater, then Anselm's concept of God will apply to no possible object whatever. Therefore, if we allow Anselm his concept of God and his principle that existence is a greatmaking quality, then in granting that God, as Anselm conceives of Him, is a possible being we will be granting much more than that his concept of God is not contradictory. We will be conceding, for example, that some existing thing is as perfect as it can be. The fact is that Anselm's God is a possible thing only if some existing thing is as perfect as it can be.

362

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION WILLIAM L. ROWE

Our final critique of Anselm's argument is simply this. In granting that Anselm's God is a possible thing we are in fact granting that Anselm's God actually exists. But since the purpose of the argument is to prove to us that Anselm's God exists, we cannot be asked to grant as a premise a statement which is virtually equivalent to the conclusion that is to be proved. Anselm's concept of God may be coherent and his principle that existence is a great-making quality may be true. But all that follows from this is that no nonexisting thing can be Anselm's God. If we add to all of this the premise that God is a possible thing it will follow that God actually exists. But the additional premise claims more than that Anselm's concept of God isn't incoherent or contradictory. It amounts to the assertion that some existing being is supremely great. And since this is, in part, the point the argument endeavors to prove, the argument begs the question: it assumes the point it is supposed to prove. If the above critique is correct, Anselm's argument fails as a proof of the existence of God. This is not to say, however, that the argument is not a work of genius. Perhaps no other argument in the history of thought has raised so many basic philosophical questions and stimulated so much hard thought. Even if it fails as a proof of the existence of God, it will remain as one of the high achievements of the human intellect. NOTES 1. Some philosophers believe that Anselm sets forth a different and more cogent argument in chapter III of his Proslogium. For this viewpoint see Charles Hartshorne, Anselm's Discovery (LaSalle, 111.: Open Court Publishing Co.,

17

1965); and Norman Malcolm, "Anselm's Ontological Arguments," The Philosophical Review LXIX, No. 1 (January 1960): 41-62. For an illuminating account both of Anselm's intentions in Proslogium II and III and of recent interpretations of Anselm see Arthur C. McGill's essay "Recent Discussions of Anselm's Argument" in The Many-faced Argument, ed. John Hick and Arthur C. McGill (New York: The Macmillan Co., 1967), pp. 33-110. 2. Anselm does allow that someone may assert the sentence "God does not exist" without having in his understanding the object or idea for which the word 'God' stands (see Proslogium, chapter IV). But when a person does understand the object for which a word stands, then when he uses that word in a sentence denying the existence of that object he must have that object in his understanding. It is doubtful, however, that Anselm thought that incoherent or contradictory expressions like 'the round square' stand for objects which may exist in the understanding. 3. Anselm speaks of "a being" rather than "the being" than which none greater can be conceived. His argument is easier to present if we express his idea of God in terms of "the being." Secondly, to avoid the psychological connotations of "can be conceived" I have substituted "possible." 4. See Monologium, chapter XV. 5. The numbers in parentheses refer to the earlier steps in the argument from which the present step is derived. 6. Gaunilo's brief essay, Anselm's reply, and several of Anselm's major works, as translated by S. N. Deane, are collected in Saint Anselm: Basic Writings (LaSalle, 111.: Open Court Publishing Co., 1962). 7. Saint Anselm: Basic Writings, p. 151. 8. Perhaps the most sophisticated presentation of the objection that existence is not a predicate is William P. Alston's "The Ontological Argument Revisited" in The Philosophical Review 69 (1960): 452-74. 9. See, for example, C. D. Broad's discussion of the Ontological Argument in Religion, Philosophy, and Psychical Research (New York: Harcourt, Brace & World, 1953). 10. An argument along the lines just presented may be found in J. Shaffer's illuminating essay "Existence, Predication and the Ontological Argument," Mind 71 (1962): 307325. 11. I am indebted to Professor William Wainwright for bringing this point to my attention.

CHAPTER TWENTY-FIVE

Modal Versions of the Ontological Argument

William Rowe is professor of philosophy at Purdue University. In this essay he critically analyzes Plantinga's version of the ontological argument, appreciating its brilliance but leveling some objections at what it claims to have accomplished.

It has sometimes been thought that two distinct ontological arguments can be found in chapters 2 and 3 of Anselm's Proslogium. It is clear that in chapter 2 Anselm intended to set forth an argument for God's existence. He there introduces his concept of God as a being than which none greater is possible, and he advances the principle that existence in reality contributes to the greatness of a being. He then argues that God, as conceived by him, exists in reality—for otherwise a being greater than the greatest possible being would be possible.1 Having satisfied himself that God's existence has been established, in chapter 3 Ansel m turns to consider the mode or way in which God exists. Some things, like cabbages and kings, exist only contingently. It is possible that they should not have existed at all. Put in the language of possible worlds, we might say that the possible world that happens to be actual contains cabbages and kings.2 Other possible worlds, however, do no contain them; and had one of those worlds been the actual world, cabbages and kings would not have existed. Does God exist only contingently? Anselm thought not, for a being would be greater if it ex-

This article was written specifically for this volume and appears here for the first time.

isted in such a way that it logically could not fail to exist. Put in the language of possible worlds, a being would be greater if it is contained in every possible world rather than in just some possible worlds. So if God exists contingently, it would be possible for God to be greater than he is. Since it is not possible for God to be greater than he is, God must exist necessarily. As I have interpreted Anselm, he did not intend in chapter 3 to be offering a further argument for God's existence. Instead, he wanted to determine whether God, whose existence he had already established in chapter 2 exists contingently or necessarily. But whatever his intentions may have been, it is not difficult to see in chapter 3 the makings of a distinct argument for God's existence. For chapter 3 presents us with the principle that necessary existence, no less than the existence in reality of chapter 2, contributes to the greatness of a being. If it is possible for Anselm's God to possess necessary existence, then that is the sort of existence he does possess—otherwise it would be possible for God to be greater. Reflection on chapter 3 of Anselm's Proslogium has led philosophers to create various modal versions of the ontological argument. Among the most interesting versions is one set forth by Alvin Plantinga. Plantinga's version has the merit of extraordinary simplicity. By defining the concept of maximal greatness in a certain manner, Plantinga is able to boil down his version of the argument to the assertion of a single premise: that there is a possible world in which the property of maximal greatness is instantiated. Another merit of Plantinga's version is that it makes use of the idea of possible worlds, thus reducing the logic of the mo-

364 70

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION Traditional Arguments for the Existence of God

dal argument to its most intuitive level. Before we consider his version, however, let's prepare ourselves for some of the questions we need to raise by examining two quite simple ontological arguments that are suggested by the reasoning in chapters 2 and 3 of AnselnVs Proslogium. Consider two distinct concepts of God that I will call d and G2. We shall define d as follows: G! = the concept of an omnipotent, omniscient, wholly good being who is such that he exists with these perfections in the actual world. G2 is defined as follows: G2 = the concept of an omnipotent, omniscient, wholly good being who is such that he exists with these perfections in every possible world. Let us say that a normal concept C of a being or kind of being is satisfied in a given possible world just in case, were that world actual, that being or a being of that kind would exist. Thus the concept elephant is satisfied in our world, but the concept unicorn is not. For our world is actual and elephants do exist, but unicorns do not. In some other possible world, however, just the reverse is true—the concept unicorn is satisfied, but the concept elephant is not. For if that world were actual, at least one unicorn would exist, but no elephants would exist. Armed with this idea of what it is for a normal concept of a being (or kind of being) to be satisfied in a possible world, let's consider our two concepts of God introduced above. With a little reflection, I think we can see that our first concept, Gj, may not be a normal concept. To ask whether the normal concept unicorn is satisfied in w is simply to ask whether a unicorn would exist if w were actual. Where w is some possible world other than the possible world that is in fact actual, the question of whether the concept unicorn is satisfied in w has nothing to do with whether unicorns, elephants, cabbages, or kings exist in the actual world. But this is not so with d. Whether GI is satisfied in w depends in part on what sorts of beings actually exist, what

sorts of beings exist in the actual world. It is not enough that w contains an omnipotent, omniscient, wholly good being. For unless that being exists in the actual world with just those perfections, GI is not satisfied in w. The important point to grasp here is that the satisfaction of G! in any possible world depends on the actual existence of an omnipotent, omniscient, wholly good being.3 The following argument is suggested by the reasoning of Proslogium 2: (1) There is a possible world in which Ci is satisfied. Therefore, (2) There exists an omnipotent, wholly good being.

omniscient,

This argument is logically valid. Is its premise true? Well, that depends, as we've seen, on what beings are contained in the actual world. If every existing being has some moral defect, then there is no possible world in which GT is in fact satisfied. For an argument to be a proof of its conclusion we must know its premise(s) to be true without basing that knowledge on a prior knowledge of its conclusion. Is it logically possible for some human being to know (1) to be true without basing that knowledge on knowing (2) to be true? It would be rash to answer no to this question. For it is difficult to draw logical limits to the ways in which human beings might come to know that a certain proposition is true. But perhaps we can say this much. It is exceedingly difficult to see how some human being would in fact come to a knowledge of (1) independently of knowing (2). So it is more than likely true that this argument is not a proof of its conclusion for any human being. G2 is a more far-reaching concept than is GI. For GI'S being satisfied in a possible world w requires that the actual world contain an omnipotent, omniscient, wholly good being but allows that many other possible worlds lack such a being. G2, however, is satisfied in a possible world w only if every possible world (including the actual world) contains an omnipotent, omniscient, wholly good being. In the spirit of Proslogium 3, if not the letter, we can construct the following argument.

THE ONTOLOGICAL ARGUMENT J.C.4 Modal Versions of the Ontological Argument

(3) There is a possible world in which G2 is satisfied. Therefore, (4) There necessarily exists an omnipotent, omniscient, wholly good being. Once we realize that the satisfaction of G2 in any possible world requires the existence of an omnipotent, omniscient, wholly good being in every possible world, we can appreciate the extraordinary difficulty of viewing this argument as a proof of its conclusion. Perhaps if we know that the actual world contains an omnipotent, omniscient, wholly good being, we might begin to ponder whether this being holds forth in all or just some possible worlds. But it is difficult to see how merely reflecting on the concept G2 can enable us to know that it is satisfied in some possible world. For, as we've noted, its satisfaction in any possible world depends on what is contained in every possible world. But again, it would be unwise to declare that it is logically impossible for someone to come to know (3) independently of knowing (4), or even (2). But few, I believe, would be inclined to view this argument as a proof of its conclusion. For reasons we need not consider here, Plantinga prefers to state his modal version of the ontological argument in terms of whether a certain property—the property of being maximally great—is instantiated in any possible world. For a property to be instantiated in a world w is for it to be true that if w were actual some thing would exist having that property. Thus the property of being an elephant is instantiated in our possible world, but the property of being a unicorn is not. Plantinga's property of being maximally great, however, is vastly different from such pedestrian properties as being an elephant or being a unicorn. The question of whether these two properties are instantiated in some possible but nonactual world w doesn't at all depend on whether the actual world contains elephants or unicorns. But Plantinga's property of being maximally great can be instantiated in some possible world w only if the actual world contains a being that is omnipotent, omniscient, and morally perfect. And even

365 71

this is not enough. Not only must the actual world contain an omnipotent, omniscient, and morally perfect being, but every possible world must contain a being having these marvelous attributes, and it must be the same being who has these attributes in all these different worlds. Once we understand all this, we can see what an extraordinary property it is to which Plantinga has drawn our attention. If any possible world whatever happens to lack an omnipotent, omniscient, and morally perfect being, then Plantinga's extraordinary property is an impossible property and is instantiated in no possible world. Analogous to our argument concerning concept G2/ the following argument is valid: (5) There is a possible world in which maximal greatness is instantiated. Therefore, (6) There necessarily exists an omnipotent, omniscient, and morally perfect being. And, for reasons given in connection with our two earlier arguments, it is extremely unlikely that this argument is a proof of its conclusion. Consider the property of being in less-than perfect company, where it is understood that a person has that property in a world w just in case every person in w (human and nonhuman) has some degree of imperfection, however slight. It may be that we enjoy (or are burdened with) this property in the actual world. But even if we are not, surely, one would think, it is possible that this property be instantiated. Surely there is some possible world in which every person has some imperfection, however slight. But if so, then Plantinga's extraordinary property is impossible; there is no possible world in which it is instantiated. If either of these properties is instantiated in some world w, then the other is uninstantiated in w and in every other possible world. Since only one can be instantiated, which, if either, might it be? The instantiation of Plantinga's extraordinary property in a pos-' sible world w is dependent on what every other possible world contains—every possible world must contain an omnipotent, omniscient, and morally perfect being. The instantiation in w of the

366 72

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION Traditional Arguments for the Existence of God

property of being in less than perfect company requires only that each person in w have some flaw, however slight. If you know nothing else relevant to your decision and had to bet on which property is possibly instantiated, knowing that both cannot be, which would you bet on? Although Plantinga accepts the version of the ontological argument that he sets forth, he acknowledges that it is not a proof of its conclusion. It does not, he notes, establish the truth of theism. What then does the argument do? It establishes, Plantinga claims, the rational acceptability of theism. It does this, Plantinga argues, because the premise of the argument, proposition (5), is something that can be believed without violating any rule of reason concerning what we may or may not believe. Since we do no wrong in accepting (5), and since we acknowledge that (5) entails the truth of theism, we do no wrong in accepting the truth of theism. If it is not wrong for me to believe a proposition, then that proposition is rationally acceptable for me. Perhaps the first point to note about Plantinga's claim is that in his view the premise of an argument may be rationally acceptable and may thus establish the rational acceptability of the argument's conclusion, even though one doesn't know the premise to be true, and even though the truth of the premise is a matter of significant controversy. After all, some who reflect on the amount of tragic evil in our world are committed to the view that the property of being in less than perfect company is instantiated in our world. Others, including a number who believe that there exists an omnipotent, omniscient, and morally perfect being, would insist that there is some possible world in which the property of being in less than perfect company is instantiated. Both groups, therefore, are committed to the denial of Plantinga's premise that there is some possible world in which maximal greatness is instantiated. Still others may hold that there is simply no way of telling whether maximal greatness is possibly instantiated. So the premise of Plantinga's argument is denied by many and held in question by others. Moreover, Plantinga offers no argument for his premise and acknowledges that reflecting on it

does not enable us to somehow see that it must be true; he does not claim that after sufficient reflection the inquiring mind somehow comes to find his premise self-evident. What, then, does Plantinga claim for his premise? He claims, as we've seen, that it is not irrational to accept it, that in accepting it one does not violate any rules concerning what we may or may not believe. Of course, if it were a rule that one must not accept a premise unless one can prove it or has some good evidence for it, Plantinga would be unjustified in accepting his premise. But the "rule" just mentioned is difficult to defend. Perhaps what Plantinga holds is this: There are circumstances in which it is permissible to believe a proposition even though you cannot prove it and don't have good evidence for it. What are these circumstances? Well, one circumstance, surely, is that you have no good reason to think the proposition false. (Some think that the idea of a maximally great being is like the idea of a largest integer—an impossible object. But this may well be wrong. It might be that we have no good reason to think Plantinga's premise false.) The other circumstances that must obtain are difficult to specify. But if we agree with Plantinga about this, then I think we can say that it may be permissible for someone to believe Plantinga's premise. Plantinga says something much stronger. He says it is "evident" that believing his premise is permissible. This claim, I believe, is excessive. We need to be much clearer about the circumstances that must obtain for Plantinga's premise to be acceptable before we declare its acceptability with the unabashed assurance Plantinga here expresses. Some philosophers declare that the ontological argument in all its versions commits some gross fallacy or contains some obviously false premise. Plantinga's careful work on the argument helps us to see that we can confidently reject such criticisms. But when the argument is set forth with care and rigor, we can see, I believe, how very difficult it is to know or establish the truth of its premise(s). (Indeed, in some versions one has great difficulty in even imagining how one might know the premise(s) without basing such knowledge on a prior knowledge of the conclusion.) I

THE ONTOLOGICAL ARGUMENT I.C.5 An Ontological Argument for the Devil think Plantinga sees this as well. Anselm's high hope of discovering an argument that would conclusively establish God's existence remains unfulfilled, even in Plantinga's skillful hands. As a consolation prize, Plantinga proposes a weak sense of rational acceptability that he claims is satisfied by the premise of his modal version of the argument. Anselm thought that if we really understood the argument it would be obvious that it is a sound demonstration of the existence of God. To reject the argument, therefore, is to be foolish. Plantinga makes no such claim. He holds only that it is clear that one is not foolish to accept it. If I am right, all that has been shown is that it may not be foolish to accept it. To establish that it is not foolish requires that we become clear that its premise satisfies all the circumstances (whatever they are) that are required for it to be permissible to believe a proposition even though we cannot prove it and don't have good evidence for it.

367 73

Notes 1. This argument has fascinated philosophers and theologians for centuries. For an exposition of the argument and the major objections to it, see my essay, "The Ontological Argument/' in my Philosophy of Religion (Belmont, Calif.: Wadsworth, 1978). 2. The idea of possible worlds is explained briefly and clearly in Alvin Plantinga's Gocf, Freedom and Evil (New York: Harper & Row, 1974), 34-39. 3. GI is an abnormal concept if there is a possible world in which no perfect being exists. For in that case, G/s being satisfied in w depends in part upon which possible world is actual. Suppose possible world w* contains an omnipotent, omniscient, wholly good being, but possible world w** does not. If vv* is the actual world, then depending on what w contains, GI may be satisfied in w. But if w** is the actual world, then no matter what w contains, G\ is not satisfied in w. G2, although a more far-reaching concept than d, is, however, a normal concept. Although its being satisfied in w depends upon what is contained in every other possible world, its being satisfied in wdoes not depend on which possible world is actual.

This page intentionally left blank

PART VII OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

This page intentionally left blank

CHAPTER TWENTY-SIX

Augustine on Foreknowledge and Free Will

JL HE PROBLEM with which this paper is concerned is raised in the following passage in Book III of St. Augustine's treatise On Free Witt: ii, 4. (Evodius) . . . I have a deep desire to know how it can be that God knows all things beforehand and that, nevertheless, we do not sin by necessity. . . . Since God knew that man would sin, that which God foreknew must necessarily come to pass. How then is the will free when there is apparently this unavoidable necessity? iii. 6 (Augustine) Your trouble is this. You wonder how it can be that these two propositions are not contradictory and incompatible, namely that God has foreknowledge of all future events, and that we sin voluntarily and not by necessity. For if, you say, God foreknew that a man will sin, he must necessarily sin. But if there is necessity there is no voluntary choice in sinning, but rather fixed and unavoidable necessity. You are afraid that by that reasoning the conclusion may be reached either that God's foreknowledge of all future events must be impiously denied, or, if that cannot be denied, that sin is committed not voluntarily but by necessity. Isn't that your difficulty? (Evodius) Exactly that.1 The problem, as Augustine sees it, is to show how it is possible both that we voluntary (freely) will to perform certain actions and that God foreknows that we shall will to perform these actions. The argument which gives rise to this problem may be expressed as follows: (1) God has foreknowledge of all future events. (2) Hence, if a man is going to sin, God foreknows that he will sin. (3) Whatever God foreknows must necessarily happen. (4) Hence, if God foreknows that a man will sin, he must necessarily sin. 1

On Free Will, trans, by John H. S. Burleigh, in Augustine: Earlier Writings, The Library of Christian Classics, Vol. VI (Philadelphia, 1953), pp. 172-173.

372

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION AUGUSTINE ON FOREKNOWLEDGE AND FREE WILL

357

(5) But if such a man must necessarily sin, there is no voluntary choice in his sinning. (6) Therefore, such a man does not have free will. I shall begin by examining Augustine's solution to this problem. Augustine, if I understand him correctly, proposes to solve the problem by denying premiss (5). That is, he denies that if a man must necessarily sin, there is no voluntary choice in his sinning. Suppose that a man is going to sin and, hence, is going to will or choose to sin.2 Given that God foreknows that a man is going to will to sin, it appears to follow from premiss (3) that it is necessary that the man is going to will or choose to sin. But if a man must necessarily will or choose to sin, how can we claim that he, nevertheless, voluntary or freely wills to sin? It is this question that Augustine sets himself to answer. Augustine's answer to the question just raised is that even though a man necessarily wills to sin he, nevertheless, freely wills to sin. The reason this is so is that the will is something that is always in our power, and whatever is in our power is free. The important passages in which he explains and argues this point are the following: But if he (a denier of free will) . . . says that, because he must necessarily so will, his will is not in his own power, he can be countered by the answer you gave me when I asked whether you could become happy against your will. You replied that you would be happy now if the matter were in your power; for you willed to be happy but could not achieve it ... we cannot say we do not have the power unless we do not have what we will. If we do not have the will, we may think we will but in fact we do not. If we cannot will without willing those who will have will, and all that is in our power we have by willing. Our will would not be will unless it were in our power. Because it is in our power, it is free. We have nothing that is free which is not in our power, and if we have something it cannot be nothing. Hence it is not necessary to deny that God has fore* knowledge of all things, while at the same time our wills are our own, God has foreknowledge of our will, so that of which he has foreknowledge must come to pass. In other words, we shall exercise our

2

On Augustine's theory the essential element in every sinful act is "a movement of the will away from unchangeable good to mutable good." (Ibid., p. 170.)

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

358

WILLIAM L ROWE wills in the future because he has foreknowledge that we shall do so; and there can be no will or voluntary action unless it be in our power.8 . . . there is nothing so much in our power as is the will itself. For as soon as we will (volumus) immediately will (voluntas) is there. We can say rightly that we do not grow old voluntarily but necessarily, or that we do not die voluntarily but from necessity, and so with other similar things. But who but a raving fool would say that it is not voluntarily that we will? Therefore though God knows how we are going to will in the future, it is not proved that we do not voluntarily will anything.4 For if that is to be called our necessity which is not in our power, but even though we be unwilling effects what it can effect—as, for instance, the necessity of death—it is manifest that our wills by which we live uprightly or wickedly are not under such a necessity; for we do many things which, if we were not willing, we should certainly not do. This is primarirly true of the act of willing itself—for if we will, it is; if we will not, it is not—for we should not will if we were unwilling. But if we define necessity to be that according to which we say that it is necessary that anything be of such a nature, or be done in such and such a manner, I know not why we should have any dread of that necessity taking away the freedom of our will.5

Augustine's proposed solution, I believe, proceeds as follows. Granted that a man necessarily wills to sin, it does not follow that his will is not in his power, that he does not freely will to sin. To see that this is so we must first see what it means to say of something that it is not in a man's power. Augustine's view is that to say that x is npt in a man's power is to say (roughly) that the presence or absence of x is not a result of the man's will. That is, x is not in a man's power if and only if either (1) x fails to occur even though the man wills to do x—for example, running a mile in four minutes is not in my power because the feat fails to be accomplished even though I will to accomplish it—or (2) x occurs even though the man does not will to do x—for example, my growing old is not in my power because it occurs even though I do not will it to occur. Thus where x is something that occurs, the question whether x is in our power reduces to the question whether x would occur even though we do not will it. This much 9

Ibid., pp. 175-176. /bid., pp. 174-175. Augustine, The City of God, Bk. V. Ch. X, trans, by J. J. Smith, in Basic Writings of St. Augustine, Vol. II (New York, 1948), p. 68. 4 5

373

374

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION AUGUSTINE ON FOREKNOWLEDGE AND FREE WILL

359

granted, Augustine argues as follows. Even though a man necessarily wills to sin, we cannot say that this sinful act of will is not in the man's power. For clearly the act of will would not occur if the man did not will. To say that the man's willing to sin is not in his power is to say that the man wills to sin even though he does not will to sin—and this is impossible. Thus Augustine seems to hold that it is a necessary truth that the will is in our power and, therefore, free. As he puts it, "Our will would not be will unless it were in our power. Because it is in our power it is free." The fact that a man necessarily wills to sin does not conflict with his freely willing to sin because his willing to sin (although necessary) is still in his power—for it would not occur were he not to will to sin. In this way Augustine is led to reject premiss (5) in the argument under consideration. That is, he rejects the claim that if a man must necessarily sin, there is no voluntary choice in his sinning. There is, I believe, a mistake in Augustine's reasoning in support of the claim that even though a sinful act of will is necessary it is, nevertheless, in one's power. His analysis of "x is not in our power" is, in part, "x occurs even though we do not will x." Thus my growing old is not in my power because I grow old even though I do not will to grow old. He argues, as we saw, that to say my willing to sin is not in my power is to say that I will to sin even though I do not will to sin—which, of course, is impossible. But surely there is a mistake here. If the case of my willing to sin is to parallel the case of my growing old then to say that my willing to sin is not in my power is not to say that I will to sin even though I do not will to sin; rather, it is to say that I will to sin even though I do not will to will to sin. The point I am making can be expressed just as well by taking as an example an act of will that does not occur—say, the act of willing to refrain from sinning. This act of will which does not occur corresponds to my act of running a mile in four minutes, which does not occur either. Now on Augustine's analysis of "x is not in our power" it follows that x is not in my power if I fail to do x even though I will to do x. Thus it follows that running a mile in four minutes is not in my power if I fail to run such a mile even though I will to run a mile in four minutes. Consider now the case of my

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

360

WILLIAM L ROWE

willing to refrain from sinning. Parallel to the case of running a four-minute mile, the proper analysis of "willing to refrain from sinning is not in my power" should be, on Augustine's view, the following: (a) I fail to will to refrain from sinning even though I will to will to refrain from sinning. rather than, (b) I fail to will to refrain from sinning even though I will to refrain from sinning. (a) and (b) are not the same, (b) seems to express an impossibility—hence Augustine's conviction that the will would not be will unless it were in our power. About (a) I wish to make two comments. First it is not clear that the phrase "I will to will to refrain from sinning" makes any sense at all. G. E. Moore suggested that there is such a thing as " . . . making an effort to induce ourselves to choose a particular course."6 If there is such a thing as Moore suggested and if we take the phrase in question as expressing that thing, then perhaps we can interpret (a) in such a way that it expresses an intelligible proposition—namely, that I fail to will to refrain from sinning even though I make an effort to will or choose to refrain from sinning. The second comment is that if we so interpret (a) then it seems clear that (a) may be true and, consequently, we cannot conclude, in the simple way Augustine does, that it is in a man's power to will to refrain from sinning, that the will is something that is always in our power. Therefore, I conclude that Augustine has not made good his claim that, even though a man necessarily wills to sin, his willing or choosing to sin is voluntary or free. If this is so, it follows that Augustine has not succeeded in showing that premiss (5) is not true, and, therefore, has not succeeded in solving Evodius' problem of how God's foreknowledge is compatible with the freedom of the will. Returning to the argument that Augustine wishes to refute, it would seem that there is a logical mistake either in the drawing 6

Ethics (London, 1912), pp. 135-136.

375

376

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION AUGUSTINE ON FOREKNOWLEDGE AND FREE WILL

361

of (4) from (3) or in the interpretation of (3). Let's look at (3) for a moment. Let "p" pick out some event. We must distinguish between interpreting (3) as (3a) "It is necessary that if God foreknows p, p will happen"—here the necessity applies to the conditional "If God foreknows p, p will happen"; and (3b) "If God foreknows p, p will happen necessarily"—here the necessity applies not to the conditional but to its consequent "p will happen." Since (4) is drawn from (3), and since (5) presupposes that (4) is an instance of (3b) rather than of (3a)—otherwise we have no reason for supposing in the antecedent of (5) that the man in question must necessarily sin—it is clear that the inference from (3) to (4) is valid only if (3) is interpreted as (3b) rather than (3a). But, although (3a) is accepted as true within the context of classical theology, classical theology need not—and to my knowledge does not—hold (3b) to be true/ Hence, either 7

Both Boethius (see The Consolation o/ Philosophy, Book V, 6) and Aquinas are clear on this point. In considering a proposition similar to (3), Aquinas distinguishes between taking it as (3a) and as (3b). He then claims, in effect, that (3a) is true but (3b) is false. Thus he says: "Hence also this proposition, Everything known by God must necessarily be, is usually distinguished, for it may refer to the thing or to the saying. If it refers to the thing, it is divided and false; for the sense is, Everything which God knows is necessary. If understood of the saying, it is composite and true, for the sense is, This proposition, 'that which is known by God is9 is necessary" (Summa Theologiae, 14, 13, ad 3). Thus, no problem arises in classical theology in connection with (3b)—for that proposition is rejected as false. However, a problem does arise out of accepting (3a) and claiming, in addition, that the antecedent of (3a) is itself necessary. Thus in discussing the necessarily true conditional statement "If God knew that this thing will be, it will be "Aquinas explicitly points out concerning the antecedent "God knew this contingent to be future" that "it must be said that this antecedent is absolutely necessary" (ST., I, 14, 13, ad 2). In order to avoid the conclusion that no future event is contingent, Aquinas distinguishes between things as they exist in the knower (God) and as they exist in themselves. His suggestion is that in saying that "this thing will be" is absolutely necessary we are speaking of it only as it exists in God, and not as it exists in itself. As it exists in itself it is contingent, as it exists in God it is absolutely necessary. It is interesting to note that this difficulty does not arise in Aquinas' discussion of the divine will. ". . . but the divine knowledge has a necessary relation to the thing known; not the divine will to the thing willed. The reason for this is that knowledge is of things as they exist in the knower; but the will is directed to things as they exist in themselves. Since then all other things have necessary exist-

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

362

WILLIAMLROWE

the inference from (3) to (4) is valid but (3) (interpreted as (3b)) is not true or (3) (interpreted as (3a)) is true but the inference from (3) to (4) is invalid. On either account, we have sufficient grounds for rejecting the argument Augustine presents.8 In this paper I have (1) explained and criticized what I take to be Augustine's main reason for rejecting Evodius' argument that God's foreknowledge is incompatible with free will, and (2) suggested another way of rejecting the argument. Augustine does make, however, one further point which merits attention. He points out that the line of argument Evodius has taken implies not simply that God's foreknowledge is incompatible with free will but that anyone's foreknowledge is incompatible with free will. For he shows that on Evodius' reasoning it is foreknowledge generally and not God's foreknowledge specially that causes the events foreknown to happen by necessity. This creates a difficulty for Evodius since he believes that we sometimes foreknow the decisions and actions of men without thereby rendering those decisions and actions involuntary. The importance of this ad hominem argument against Evodius must not be overlooked. For if we claim that there is a special problem for the theologian who believes in divine foreknowledge and human freedom then the reasons we give in support of that ence inasmuch as they exist in God; but no absolute necessity so as to be necessary in themselves, insofar as they exist in themselves; it follows that God knows necessarily whatever He knows, but does not will necessarily whatever He wills" (ST., I, 19, 3, ad 6; Cf. Summa contra Gentiles, I, 81). Having admitted that God's willing p is not necessary, Aquinas is free to assert that p need not happen by necessity. He points out, of course, that the conditional proposition "If God wills p, p will happen" is necessary, but denies that any necessity need attach to the consequent "p will happen" (CG., I, 85). Thus he is able to say, "Although the non-existence of an effect of the divine will is incompatible with the divine will, the possibility that the effect should be lacking is given simultaneously with the divine will. God's willing someone to be saved and the possibility that that person be damned are not incompatible; but God's willing him to be saved and his actually being damned are incompatible" (De Veritate, 23, 5, ad 3). 8 This way of rejecting Augustine's argument was argued in some detail by C. £. Caton and myself in a joint paper, "Divine Foreknowledge and Contingent Events," presented at the meeting of the Western Division of the American Philosophical Association, May, 1961.

377

378

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION AUGUSTINE ON FOREKNOWLEDGE AND FREE WILL

363

claim must at some point concern God's foreknowledge, rather than foreknowledge generally. Otherwise, the theologian is involved in no more or less a difficulty than anyone who believes that we sometimes foreknow the free decisions and actions of men.

CHAPTER TWENTY-SEVEN

Religious Experience and the Principle of Credulity

In 1939 there appeared a very important essay by C. D. Broad on the argument from religious experience to the existence of God.1 In that essay Broad set forth what he took to be a basic principle which we commonly apply to various experiences in our efforts to make some rational judgments as to whether the experiences are veridical or delusory. Roughly put, Broad suggested that when someone has an experience which he takes to be an experience of x, it is rational to conclude that he really did experience x unless we have some positive reason for thinking otherwise. Broad then argued (1) that it is unreasonable to refuse to apply this principle to religious experience, (2) that there are no reasons of genuine merit for thinking that religious experiences are delusive, and concluded (3) that, there fore, probably religious experiences are veridical, probably they really are experiences of the objects which those who have the experiences take them to be of.2 Recently, Richard Swinburne has set forth a principle similar to Broad's, labeled it "The Principle of Credulity," and argued, like Broad, that the principle is applicable to religious experiences, that the special considerations that limit the principle (Broad's "positive reasons" for thinking some experience/s is/are delusive) are not present in any significant degree in many religious experiences, and concluded that probably religious experiences really are experiences of the objects which they seem to be of.3 In this paper I wish to discuss the Broad-Swinburne defense of the argument from religious experience. Before engaging in this discussion, certain remarks must be made concerning the nature and content of religious experience. These remarks are not altogether uncontroversial and they will have some bearing on the ultimate judgment I make on the Broad-Swinburne defense of the argument from religious experience. My aim in making these remarks, however, is not to weaken the Broad-Swinburne defense, but to set it on as strong a foundation as can reasonably be constructed for it. I shall define a religious experience as an experience which the subject takes to be an experience of God or some supernatural thing. This definition may be too narrow. Perhaps someone has an experience which he takes to be an awareness of some natural thing endowed momentarily with some supernatural quality. Perhaps such an experience should be regarded as religious. If so, then we can expand my

380

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 86

definition to include it. In any case, the argument from religious experience is generally constructed as.an argument for the existence of a supernatural object. So my definition, somewhat narrow as it may be, will suffice to select out those experiences which figure in the argument from religious experience. Why not define a religious experience more simply as an experience of God or some supernatural being? Why, that is, introduce the subjective element Hakes to be' into the definition? I do this for two reasons. First, it seems an obvious fact that religious experiences occur, a fact that even an atheist would wish to assent to. But if 'S experiences x' implies 'x exists' then without the qualification 'takes to be', an atheist would be in the uncomfortable position of denying that anyone has ever had a religious experience. Clearly this is an unwelcome conclusion. Our question should be not whether religious experiences occur — they surely do — but whether some or all religious experiences are veridical rather than delusory. Given my definition, theist and atheist alike can agree, for example, that on the road to Damascus Saul of Tarsus had a religious experience. The issue that divides them is whether his experience was veridical, an experience in which he really did encounter a divine being, or delusory, an experience which he took to be an experience of a divine being but was mistaken in so taking it. The second reason for insisting on the qualification 'takes to be' derives from the fact that it is possible for someone to experience x without being aware that he is experiencing x. Not being able to recognize a black walnut tree, you may look at a certain tree, be having an experience of seeing a black walnut tree, but not be having an experience which you take to be an experience of a black walnut tree. If you later say that you've never seen a black walnut tree, I may correct you by pointing out that the tree you were looking at was indeed a black walnut. It is one thing to experience x and another thing to recognize that what you are experiencing is x. It is, therefore, possible, for example, that the Virgin Mary might appear to two people A and B, but A have a religious experience while B does not. For A may recognize that what he is experiencing is indeed Mary, whereas B may think that he has had an experience merely of a woman in a strange white garment. A's experience is truly religious whereas B's experience is not, and this notwithstanding the fact that B really did experience the Virgin Mary. I take the term 'religious experience' to imply something about the subject of the experience. It is not enough that the object of the experience be a supernatural figure. Unless the subject experiences that object as supernatural, the subject has not had a religious experience. Therefore, it is essential to a religious experience that the subject takes the experience to be an experience of God or a supernatural being. One consequence of my definition must be again emphasized. Someone may have a religious experience even though he does not really experience God or any supernatural thing. For someone may have an experience which he takes to be an experience of x even though there is no x that he is experiencing. Macbeth had an experience which he took to be an experience of a dagger, but there was no dagger that he was then experiencing. So far as the argument from religious experience is concerned, the implication of this consequence is that the following inference is invalid.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

381 87

1. S has a religious experience, therefore, 2. God (or some supernatural thing) exists. As I noted earlier, this consequence is what we should expect. It does not prejudice the case against the argument from religious experience. Indeed, both Broad and Swinburne, who defend the argument from religious experience, insist on this consequence. The question, then, is whether we have reasons for thinking that religious experiences are veridical, whether they really are experiences of the objects which the subjects of the experiences take them to be of. What is it for a religious experience to be veridical, to really be an experience of the object which the subject takes it to be an experience of? Saul had an experience on the road to Damascus which he took to be an experience of a divine being. What would be the case if Saul's experience was veridical? Roughly, it would be the case that what he took to be the object of his experience actually existed, was present, and caused him to have the experience which he took to be of that object. Given, then, that religious experiences occur, the question we must raise is whether we have rational grounds for thinking that the putative objects of the experience actually exist, are present, and cause the subjects to have those experiences. It is with respect to this question that Broad and Swinburne advance the principle of credulity and argue forcefully that it ultimately leads us to give an affirmative answer to the question. We are now in a position to discuss their arguments on this matter. Adapting Broad's reasoning concerning mystical experience to the argument from the experience taken to be of God, we can set forth his use of the principle of credulity as follows. 1. Experiences occur which their subjects take to be experiences of God. 2. When subjects have an experience which they take to be of x, it is rational to conclude that they really do experience x unless we have some positive reasons to think their experiences are delusive. 3. There are no good, positive reasons for thinking that all or most experiences which their subjects take to be of God are delusive. therefore, 4. It is rational to believe that at least some experiences which their subjects take to be of God really are experiences of God. therefore, 5. It is rational to believe that God exists. One important question to be raised about this argument is the question of the phenomenological content of the 'experience of God'. What is the actual content of the experience and just how does God enter into the experience? If we think of God in terms of the classical concept developed in the religions of the West, God is

382

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 88

a person who is, in part, omnipotent, omniscient, perfectly good, and the creator of all things. Many have claimed to experience such a being, but what is the actual, phenomenological content of these experiences? Apparently the phenomenological content may vary considerably. It may consists largely of feelings without much in the way of visual or auditory sensations. Alternatively, one may hear a voice, "see" a divine figure, etc. In view of this and our concept of God it is tempting to conclude that God as such is never an immediate object of experience, the immediate objects being certain feelings or sensible characteristics. Rather God enters in by way of an interpretation of the experience, the feelings and/or sensible characteristics being such as to give rise to the conviction that God is truly present. This way of describing the experience of God leads to the skeptical question: what is it about the content of your experience that justifies you in thinking that God is present? Indeed, as Otto observed in rejecting Schleiermacher's description of religious experience in terms of the feeling of absolute dependence, the end result of this view of religious experience is twofold: first we end in subjectivism, religious experience being reduced to experiencing one's self as feeling a certain way; second, God is reached only by way of inference,as being the external cause of the content of the experience, the feeling of absolute dependence. Thus by severely restricting what may properly be construed as the objects of an experience, the argument from religious experience to the existence of God is defeated at the outset. The so-called experience of God is really the experience of things other than God which, perhaps for no good reason, give rise to a conviction that God is indeed present. Although the distinction between the actual content of an experience and that is only an interpretation of the experience is undoubtedly a viable distinction and may have some role to play in our understanding of religious experience, in this paper I don't wish to press that distinction in order to eliminate the argument from religious experience. Just as we may say at the level of common-sense that Jones saw a tree, and not that the tree enters into consideration only by way of an interpretation of his actual experience, so I shall allow that if there is a God then there may be an experience of God in which God is indeed the object of the experience. To hold otherwise is to stack the deck against the argument from religious experience. What, if anything, is wrong with the argument from religious experience to the existence of God? There are, it seems, only two points in the argument which may be attacked. We may attack the third premise, arguing that there are positive reasons for thinking that religious experiences are likely to be delusive. Alternatively, we may reject the second premise, denying the principle of credulity, or at least denying its applicability to religious experience. What positive reasons might we possess or come to posses which would be grounds for taking someone's experience as delusive? If someone has an experience which he takes to be an experience of x, (1) we might have good independent grounds for thinking that x does not exist, that x was not present, or that x was not the cause of his experience, or (2) we might have grounds for thinking that the

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

383 89

experience occurred under circumstances that often result in delusory experiences. So far as (1) is concerned we may hold that there are independent grounds for thinking that God does not exist. I have argued elsewhere, for example, that the variety and profusion of evil in the world does render it probable that God does not exist.4 But this is a notoriously controversial issue. Some philosophers have argued that the facts about evil do not add in the least to the probability that God does not exist.5 Moreover, even though I think that the variety and profusion of evil does render God's existence improbable, it does seem to me that the evidence for theism may be such as to outweigh the negative evidence based on evil. So the argument from evil is perhaps not a sufficient reason for rejecting religious experience of God as delusive. A second reason that has been given for the non-existence of God is that the concept of God is incoherent. But here, too, there are strong arguments on the other side.6 The fact of the matter is that there are no generally accepted arguments that prove or render very probable the non-existence of God. Nor have we any good reasons for thinking that God is not present where the experience takes place. If God exists he is present everywhere. There are causal accounts of the experience of God that make factors other than God the cause of the experience. But, again, these causal accounts are open to serious criticisms and, in any case, are not generally accepted.7 Some philosophers have focussed on the circumstances in which religious experiences occur and have sought to find in those circumstances positive reasons for thinking that the experiences are delusive. Bertrand Russell, for example, has argued that those who enjoy religious experiences generally have those experiences only after they have brought about physical and mental changes within themselves by fasting, meditation, and the like. He compares these "abnormal states" with those produced by the excessive use of alcohol. With characteristic wit and style he concludes: "From a scientific point of view, we can make no distinction between the man who eats little and sees heaven and the man who drinks much and sees snakes. Each is in an abnormal physical condition and therefore has abnormal perceptions."8 The trouble with Russell's argument is that it assumes that bodily and mental conditions that make for unreliable and/or delusory experiences of the natural world also would make for unreliable and/or delusory experiences of the supernatural realm, if there is such a realm to be experienced. Perhaps this assumption is merited, but it certainly is not obviously true. Indeed, for all we know, the opposite may be the case. As Broad has remarked: Suppose, for the sake of argument, that there is an aspect of the world which remains altogether outside the ken of ordinary persons, in their daily life. Then it seems very likely that some degree of mental and physical abnormality would be a necessary condition for getting sufficiently loosened from the objects of ordinary sense-perception to come into cognitive contact with this aspect of reality. Therefore the fact that those persons who claim this peculiar kind of cognition generally exhibit certain mental and physical abnormalities is rather what might be anticipated if their claims were true.

384

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 90

One might need to be slightly "cracked" in order to have some peep-holes into the super-sensible world.9 The problem is that we don't know what bodily and mental conditions are likely to lead to delusory experiences of God. Nor, for that matter, do we know what conditions are such that if we satisfy them we will have an experience of God, if there is a God to experienced. This last point marks another difference between religious experience, particularly the experience of God, and our experiences of natural objects. Most existing objects are such that there are conditions which if satisfied by the subject, the experience will follow. This is an important point we often rely on in judging whether a particular perceptual experience is veridical or delusory. A book, for example, is such that if a normal perceiver is rightly placed, eyes open, attentive, etc., the perceiver can be expected to have an experience which he takes to be one of seeing a book. If we know that several subjects satisfy the conditions but do not have the experience this will be grounds for taking a particular subject's putative experience of the book as delusive rather than veridical. God, however, is not such an object. God may choose to reveal himself to A but not reveal himself to B under similar conditions. This means that the failure of a number of subjects to have the experience of God under conditions similar to those in which A had such an experience need not count against A's experience being veridical. The conclusion I draw from these points is that the attempt to defeat our adaptation of Broad's argument by attacking premise (3), by trying to find positive reasons for thinking religious experience to be delusive, is unsuccessful. Should we then, following Broad and Swinburne, conclude that the argument from religious experience to the existence of God is successful, that it renders it at least probable that God exists? My final remarks are addressed to this question. As I indicated earlier, the only other point of attack is the second premise, the principle of credulity itself. Here I am inclined to agree with Broad and Swinburne, at least to the extent of admitting that we do assume a principle akin to their principle of credulity. My question is not whether we assume such a principle but whether the principle is applicable to religious experience. The problem concerns the finding of positive reasons for rejecting a particular experience or type of experience as probably delusive. Here I think there is an important difference between (1) knowing how to proceed to find positive reasons, if there should be any, for rejecting an experience as probably delusive, and (2) not knowing how to proceed to find such positive reasons if there should be any. When we are in situation (1), as we clearly are in the case of those who habitually drink alcohol to excess and report experiences which they take to be of rats and snakes, the application of the principle of credulity is clearly in order. And in that case, we are successful in finding, positive reasons for thinking their experiences delusive. But when we are in situation (2), as we seem to be in the case of religious experience, I am doubtful that the application of the principle of credulity is warranted. Since we don't know what circumstances make for delusory religious experiences and we

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

385 91

don't know what the conditions are which, if satisfied, one would have the experience of God if there is a God to be experienced, we can't really go about the process of determining whether there are or are not positive reasons for thinking religious experiences to be probably delusive. The principle of credulity presupposes, if you like, that we are in situation (1) with respect to the experiences whose cognitive status we are trying to determine. When we are in situation (2), as I believe we are with religious experiences, the principle of credulity does not, or should not, apply. My response, then, to the Broad-Swinburne defense of the argument from religious experience is that they are right in holding that we do assume something like the principle of credulity, that they are right in thinking that if we apply the principle to the experiences which seem to be of God we will be able to conclude that probably God exists, but that they are wrong in thinking that we are justified in applying the principle to religious experiences. My objection to the Broad-Swinburne defense of the argument from religious experience can be put a bit more clearly by distinguishing two versions of the principle of credulity. I. When subjects have an experience which they take tb be of x, it is rational to conclude that they really do experience x unless we have some positive reasons to think their experiences are delusive. II. When subjects have an experience which they take to be of x, and we know how to discover positive reasons for thinking their experiences delusive, if such reasons do exist, then it is rational to conclude that they really do experience x unless we have some positive reasons to think their experiences are delusive. My thesis is twofold; first I think that II, not I, better expresses the principle of credulity which is (or ought to be) a basic principle of rationality; second, I think that we do not know how to discover positive reasons for thinking that religious experiences are delusive, if such reasons do exist. I conclude, therefore, that the principle of credulity cannot play the role in the argument from religious experience that Broad and Swinburne assign to it. And since the argument from religious experience appears to depend on the principle of credulity, I conclude that the argument is defective and, therefore, fails to show that it is probable that God exists.

NOTES 1. C. D. Broad, "Arguments for the Existence of God, II," The Journal of Theological Studies, XL (1939), 157-167. 2. Although Broad indicates that his intention is to discuss religious experience, and to some extent carries out this intention, much of his discussion is centered more narrowly on mystical experience. I have adapted this discussion and argument concerning mystical experience to the broader category of religious experience. 3. See Chapter 13, "The Argument from Religious Experience," in Richard Swinburne, The

386

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 92 Existence of God (Oxford: Oxford University Press, 1979), pp. 244-276. 4. William L. Rowe, *The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly, XVI, No. 4, (October, 1979), 335-341. 5. See Richard Swinburne, "The Argument from Religious Experience," and Alvin Plantinga, "The Probabilistic Argument From Evil,"Philosophical Studies, 35(1979), 1-53. 6. See, for example, Richard Swinburne, The Coherence of Theism (Oxford: Oxford University Press, 1977). 7. An excellent exposition and critical account of the Freudian causal account of theistic belief may be found in William P. Alston's "Psychoanalytic Theory and Theistic Belief," Faith and the Philosophers, ed. by John Hick (New York: St. Martin's Press, Inc., 1964), pp. 63-102. 8. Bertrand Russel, Religion and Science (London: Oxford University Press, 1935), p. 188. 9. Broad, "Arguments for the Existence of God, II," p. 164.

CHAPTER TWENTY-EIGHT

The Rationality of Religious Belief

Some beliefs are rational, others are irrational, and still others may be neither rational nor irrational. It is, of course, one thing for a belief to be rational and another thing for it to be true; but we believe that rational beliefs are more likely to be true than either irrational beliefs or beliefs that are neither rational nor irrational. For this reason, as well as others, we value a person's being rational with respect to her beliefs and regard it as a failing for a person to be irrational with respect to her beliefs. For several reasons it is no easy matter to determine whether someone's belief is rational. First, there is no accepted, non-trivial definition of what it is for a belief to be rational. Second, what is rational for one person to believe may be irrational for another person to believe. And finally, as we shall see when we consider basic beliefs, there may not be any way to establish an objective criterion of rationality. Despite these rather severe difficulties, I propose to investigate the question of whether religious belief is rational. Actually, it is the rationality of the belief that God exists that I want to consider, where the concept of God in question is that of a supremely good being who is all-powerful, all-knowing, and creator of the world. The question I want to discuss is whether that belief is rational. Affirmative answers to the question of the rationality of belief in God may be divided into two types. The first type holds that it is rational to believe in God because the proposition that God exists can be derived from, or is probable in relation to, other propositions that we know or which it is rational for us to believe. Some philosophers and theologians who embrace the Ontological Argument contend that the belief that God exists is rational because it is logically derivable from propositions that are selfevidently true. Others may hold that belief in God is rational because the proposition that God exists is logically derivable from propositions that are known to be true, but are not all self-evident. And still others may hold that belief in God is rational because the proposition that God exists is probable in relation to other propositions that it is rational for us to accept. Here it is important that the case for God's existence take into account all the things we know or rationally accept which bear on the truth or falsity of the claim that God exists. For a proposition may be probable in relation to some of the things we know but improbable in relation to all of the things we know. God's existence may be probable relative to the existence of a world with certain features, but improbable in relation to all the features we know the world to contain. On the other hand, one may admit that the existence of God is improbable in relation to the profusion and variety of evil in the world, but insist that it is still more probable than not given all that we know about the

388

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

world.1 The second type of affirmative answer to the question of whether the belief that God exists is rational rests on the distinction between basic beliefs and non-basic beliefs. A belief p is basic for a person S at a time t just in case S accepts p at t but does not accept p on the basis of other beliefs he holds at t. Although the notion of what it is to accept one belief on the basis of another is difficult to explicate, examples of basic beliefs and nonbasic beliefs are not hard to come by. The belief that I have a headache is such that when I hold it, it is a basic belief for me. My belief that you have a headache is normally not a basic belief. For I generally believe it only on the basis of other beliefs that I have, such as the belief that you have just told me that you have a headache. My belief that 2 + 1 = 3 is a basic belief for me. But my belief that 241 + 159 = 400 is not a basic belief; it is based on other beliefs I hold, such as that 1+9=10, 5+4=9, etc.2 Of course, the fact that a belief is basic for a certain person in no way implies that the belief is rational. Some basic beliefs are rational, others are irrational, and still others may be neither rational nor irrational. In addition to the notion of basic belief, we need the concept of a properly basic belief.3 A belief is properly basic for a person S at t just in case it is basic for S at t and S is rationally justified in accepting it at t. My belief that I have a headache is not only basic, it is properly basic, it is a belief that I am rationally justified in holding. So too, for my basic belief that 2+1=3. Being self-evident to me, the proposition that 2 + 1 = 3 is such that I am rationally justified in believing it. On the other hand, a compulsive gambler may suddenly come to believe that the next hand dealt to him will be the winning hand. He may have no other beliefs that he takes to be significant evidence for this belief. Perhaps some deep psychological need has brought about this belief. His belief is basic, but not properly basic. For there is nothing about him or the situation he is in that would render the belief rational. Contrast this with someone who upon looking out the window and having a visual experience which she takes to be of a cat climbing a tree immediately forms the belief that there is a cat climbing a tree. Her belief that there is a cat climbing a tree is rendered rational by the situation she is in-her looking out the window and seeming to see a cat climbing a tree, etc. It's not that she has some other rational beliefs and infers her present belief from them—she doesn't say to herself: "I'm looking out the window and seem to see a cat climbing a tree. Let's see. What can I infer 1 A good example of the attempt to establish that the existence of God is probable in relation to what we know about the world is Richard Swinburne's The Existence of God (Oxford: Oxford University Press, 1979). Swinburne argues that the variety and profusion of evil does not add anything to the probability that God does not exist, but that the evidence from religious experience taken in conjunction with what we know about our world makes the theistic hypothesis more probable than not. 2 Can a belief be basic and be based on other beliefs we hold? I think so. Some propositions may seem self-evident and also be clearly seen by us to follow from other propositions which we take to be selfevident. For example, the logical principle that if p implies q then not-q implies not-p seems self-evident to me. But I may also hold other self-evident beliefs from which I clearly see that this principle follows; I do then hold this principle on the basis of other beliefs I hold, but I do not hold it only on the basis of those other beliefs. Taking this possibility into account we perhaps should characterize a basic belief as one that a person accepts at a time but does not accept only on the basis of other beliefs he accepts at that time. I will ignore this point in my discussion, but its importance for religious belief should not be overlooked. Anselm may have held God's existence to be self-evident and therefore basic, and yet also have held that we see .that it immediately follows from other self-evident beliefs we hold. 3 The expression 'properly basic belief is Plantinga's. See references in footnote 5.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

389

from that belief? Oh yes, I can infer that I see a cat climbing a tree." She has no evidence for her belief in the sense of other beliefs on the basis of which she holds her belief that there is a cat climbing a tree. Her belief is thus basic and rational (a properly basic belief). We might say of her belief that it is grounded in a situation that renders her rationally justified in holding that belief. The gambler's belief is either groundless or grounded in some situation that fails to render his belief rational. As we've noted, many of our beliefs are rationally justified by virtue of being based on other rationally held beliefs that constitute evidence for them. This, however, is not the only way in which a belief can be rationally justified. For if it were then, barring a circle of justification, if we were rationally justified in accepting any belief there would have to be an absolutely infinite number of beliefs that we accept, each of which we would also have to be rationally justified in accepting. Given then that we have rationally justified beliefs, some beliefs we hold must be such that we are rationally justified in accepting them without having evidence for them, without believing them on the basis of other rationally justified beliefs which support them.4 Properly basic beliefs are beliefs of this sort. The second type of affirmative answer to the question of whether the belief that God exists is rational holds that it is rational to so believe because the belief that God exists is a properly basic belief. This view has been set forth and vigorously defended by Alvin Plantinga in two important papers.5According to Plantinga the view that belief in God is properly basic is traceable to John Calvin and is a fairly standard view among Reformed theologians. To this I think we can add a long tradition of thought that can be found in Augustine and receives its most significant expression in Bonaventure. Bonaventure held that the knowledge of God's existence is implanted within us, that the proposition that God exists is self-evident and indubitable — anyone who correctly apprehends that proposition sees its truth clearly and cannot entertain any doubts of its certainty.6 Clearly, then, according to Bonaventure the proposition that God exists can be known immediately to be true, we don't need to infer it from other things we know in order to be rationally justified in accepting it. The belief that God exists is properly basic for those who correctly apprehend the proposition that God exists. Is belief in God rationally justified? A complete discussion of this question would require an examination of both types of affirmative answers that have been given to it. However, since the second type of answer has received considerably less treatment than the first, I will limit the following discussion to the view that belief in God is rational because the belief that God exists is properly basic.

4 If all that is required for justification is that we be able to justify one belief by another, as opposed to actually having justified the one by the other, the infinite regress argument for properly basic beliefs may not be sound. For another argument in support of epistemologically basic beliefs see John Pollock, Knowledge and Justification (Princeton: Princeton University Press, 1974) 27-29. 5 "Is Belief in God Rational?" in Rationality and Religious Belief, ed. C. Delaney (Notre Dame: University of Notre Dame Press, 1979), 7-27; "Is Belief in God Properly Basic?" Nous, Vol. XV, No. LMarch 1981,41-51. 6 See Question I, Article I of Bonaventure's Disputed Questions On The Trinity.

390

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

Our question then is whether belief in God is properly basic. To contend that it is not properly basic one must hold either that it is not a basic belief at all or that, although basic, it is not properly basic. Let's consider the first possibility. Can we reasonably hold that belief in God is not a basic belief, that those who hold this belief hold it on the basis of other beliefs they accept? I think the answer to this question is no, but proving this is rather difficult in the absence of a clear criterion of basicality of belief. A belief is basic for a person just in case he accepts it without accepting it on the basis of any of his other beliefs. But what is it for a person to hold a belief on the basis of another belief and how can we tell that a person holds a given belief but does not hold it on the basis of any of his other beliefs? Often, when one belief is based on another the second plays a causal role in the formation of the first. My belief that you have just told me that you have a headache gives rise to my belief that you have a headache. But it would be a mistake to identify the based on relation with the causal role that some beliefs play in the formation of new beliefs. For we want the based on relation to be essentially connected to the notion of justification, or at least to what one takes to be one's justification for his belief. And the plain fact is that one belief may cause another even though the second is not based on the first, for the first belief may not figure in the justification of the second. My belief that a colleague has spoken well of my work may casually contribute to my believing that his work has greater merit than I hitherto thought. But my new belief is not based on my belief that he has spoken very highly of my work. Indeed, my new belief may be a basic belief, one I accept without holding it on the basis of my other beliefs. Holding a belief on the basis of other beliefs is also not to be confused with the fact that many of my beliefs are such that I would not accept them unless I believed other things. My belief that my colleague's work is better than I earlier .thought is something I would not believe unless I also believed that I have a colleague and that he has produced some work. But I do not believe the former on the basis of the latter. Holding one belief on the basis of another perhaps can be paraphrased as taking the second as rational support for or evidence for the former. A belief p is basic for me if I do not hold other beliefs which I take to be evidence or support for my belief that p. Although this paraphrase may serve to distinguish the based on relation from some other notions, it still leaves us with difficulties. For one thing, how can we be sure that someone holds a belief without taking any of his other beliefs as evidence for it? We might ask him. But couldn't he be mistaken in his answer? Second, we need to note that the paraphrase renders the based on relation entirely subjective: a belief is based on another if the person takes the second to be evidence for the first. And, as we all know, it is entirely possible to regard one belief as evidential support for another when, objectively speaking, it is not evidence for it at all. Moreover, we also have the possibility that a person holds a belief which is in fact good evidence for another belief that he holds, but the person's second belief is not based on the first because he does not take his first belief to provide evidential support for the second. I may believe that a certain horse will win the race and also believe with good reason that the fix is in for this horse to win. My second belief may provide evidential support for the first; and yet I may believe the first without taking the

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

391

second as evidence for it. Indeed what I take as evidence for it may be my belief that I saw his winning in my morning tea cup. In this case I hold one belief (the fix is in) which is evidence for my belief that a certain horse will win, and a second belief (I saw it in the tea cup) which isn't evidence for it; but I hold the belief in question on the basis of the second belief and not on the basis of the first. Perhaps it would be irrational for me to do this. But that doesn't mean that it can't be done.7 Despite these problems, I shall continue to view the based on relation in terms of a person's taking one belief as evidence for or support for another. Our question then is whether belief in God is a basic belief. Is it a belief which people hold without holding it on the basis of any of their other beliefs, without, that is, taking any of their other beliefs as evidential support for it? Perhaps we can now answer this question. Many of us were taught religious beliefs in much the same way as we were taught some of our beliefs about the world and some of our beliefs about arithmetic. At a very early age we were taught that Columbus discovered America, that 2+1=3, and that there is a God. We accepted these beliefs without question. They were not presented to us as beliefs for which we needed evidence. Nor were they brought into question by anyone with whom we associated. Having no need for evidence, we accepted them without evidence, we did not accept them on the basis of other beliefs we held at the time. They were basic beliefs for us, beliefs we held without taking any of our other beliefs as evidence for them. Of course, as time went on some of our basic beliefs became non-basic, we came to see that other beliefs we held were indeed evidence for them and to hold them on the basis of those other beliefs. And still other basic beliefs we gave up believing altogether. For example, as children some of us were told by our parents that Santa Claus really does bring presents, and we immediately believed that there is such a being as Santa Claus without inferring this belief from other beliefs we hold.8 Of course after a relatively short period of time our peers managed to disabuse us of this belief so that it ceased to be basic and was abandoned altogether. For some of us, however, belief in God may have remained a basic belief, a belief so unquestioned that we have never felt the need to abandon it or to take any of our other beliefs as evidence for it.9 If the preceding account is correct, as I think it is, belief in God is a basic belief for an enormous number of people. Thus if our question is whether belief in God is properly basic we cannot contend that it is not by arguing that it is not basic. We must, admitting that it is basic, argue that it is not properly basic, that it is not a rationally justified basic belief. 7 This example may overly strain our intuitions. We might say of such a person that he does not have our concept of rational justification, that he means something different by "evidence" from what we mean. Perhaps that is so, but also perhaps not. He may mean what we do but just have different beliefs about what constitutes evidence for something. 8 Actually, the basic belief here is that Santa Claus really does bring presents, from which our belief that there is such a being immediately follows. I assume here that on hearing our parents we immediately formed the belief that Santa Claus really does bring presents. 9 Perhaps on learning of the arguments for God's existence we came to believe that certain of bur beliefs about the world did provide evidential support for our belief in God. At this point did our belief in God cease to be basic? I think the answer to this question-may be no. See footnote 2.

392

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

When a person holds a belief without holding it on the basis of any of her other beliefs, how can we determine whether her belief is properly basic, whether she is rationally justified in holding it? A number of philosophers have held that we can answer this question only by having a criterion of proper basicality, a principle the application of which will tell us whether a basic belief is or is not a rationally justified belief.10 Modern foundationalists, for example, hold the principle that a belief is properly basic just in case it is either self-evident or incorrigible. Believing as he did that the proposition that God exists is self-evident to those who correctly apprehend it, Bonaventure might not be unduly troubled by this criterion. But many who hold that belief in God is properly basic also hold that the belief is neither self-evident (or at least not self-evident to us) nor incorrigible. They then will reject this criterion, or at least will ask the foundationalist to justify it. But how can the foundationalist justify his criterion? Perhaps he thinks belief in his criterion needs no justification, that belief in his criterion is properly basic. But then as Planting has argued, his position is self-referentially incoherent.11 For the criterion is not self-evidently true, nor is it incorrigible. So, as Plantinga observes, in accepting his criterion as properly basic the foundationalist violates the conditions of proper basicality he himself lays down in accepting it. The truth seems to be that the foundationalist can support his criterion only by an appeal to cases. He must show that clear examples of properly basic beliefs satisfy the criterion, and clear examples of beliefs that aren't properly basic don't satisfy it.12 But this procedure presupposes that we already know of some beliefs that they are properly basic and of some other beliefs that they aren't; a knowledge that is prior to our knowledge of any criterion of proper basicality. And it is just at this point that we run into a perhaps insurmountable difficulty. Suppose that the theist takes belief in God as one of his clear cases of a properly basic belief. The foundation list will not include belief in God among his set of clear examples of properly basic beliefs. But so what? Does that suffice to show that the theist is wrong in accepting belief in God as properly basic? I don't think so. There is no reason in principle why the theist's set of examples must conform to the foundationalist's set, any more than there is for thinking that the foundationalist's set must conform to the theist's set. And, as we've already seen, we cannot resolve this matter by appealing to a criterion of proper basicality. Lacking a criterion, neither the theist nor the foundationalist can rationally refute the other's set of "properly basic" beliefs. Each will hold that the other is mistaken about the belief that God exists — but it is one thing to hold that someone is mistaken, and quite another thing to show that his view isn't rationally justified. Further argument here is likely to be question-begging. Perhaps we are at the level of "persuasion." We (theist and non-theist) must each tell our respective stories in the hope of convincing the other of the questionable character of his position. 10

Those who hold this sort of position — that is, the position that in order to determine what things we know we must first establish a criterion or method for selecting out what we really know from what we only think we know — are called "methodists" by Roderick Chisholm, See his The Problem of the Criterion (Milwaukee: Marquette University Press, 1973). Those who hold that we must first start with clear examples of things we know and then arrive at a criterion of knowledge are called "particularists" by Chisholm. 11 "Is Belief in God Rational?" 25-26; "Is Belief in God Properly Basic?" 49-50. 12 I agree with Plantinga and Chisholm in adopting the position of the "particularists."

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

393

In what follows, I'm going to tell the non-theist's story (or one of them). It is a story I find convincing. Whether it will convince the theist who takes belief in God as properly basic is another matter. One notable difference between the modem f oundationalist's initial set of properly basic beliefs and the theist's - that is, a theist whose initial set of properly basic beliefs includes a belief in God - is that the theist's set will include beliefs about the existence of objects, both physical and spiritual, beliefs that are neither self-evident nor incorrigible. Thus Plantinga holds that beliefs such as I see a tree, that person is angry and / had breakfast this morning are in certain circumstances properly basic. Here I think Plantinga is right. When I have the experience of seeming to see a tree, and other circumstances obtain, my belief that I saw a tree is one that I am rationally justified in accepting. That I am in these circumstances is what justifies me in believing that I have seen a tree. But often enough my belief that I see a tree is not based on my belief that I seem to see a tree or my belief that the other circumstances obtain. For at the time I believe that I see a tree I may be so preoccupied with seeing the tree that I don't even form the belief that my experience is one of seeming to see a tree. And if I don't at the time in question have the latter belief, the belief that I see a tree can't be based on it. Upon having experience of a certain sort, I believe that I am perceiving a tree. In the typical case I do not hold this belief on the basis of other beliefs; it is nonetheless not groundless. . . We might say that this experience, together, perhaps, with other circumstances, is what justifies me in holding it; this is the ground of my justification, and, by extension, the ground of the belief itself.13

Turning to belief in God Plantinga argues that the case is similar. It is in particular circumstances that belief in God is properly basic. Upon reading the Bible, one may be impressed with a deep sense that God is speaking to him. Upon having done what I know is cheap, or wrong, or wicked I may feel guilty in God's sight and form the belief God disapproves of what I've done. Upon confession and repentance, I may feel forgiven, forming the belief God forgives me for what I've done. A person in grave danger may turn to God, asking for his protection and help; and of course he or she then forms the belief that God is indeed able to hear and help if he sees fit.14

In circumstances such as these one forms and accepts beliefs such as God disapproves of what I have done. Got forgives me, and God is to be thanked and praised. Each of these beliefs, so Plantinga argues, is properly basic in its respective circumstances. Since each selfevidently entails that God exists, the belief that God exists, if not itself properly basic, is immediately based on beliefs that are. Of course, feeling that one is guilty in God's sight is not itself sufficient to rationally justify the basic belief that God disapproves of what you've done. For suppose you also have very good reason to believe that given your severe religious upbringing you cannot avoid having a sense of religious guilt when you do something wicked. That is, 13

14

"Is Belief in God Properly Basic?/' 44-45.

Ibid.,46.

394

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

you have good reason to believe that given your religious upbringing you would have a sense of religious guilt when you do something wicked Whether or not there is a God. In this situation you may not be rational in holding the belief that God disapproves of what you've done. For you know that even if there is no God at all you still would feel religiously guilty upon doing something wicked. Just what must the circumstances be if my being in those circumstances justifies me in holding a basic belief, for example the belief that I see a tree. Well, in the normal case one very important circumstance is that I have the experience of seeming to see a tree, that I am appeared to be in the way that I am when I see a tree. Is being in this circumstance at t a sufficient condition of my belief at t that I see a tree being properly basic? Presumably not, for we know that if other circumstances are present we could seem to see a tree without actually seeing a tree, we could be hallucinating a tree. So something else is required beyond its merely seeming to me that I see a tree if I am to be rationally justified in accepting my basic belief that I see a tree. Although what else is required can't be stated with any precision, we can perhaps approach it by including the condition that we know of no positive reason for thinking that our experience may be delusive. Tentatively, then, we can say that when a person is in the circumstance of having the experience of seeming to see a tree and knows of no positive reason for thinking that his experience may be delusive, then if his belief that he sees a tree is basic he is rationally justified in accepting it, his belief that he sees a tree is properly basic. Of course, the same must hold for the circumstances in which one is justified in accepting belief in God as basic. It is not enough that I be in the circumstance of feeling forgiven by God in order to be rationally justified in accepting as basic the belief that God forgives me for what I've done. Perhaps my experience was delusive, not brought about by God at all. It would seem that here too we must add some further condition: something like my knowing of no positive reason for thinking that my experience may be delusive. Perhaps, then, Plantinga's position comes roughly to this. When a person's circumstances are such that he has the experience of, say, feeling forgiven by God, and does not know of any positive reason for thinking that the experience may be delusive, then his basic belief that God forgives him for what he's done is rationally justified, it is a properly basic belief. Let's begin our criticism of this view by returning to the case of seeing a tree. Consider again the, added condition that the person does not know of any positive reason for thinking that his experience may be delusive. This condition is a good bit weaker than the condition that the person knows that no positive reason exists for thinking that his experience may be delusive. The latter condition is too strong. Of course, if a person has the experience of seeming to see a tree and does know that no positive reason exists for thinking that his experience may be delusive, then he is rationally justified in accepting his basic belief that he sees a tree. But we are often rationally justified in our belief that we see a tree even though we do not know that no positive reason exists for thinking that our experience of seeming to see a tree may be delusive. The range of factors that could render our experience delusive is probably quite large, and some of those factors hardly ever occur. It is simply too stringent to require that we know that absolutely none of those

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

395

factors is present. On the other hand, it is perhaps not enough simply to require that we do not know that any such factors are present. For we may not know what sorts of factors interfere either with veridical perception in general or of trees in particular and, for this reason alone, have no reason to believe that any interfering factors are present. In this case we may not judge that the person is rationally justified in accepting as basic the belief that he sees a tree when he has the experience of seeming to see a tree. Suppose someone has an experience of seeming to see a tree and does not know of any positive reason for thinking his experience may be delusive. This lack of knowledge may be informed or uninformed. It is informed when the person knows what sorts of factors interfere with veridical perception of trees and has ho reason to think that any of these factors is present. His lack of knowledge of any positive reasons for thinking his experience to be delusive is uninformed when the person either has little or no knowledge of what sorts of factors interfere with veridical perception of trees or the factors are such that he has no way of finding out whether they are present. With this distinction in mind we can perhaps say that in the normal case when a person forms the basic belief that he is seeing a tree, his belief is properly basic just in case he has the experience of seeming to see a tree and does not know (in an informed way) of any positive reason for thinking his experience may be delusive. This is to say that the person has some knowledge or understanding of the factors that, if present, would render the experience delusive and is able to tell whether these factors are likely to be present. Turning once again to the belief that God forgives me for what I've done, we are now in a position to see why it may not be a properly basic belief. As Plantinga notes, the belief may be properly basic in the circumstance of the person's having an experience like feeling forgiven by God. But, as Plantinga also realizes, some other condition must obtain as well. I've suggested that the further condition is something like not knowing of any positive reason for thinking the experience to be delusive. But, as we've just seen, this absence of knowledge may be informed or uninformed, and normally only if it is informed is the person's belief that God forgives him for what he has done properly basic. But it is precisely at this point that the analogy between belief in God and belief in physical objects collapses. In the latter case we do know what the interfering factors are and we do know (roughly) how to determine whether or not they are present. Hence our not knowing of any positive reason for thinking our experience to be delusive often is an informed lack of knowledge. We do not lack such knowledge simply because we either don't know what factors would interfere with veridical perception or, if we do know, have no way of telling whether any of them are present. But when we turn to experiences which seem to be of God's actions or attributes our lack of knowledge of any positive reasons for thinking that these experiences may be delusive is, I think, an uninformed lack of knowledge - we don't know of the presence of such positive reasons because either we don't know what factors make for delusive experience here, or if we do know what these factors are we have no way of telling whether they are present. This being so, our initial judgment is that circumstances in which people have these experiences do not rationally justify them in holding such basic beliefs as God forgives me for what I've done. Although

396

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

there are circumstances in which such basic beliefs are formed, the circumstances fail to render these beliefs properly basic. I cannot undertake here to prove that the theist's not knowing any positive reasons for thinking that his experience (feeling forgiven by God, for example) may be delusive is an uninformed lack of knowledge. I can at best indicate some considerations in support of this judgment. These considerations appear when we contrast experiences which seem to be of physical objects with experiences which seem to be of God. In the former case we do have two important sorts of knowledge. First, we know some circumstances that often make for delusory experiences of physical objects: excessive alcohol, extended periods of sensory deprivation, etc. We don't have a similar knowledge of what circumstances lead to delusory experiences of God. Second, we know a good bit about what the conditions are which, if satisfied, one would have the experience of a physical object if that physical object actually exists. A book, for example, is such that if a normal perceiver is rightly placed, eyes open, attentive etc., the perceiver can be expected to have an experience which he takes to be one of seeing a book. If we know that several subjects satisfy the conditions but do not have the experience, this will be grounds for taking a particular subject's putative experience of the book as delusive rather than veridical. God, however, is not such an object. God may choose to reveal himself to A but not to B under similar conditions. This means that the failure of a number of subjects to have the experience under conditions similar to those in which A had such an experience need not count against A's experience being veridical. The points of contrast we've just made between the sorts of knowledge we have concerning whether experiences of physical objects are veridical or delusory, and concerning whether experiences of God are veridical or delusory, support the view that our lack of knowledge of positive reasons for thinking that an experience of God to be delusive is an uninformed lack of knowledge. The truth is that we simply don't know how to proceed to discover positive reasons for thinking that ah experience which seems to be of God is or may be delusive.15 Hence, our lack of knowledge of the existence of such reasons is an uninformed lack of knowledge. And if this be so then the circumstances in which a person has an experience of feeling forgiven by God and does not know of any positive reasons for thinking that the experience may be delusive are not such as to rationally justify that person in accepting his basic belief that God forgives him. Against this view that a belief is properly basic for a person only if that person is aware of a variety of factors (and can tell if they are present) that would indicate that the circumstances in which the belief is formed are insufficient to rationally justify the belief, a serious objection can be raised. For we can all think of examples in which an understanding of defeating factors is totally lacking, and yet it would seem plainly incorrect to hold 15

Of course, since God is a perfectly good being, we can from that fact alone discover some reason for thinking an experience that purports to be of God is delusive. For suppose someone reports an experience which he takes to be a perception of God commanding him to kill all those who sincerely seek to live a moral and holy life. We can be confident that God did not reveal that message and thus have a reason for thinking the experience to be delusive. But it is very doubtful that there is an adequate range of reasons for questioning such experiences. For a full discussion of this matter, see Richard M. Gale, On the Nature and Existence of God (Cambridge: Cambridge University Press, 1991) 285-343.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

397

that the belief in question is not properly basic. In these examples the persons have had no opportunity or occasion to think about or consider what positive reasons (defeating factors) there might be for questioning whether their beliefs are properly basic. Plantinga provides an example of a 14-year old theist brought up to believe theism in a community where everyone so believes. This 14-year-old theist, we may suppose, doesn't believe in God on the basis of evidence. He has never heard of the cosmological, teleological, or ontological arguments; in fact no one has ever presented him with any evidence at all. And although he has often been told about God, he doesn't take that testimony as evidence; he doesn't reason thus: everyone around here says that God loves us and cares for us; most of what everyone around here says is true; so probably that's true. Instead, he simply believes what he's taught.16 Clearly, in the situation described our 14-year-old has a basic belief in God and is rationally justified in holding that belief. For part of the stipulation of the case is that he has no reasons to think that God doesn't exist or that his community could be wrong in its religious beliefs. In fact, and this is the important point, it would be unreasonable to expect him to understand what such reasons might be.17 So it doesn't seem quite right to insist that a belief is properly basic for a person in certain circumstances only if (1) the person understands what factors would render those circumstances insufficient to make his belief rational, and (2) the person is able to find out if those factors are present. We earlier noted that one's basic belief is rationally justified (and thus properly basic) in certain circumstances provided one has no good reason to think otherwise. We then noted that one's having no good reason to think otherwise may be informed or uninformed. In the informed case, one understands what factors would defeat the prima facie justification of one's basic belief, has an idea of what things would be like if those factors were present, and has no indication that the factors are present. In the uninformed case, one either has no idea of what such factors might be or has no idea of how to go about determining whether such facts are present. Initially, I proposed that one's having no good reason to think otherwise must be informed if one's belief is to be properly basic. The case of the 14-year-old theist shows that this proposal is incorrect. For his having no good reason to think otherwise is surely uninformed. How then shall we revise our proposal? My suggestion is that we need to distinguish two different sub-classes of the uninformed case of having no good reason to think otherwise. When one has no knowledge of what factors would defeat the grounds of a basic belief, this may be due to the fact that one has never considered the question of whether such factors exist—and one is not at fault, epistemically, for failing to consider the question. Our 14-year-old theist fits in this sub-class. But we can easily imagine our 14-year-old theist going off to college and coming into contact with people from other religious traditions (Hinduism, 16

"Reason and Belief in God/' Faith and Rationality, edited by Alvin Plantinga and Nicholas Wolterstorff (Notre Dame: Indiana, University of Notre Dame Press, 1983) 33. 17 It's clear, moreover, that properly basic beliefs are fairly easy to come by, particularly when we place the believer at a tender age in a community of believers. Had our 14-year-old been raised in a community of atheists, his belief that there is no God would have been properly basic.

398

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

Buddhism, etc.) who believe quite differently, but on grounds quite similar to his. We can imagine, that is, that our young theist emerges from his state of epistemic innocence. Considering the basic beliefs of his Hindu and Buddhist counterparts, he realizes that his and their basic beliefs about the divine cannot all be true, that some of these communities of believers must hold a false belief about the divine, and begins to wonder whether the grounds for his own belief really do provide adequate rational support for that belief. Clearly, everyone in his community believed in the theistic God and taught him so to believe. But the same is true for the Hindu and Buddhist who believe in some other sort of divine being, and, he realizes, the same would be true for him had he been born and raised in a community of committed atheists and therefore held a basic belief in atheism. Suppose, then, he considers the question of whether there are any good reasons to think that the grounds on the basis of which he believes in God might not provide sufficient rational support for his belief. What factors, he asks himself, were they present, would defeat the support for his belief. Suppose, for the moment, that I am right in thinking that the result of this process of thought will not turn up a variety of factors that meet these two conditions: (1) if they are present they would defeat the rational support afforded his belief by the grounds on the basis of which he believed, and (2) he is able to find out if these factors are present. If so, then our theist's situation is this: he believes in God on the basis of being taught so to believe by the community of believers in which he is raised and he has no good reason to think otherwise, no good reason to think either that his belief is false or that the grounds on which he believed are not sufficiently indicative of the truth of his belief. But, as we've seen, his not knowing of any good reasons to think otherwise is uninformed. For he is not aware of defeating factors about which he is able to determine whether or not they are present. This is the second sub-class of the uninformed case of having no good reason to think otherwise. And our revised view is that in the second sub-class of the uniformed case of having no good reason to think otherwise, one's belief ceases to be properly basic. Of course, many theists have different, if not better, prima facie grounds for belief in God than our 14-year-old theist. For example, just as one may have an experience in which one seems to see a tree, so one may have an experience in which one seems to encounter God as forgiving, disapproving, or speaking, and may form the related basic belief about God. If the above line of argument is correct, what are we to say of such basic beliefs as God forgives me for what I've done. God disapproves of what I've done, God is speaking to me, etc.? Our judgment is that when the theist moves beyond the stage of epistemic innocence to the point of inquiring whether she has any good reasons to question the rationality her basic belief in God, she has begun a journey whose end is that her belief about God is not properly basic because it is no longer rationally justified. If the circumstances and conditions in which these basic beliefs are typically formed and accepted are such that the person who has journeyed along this questioning path is no longer rationally justified in accepting them, are her beliefs therefore irrational? This is a large question that cannot be fully answered here. But some suggestions about its answer can be given.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION

399

For purposes of argument let's say that our general knowledge is such as to make belief in God neither rational nor irrational. That is, let's suppose that the attempt to gather evidence for or against belief in God is on the whole unsuccessful.18 We have then to consider whether or not belief in God is properly basic. I have sketched an argument for the conclusion that for the questioning theist it is no longer a properly basic belief. Is it, therefore, irrational or neither rational nor irrational? I believe it would be irrational in the circumstance of having an experience which seems to be of God only if we also knew of some of the factors which tend to make the experience delusive and believed or had reason to believe that these factors were present. But if I'm right, we either do not know what these factors are or are incapable of determining whether they are present. This being so, we lack the grounds we would need in order to judge correctly that belief in God is irrational. Undoubtedly, many non-theists will reject this view. They will hold that the believer's intellectual duty in the situation I've described is to suspend judgment, not to believe that God forgives him, that God disapproves of what he's done, or that God is speaking to him. I would agree with this judgment were the circumstances such that they did not include an experience in which it seems to the person that God forgives him, that God disapproves of what he's done, or that God is speaking to him. For the theist who has moved well beyond epistemic innocence, that experience is insufficient to render his belief in God rational. But it does count for something. Moreover, it must be noted that our inability to know what sorts of factors would render the experience delusive is due in large measure to the nature of the putative object of the experience. To conclude that in this case one's intellectual duty is to suspend judgment is, I think, to ignore the positive rational support we normally accord to experience. It would be nice if in such cases we did know what factors tend to render the experience delusive and could make reasonable judgments as to whether those factors are present. We could then reasonably judge belief in God to be rational, or if not rational then irrational. But as James has taught us: "In the great boarding house of nature the cakes and the butter and the syrup seldom come out so even or leave the plates so clean."

18 Actually, the assumption that we don't have good reason to think that God doesn't exist was implicit in our earlier discussion. For if a person has good reason to think that God does not exist then he does know of a positive reason for thinking that his putative experience of God's forgiveness is probably delusive.

This page intentionally left blank

CHAPTER TWENTY-NINE

Religion within the Bounds of Naturalism: Dewey and Wieman

It is fair to say that naturalism is the dominant philosophical position in America. While it is clear that naturalism provides no room for traditional theism, it is not clear what conceptions of God it can tolerate, let alone support. And it is simply unclear whether a conception of God that is congenial to philosophical naturalism can be religiously satisfying. In this paper I aim to explore and critically evaluate two attempts within this century to find in naturalism the resources both for a viable doctrine of God and for the articulation and practice of religious faith. The two proponents of naturalism I will discuss are John Dewey (1859-1952) and Henry Nelson Wieman (1884-1975). But of these two, it is Wieman, the theologian, I will discuss at somewhat greater length. In his 1946 book, The Source of Human Good,1 Wieman explicitly endorsed the philosophy of naturalism and footnoted a volume of essays, Naturalism and the Human Spirit, which appeared in 1944.2 Among the contributors to Naturalism and the Human Spirit were such luminaries as John Dewey, Ernest Nagel, and Sidney Hook. The book aimed at explaining, defending, and further developing a distinctive philosophical position: naturalism. Before examining Dewey's religious humanism and Wieman's empirical theism, we need to consider briefly the philosophical naturalism both embraced. Naturalism A careful look at Naturalism and the Human Spirit reveals that the distinctive feature of naturalism is its thesis that there is only one reliable method for gaining factual knowledge, a method common to the developed and developing sciences. Does this thesis rule out of court any subject matter that cannot be studied by this method? In particular, does the thesis of naturalism amount to a methodological materialism, the denial of the existence of non-material beings.3 As a first step toward answering these questions, we need to distinguish two different theses

402

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 18

WILLIAM L. ROWE

T.

The only things there are are things which can be investigated by the methods of science,

and Tl. The only things about which reliable knowledge can be obtained are things which can be investigated by the methods of science. Strictly speaking, it is Tl, not T, that is the fundamental thesis of naturalism. Moreover, if we restrict naturalism to its basic thesis (Tl) naturalism does not logically preclude the existence of objects in the universe that cannot be investigated by the methods of science. If there are such things, all naturalism says about them is that no reliable knowledge about them can be achieved by us. Nor does Thesis 1 entail materialism and the rejection of religious supematuralism, here taken as the view that there exists a purely spiritual being endowed with supernatural powers? All this thesis strictly tells us is that such a being cannot be an object of genuine knowledge unless it acts in observable ways in human affairs and thus can be discerned by the methods of science. As it stands, then, the fundamental thesis does not preclude the existence of a supernatural being. But naturalists tend to hold an additional thesis. T2. The only individual things of whose existence we have reliable knowledge are physical things. This thesis implies at least agnosticism concerning the existence of a purely spiritual, supernatural being. But it does not imply that religious supernaturalism is false. To get the outright rejection of supematuralism we need the third thesis naturalists tend to hold. T3. Mental states are causally dependent on physical states. So, for our purposes, we will take naturalism to embrace Tl, T2, and T3.4 What T3 is meant to imply is that any being that has a conscious life and mental states has certain physical states on which its mental states are causally dependent. And this precludes the existence of the God of supernaturalism. We've now had a brief look at what naturalism as a philosophical position was taken to be by an important group of American philosophers, a group that included Dewey and whose writings in the early forties were admired and quoted by Wieman.5 With this background, we can now turn to an examination of the naturalistic conceptions of God developed within this framework by Dewey and Wieman.

Dewey's religious humanism In A Common Faith,6 Dewey presents his view of religions, supernaturalism, and what he regards as something of abiding value that he calls by the

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

403 19

name 'the religious'. Although he has no real interest in developing a naturalistic religion to compete with the surpematualistic religious traditions, it is clear that he thinks naturalism is fully compatible with the valuable element he calls by the name *the religious'. And in the course of developing his account of the religious, he also sets forth and defends a naturalistic view of God. So, I would say that Dewey does endeavor to develop a naturalistic conception of God and does try to provide a place within naturalism for something akin to religious faith. For this reason he may be viewed as a naturalist who advances and defends a version of religious humanism. He begins his book by nothing that orthodox Catholics, Protestants, many theists and some deists hold that anything really valuable in religion is tied to the existence of a supernatural being. In opposition to these religionists, as Dewey likes to call them, there are those who think the advance of culture and science has completely discredited the supernatural and along with it the hope that there is something really valuable in religion. But however far apart the religionists and these critics of supernaturalism are, they agree that the value of religion stands or falls with supernaturalism. And it is precisely this point that Dewey disputes. For although he thinks supernaturalism is irrelevant, if not simply false, he also thinks that there is something of great value in religion, something whose value doesn't depend at all on the existence of the supernatural. In taking this view, he sees himself as subject to attack from two sides. The religionists on the one hand will insist that by rejecting the supernatural he has cut out the vital nerve that creates and sustains whatever is valuable in religion. And many modern cultural critics of supernaturalism on the other hand will charge him with being excessively tenderminded toward religion, with making a concession rather than carrying the argument to its logical conclusion. What then is it that Dewey thinks occurs in religious life and is of such value? In answering this question he thinks it important to distinguish religion, a religion, and the religious. Noting the difficulty of finding a generally accepted definition of 'religion/ he cites for his purposes the Oxford Dictionary account of religion as the recognition of some unseen higher power having control of human destiny, a power worthy of obedience and reverence (CF, p. 3). Of course, various religions characterize the unseen power in a multitude of different ways and differ just as profoundly in the ways of expressing obedience and reverence. So, Dewey quickly concludes that there is no such thing as religion in the singular, there is only a multitude of religions. Each religion involves a special body of beliefs and practices having some kind of institutional organization. However, among these diverse religions, Dewey thinks one can detect an element that he calls the religious. In part, his aim is to emancipate this element, the religious, from its entanglement with religions. For the religious, as Dewey conceives of it,

404

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 20

WILLIAM L. ROWE

is a feature that although present in significant religious experiences can also be present in experiences that would not be generally regarded as religious experiences. What, then, are these experiences that, according to Dewey, have the feature he calls the religiousl Basically, they are unifying experiences focussed on ideals, experiences in which there occur enduring changes in ouselves that enable us to live better, deeper, fuller lives in harmony with the unchangeable forces in the universe. When stripped of their explanations in terms of supernatural causes, this is what, on Dewey's account, many significant religious experiences amount to. In these experiences we are aroused and guided by ideal ends. And it is the active pursuit of these ideals that unifies the self and brings about a better, deeper, and enduring adjustment in life. When such an experience occurs in say, the Christian faith, its subject naturally interprets it in terms of the workings of the Christian God. But, according to Dewey, experiences similar in their effect 'occur to Taoists, Buddhists, Moslems, and persons of no religion including those who reject all supernatural influence and power' (CF, p. 14). Indeed, he explicitly dissociates the question of whether an experience has the quality 'the religious' from the question of what causes the experience. As best I can determine, for Dewey an experience bears the quality 'the religious' just in case the experience is one in which a person is profoundly moved by some ideal end or ends in such a manner that the self is unified in some more or less enduring way and the person's life becomes better and more properly adjusted to the conditions and forces in the universe that affect him. Of course, not just any experience in connection with the pursuit of some ideal end will be religious in quality. One may be intellectually convinced that some end should be supreme in one's life without the end sufficiently arousing one's emotions to bring about significant changes in one's life. Moreover, the end must be sufficiently comprehensive or inclusive so as to unify the various elements in the self and produce deep and far-reaching adjustments in our relation to the universe around us. But that such experiences occur, that they are of profound value, and that they occur both in religions and outside of religions, Dewey has no doubt. Having specified the sort of experience that he takes to be of extreme value in human life, Dewey turns to the question of how we should conceive of God or the divine. He considers two views: the view that God is a particular being in which the greatest ideals are already actual in the highest degree, and the view that God is 'the unity of all ideal ends arousing us to desire and actions' (CF, p. 42). Dewey, of course, doesn't think there is anything in existence answering to the first view of the divine. Indeed, his view is that the supernatualist's God in which the greatest ideals are already actual is the result of a primitive human need to convert the object of desire, the greatest ideals, into an antecedent reality. As to historical personages

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

405 21

who have been vested with divine attributes, his view is that they are 'materializations of the ends that enlist devotion and inspire endeavor. They are symbolic of the reality of ends moving us in many forms of experience' (CF, p. 41). But quite apart from the issue of explaining the human motivation for the invention of divine beings, including the supernaturalist's God, he presents some more or less pragmatic reasons in support of the second view over the first, in support, that is, of using the term 'God' to denote not a particular being but the unity of all ideal ends that arouse us to desire and actions resulting in the unification of the self and enduring changes making for a better, fuller life. I will mention three such pragmatic reasons he gives in support of his view. 1. To the point that an all-powerful being can reward those who pursue ideal ends and punish those who do not, Dewey notes that such a view fosters unworthy motives to pursue ideal ends and claims that ideals such as justice, affection, and truth have such a hold on humanity that 'it is unnecessary for the religious attitude to encumber itself with the apparatus of dogma and doctrine' (CF, p. 44). 2. He argues that human beings have never fully used the powers they possess to advance the good in life because they have waited upon some powerful being outside of nature to do the work they are responsible for doing. 3. He points out that if we pin our hopes on a supernatural being whose existence is a subject of serious intellectual debate, we risk the loss of faith if the existence of this being is shown to be rationally doubtful; whereas, the presence of ideal ends leading and guiding human endeavor is a matter beyond intellectual dispute. Against his identification of the divine with the unity of ideal ends, Dewey considers and responds to one fundamental objection. The objection is that by rejecting the divine as a particular being capable of efficient causality in the world, the ideal is left without 'roots in existence and without support from existence' (CF, p. 47). After all, as Dewey himself makes abundantly clear, ideal ends are not substantial things existing apart from us. They are created by human imagination when we consider actual goods and imagine possibilities that are not yet actual. And whatever power they have is dependent on actual human desires to see them realized on earth. In response, Dewey claims that ideals do have their roots in nature, for they are directly linked to actual entities: human beings who through imagination generate them and by desiring them vest them with the power to move us to action. Moreover, these ideal ends are connected with all the natural forces and conditions that help promote their realization. As he remarks: We are in the presence neither of ideals completely embodied in existence nor yet of ideals that are mere rootless ideals, fantasies, Utopias. For

406

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 22

WILLIAM L. ROWE

there are forces in nature and society that generate and support the ideals. They are further unified by the action that gives them coherence and solidity. It is this active relation between ideal and actual to which I would give the name 'God' (CF, pp. 50-51). We should note here that the last sentence in this quotation appears to suggest a different use of the word 'God' from the one Dewey earlier proposed. He first told us (CF, p. 42) that on his view 'God' denotes 'the unity of all ideal ends arousing us to desire and actions'. Now, in response to the objection that ideal ends aren't substantive entities and lack sufficient connection with some actually existing things, he seems to want to expand his use of the term 'God' to include whatever in us and in nonhuman nature contributes to the unification of ideal ends and their realization on earth. Indeed, that this is what he intends is, I think, made clear by his noting that aggressive atheism has something in common with traditional supernaturalism. Both see humankind as the one thing of importance in nature. While the supernaturalists focus on the human drama of sin and redemption in relation to the God beyond nature, and see nature, apart from man, as either accursed or negligible, the attitude taking by militant atheists is often 'that of man living in an indifferent and hostile world and issuing blasts of defiance'. Dewey immediately says:'Use of the words "God" or "divine" to convey the union of actual with ideal may protect man from a sense of isolation and from consequent despair or defiance' (CF, p. 53). Clearly, his point here is that by using the word 'God' to stand, in part, for certain natural forces that support our ideal ends we can diminish the sense that nature is indifferent and hostile to mankind. Moreover, Dewey then hastens to add that his use of 'God' does not encourage reverence for 'everything in general', pointing out that it selects just those factors and forces in existence that generate and support our idea of good as an end to be striven for. For, as he notes, a multitude of factors in existence are either hostile or irrelevant to the generation and realization of ideal ends. Again, then, it seems clear that Dewey here means to include certain natural forces as, in part, what he means by his use of 'God'. So it appears that Dewey gives us two ideas of God, not just one. His first idea, (CF, p. 42), is that God is 'the unity of all ideas ends arousing us to desire and actions'. On this view, ideals such as justice, friendship, etc. are literally a part of God. His second idea (CF, p. 51), is that God includes not just the ideal ends but also whatever factors and forces in ourselves and non-human nature that contribute to the unification of ideal ends and their realization on earth. Before turning to Wieman's effort to forge a naturalistic conception of God, I want to make a few general remarks, all related to the two ideas of God that we have uncovered in Dewey's A Common Faith. For these two

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

407 23

ideas suggest different directions open to the naturalist who seeks to develop a naturalistic idea of the divine. The first direction, locating the divine - as Dewey's first idea of God does - in the ideal ends generated by human imagination, ends whose pursuit by us results in growth of human good in the world, fits well with the spirit of humanism and the celebration of humankind's achievements in the use of intelligence and imagination. For when all is said and done, on Dewey's first account God turns out to be the product of human genius, the result of our ability to imagine possibilities whose active pursuit by us may bring about the kingdom of heaven on earth. Thus, it seems fitting to think of a naturalist who advocates a view of the divine in terms of humanly generated ideal ends as an advocate of religious humanism. Religious humanism, however, may branch out in either of two directions. On the one hand one may see the world of non-human nature as indifferent or hostile to human flourishing. Here human beings are seen as struggling against the forces of nature for a place in the sun and the opportunity to flourish through making changes in the conditions in which we live. My suggestion is that Dewey's first idea of God is quite compatible with a sort of humanism that pits humankind against nature, sees human achievements as victories over nature, sees the growth of human good through the pursuit of ideals as a uniquely human doing in spite of the opposition of an ultimately unfriendly universe. The reason his first idea of God is compatible with this we-against-nature-sort-of-humanism is that on this idea the divine itself is a human achievement, the divine itself is the product of human intelligence and imagination. There is, on the other hand, another direction in which religious humanism may move, a direction in which humankind is seen as continuous with nature and supported by forces in nature that are beyond our manipulation and control. On this vision of things, the divine is more likely to be seen as deeply embedded in the natural world, as something humankind does not create, as something whose sustaining power makes possible whatever we may achieve or become, as something that will itself bring about the growth of human good and the kingdom of heaven on earth if we but adjust ourselves rightly to it. Pushed to its extreme, this view of the divine sees God as a process or force in nature that is not the product of human intelligence and imagination-precisely the opposite is the case, the development of human intelligence and imagination is the result of those forces and processes in nature that rightly deserve to be called God. When we turn to Dewey's second idea of God, an idea which appears to include 'forces in nature and society that generate and support the ideals', as well as the ideal ends themselves, we have a small step being taken in the direction of a religious humanism that finds the divine in nature itself, as

408

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 24

WILLIAM L. ROWE

opposed to seeing the divine as a human creation. When Wieman read Dewey's A Common Faith and reviewed it in November of 1934 for The Christian Century; he fastened on the passages in which Dewey appears to move beyond identifying the divine solely with ideals generated by human imagination, claiming to find a movement in Dewey's thought toward a view of God as superhuman without implying either that God is supernatural or personal. In short, not surprisingly, Wieman took Dewey to be moving in the direction of his own view of God as a real force in the world of nature to which humankind must give loyalty and devotion if we are to be saved from evil and led to the best that human life can ever reach. While I think Dewey's religious humanism clearly develops in the second of the two directions I've outlined, a direction that sees human life as continuous with non-human nature and supported by various natural forces, I think it was a mistake on Wieman's part to interpret Dewey as moving significantly toward his own conception of God. Dewey's most persistent view in A Common Faith is to identify the divine with ideal ends of human conduct. And even if he does broaden his idea of God to include natural forces within and without us that provide support for our highest ideals, it is still true that God is, at least in part, the product of human thought and imagination. One final comment. I have used the expression 'religious humanism', and the expression 'empirical theism', taking Dewey's position to be a version of the former and Wieman's position to be a version of the latter. But shouldn't Dewey also be thought of as a theist? Perhaps he should. But I have not described him as a theist for two reasons. First, he does not himself profess any strong attachment to using the word 'God' for the ideal ends arousing us to desire and action. But my main reason is that Dewey's God is, as I said, a human creation, the result of the play of human imagination in creating possibilities that extend beyond the realities of the present world. Any such view, I think, is more aptly described as humanism. And, since he holds that experiences in which the self is unified in the pursuit of ideal ends are essentially religious, it seems proper to describe his view in A Common Faith as religious humanism. However, if someone insists on describing his view as a version of theism, I suppose the label "humanistic theism" would not be inappropriate. Wieman's empirical theism In his intellectual autobiography Wieman writes: My intellectual life has been focused on a single problem. Every significant influence which has played upon me has been directed to this inquiry. The problem which has engaged me for the past fifty years can be put in the form of a question: What operates in human life with such

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

409 25

character and power that it will transform man as he cannot transform himself, saving him from evil and leading him to the best that human life can ever reach, provided he meet the required conditions?7 Since Wieman identifies God with whatever does operate in human life with such character and power that it will save us from evil and lead us to the best, we must note at the outset that in posing the question as he does Wieman begs two important questions: Ql. Is there anything in the universe that operates in this way - the question about God's existence? and Q2. Is what operates in this way in the universe one or many distinct things or processes - the question about the unity of God? Does Wieman have any credible responses to these two objections? My own view is that he does have a reasonable response to the first objection, but I am doubtful that he has an adequate response to the second. In discussing the question 'Does God exist?' Wieman explains that he proposes to 'formulate the idea of God so that the question of God's existence becomes a dead issue'. It will be a matter of common knowledge that God exists. The remaining question will be the discovery of important truths about God, truths that will serve to guide us in relating ourselves properly to him. And in his first book, Religious Experience and Scientific Method, published in 1926, he starts by saying this: Whatever else the word God may mean, it is a term used to designate that Something upon which human life is most dependent for its security, welfare and increasing abundance. That there is such a Something cannot be doubted. The mere fact that human life happens, and continues to happen, proves that this Something, however unknown, does certainly exist.8 Suppose I were to say to you that what I mean by 'God' is whatever has brought it about that human beings exist. Would you protest if I then said that the question for empirical inquiry is to find out what it is that has this character? Would you say that I have begged a very important question here: the question of whether there is any thing or things or processes that have resulted in the existence of human beings? Well, I suspect you would say that only if you had been tainted by bad philosophy and thought we must first prove that human beings haven't just popped into existence. We all know that there are conditions and forces that generate and support human life. The only question at issue would be a more precise understanding of those conditions and forces. Similarly for Wieman. He proposes a formal characterization of God such that given that characterization it isn't an issue of whether God exists - it is a matter of common knowledge that God, so characterized, exists. The question is not whether God exists but what empirical inquiry can tell us about God. So, I think Wieman's failure

410

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 26

WILLIAM L. ROWE

to raise the issue of whether God exists is justified given his formal definition of 'God'. This is not to say that Wieman is justified in speaking of God as a unitary being or process. Nor is it to say that Wieman's use of the term *God' is at all appropriate. All I have tried to do here is to rescue him from the charge that he has committed a serious blunder in begging the question of the existence of God, a question that he should have answered by applying the empirical method. Still, it may strike us as odd that Wieman should define the term 'God' in such a way that the existence of God is not a matter of debate. Here I can only speculate. I think Wieman never once questioned that most religions are concerned with something of extreme importance in human life, something that really exists and creates the greatest good we are able to achieve. Perhaps he was inconsistent not to question this fundamental conviction. Perhaps he should have subjected that conviction to empirical inquiry. But, so far as I can tell, Wieman never questioned it. He does think that religions have inadequately characterized that something. So, given his empiricism, Wieman sought to characterize that something so that it can be investigated, if not by the refined methods of the developed sciences, at least by the experimental methods of common sense. Thus he proposed a definition of 'God' such that (1) God's existence is a common sense fact, and (2) the nature and activity of God may be investigated empirically. In his perceptive review of Wieman's discussion of God's existence, John Dewey agreed with Wieman's rejection of supematuralism but faulted him for using the word 'God' for the forces and conditions in nature that generate and support human good. For such a use suggests that these forces and conditions constitute some sort of unitary thing or process. 'But', Dewey says, 'it will be plain in any case that the word is used simply to designate a multitude of factors and forces which are brought together simply with respect to their coincidence in producing one undesigned effect - the furtherance of good in human life'.9 When we begin to list all the things and processes in the universe upon which the emergence of human life and the realization of human good are dependent, it seems quite obvious that we are dealing with a multitude of objects, states, and processes, and not some single thing or process. For this reason I see little hope at all for the Wieman of the early thirties to succeed in answering the question about the unity of God.10 As Wieman proceeded to develop his views in the forties and later, however, he does seem to introduce changes that diminish the severity of this problem. For as his thought progressed he became less interested in speculating about cosmic processes that are involved in generating and sustaining human life and more interested in what process is responsible for the emergence and growth of human good in already existing human communities. In The Source of Human Good he identified this process as 'creative good', a process more or less

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

411 27

limited to something that goes on among human beings when they relate to one another through linguistic communication in certain characteristic ways. The expression 'creative good* later gave way to the more narrowly focussed expression 'creative interchange'. These changes in his thought have implications not only for the 'size' of Wieman's God but also for the question of unity we've been discussing. For as his thought progressed he narrowed the scope of what would be designated by the term 'God' and thereby did away with some of the multiplicity that seemed to be captured by his earlier characterizations of God. After Wieman replied to Dewey's critical remarks about his view of the unity of God, Dewey wrote a response, perhaps naively hoping to close the discussion. In his response, Dewey gives up pressing the question of the unity of Wieman's God, noting that if Wieman or anyone else gets contentment from calling these forces and conditions God, that is fine with him. But he then indicates that what troubles him most about Wieman's view is his declaration that God is the greatest good, and thus deserves our complete devotion, love, and commitment. Here is what Dewey says: The more important thing which gives me hesitation is Mr. Wieman's shift between God as the power which makes for the greatest good and God as the Greatest Good let me point out that that which makes for good, whether it be singular or collective, demands care, attention, watchfulness, from man in order that it (or they) may do its (their) work most effectively and surely. But there is nothing about it or them particularly to demand love and adoration. I can love a healthy person..., and I can feel a grateful affection for the person who brings me greater health. I do not know just what love (or adoration) of the chemico-physical processes that basically determine health would be, though of course I understand the need of thought-ful cooperation with their workings.11 These remarks appeared in The Christian Century on 22 March, 1933. Wieman's response appeared in the April 5 issue of that periodical. The remarks I just quoted from Dewey struck at the very foundation of Wieman's thinking, and Wieman was quick to make it clear how fundamental he takes this issue to be. I quote him at some length. I believe the supreme issue, not only between Mr. Dewey and me, but between non-theism and theism, and indeed, in the whole world of present day thought, is expressed by Mr. Dewey in the following words: '...that which makes for good, whether it be singular or collective, demands care, attention, watchfulness, from man in order that it may do its work most effectively and surely. But there is nothing about it or them particularly to demand love and adoration.' (Italics mine.) If there is nothing about that which makes for greatest good to demand love and adoration, then certainly the word God should not be applied to it. But if there is something about it to demand love and adoration, fiir-

412

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 28

WILLIAM L. ROWE

thermore if there is that about it to demand our greatest love and adoration because it includes all that is truly most precious, then certainly there is no word in the languages of mankind which will indicate its significance save the word God. That is to say, if it demands our highest love, service and adoration, then it demands of us those behavior patterns which have characteristically been directed to what has borne the name of God - the behavior patterns of supreme love, adoration, loyalty, selfgiving, dedication of self and all that pertains to self. Here, then, is the basic issue. Is that which makes for greatest good to be compared to chemico-physical processes which make for health, and so to be tended but not loved, used but not served? Or is that which makes for greatest good the actual present growing good itself, to be tended but also loved, served but also adored?12 There are two issues at stake here. The first, and less important issue, is whether it is proper for Wieman to use the word 'God' to designate the objective force or forces that make for human good. The second, and absolutely basic issue for Wieman, is whether it is right and proper for us to commit ourselves with love and devotion to that reality, whether one or many, that makes for human good. Or, to follow Dewey, is it only proper for us to take note of these forces, to tend to them and make use of them, rather than to respond to them with commitment and love. Of course, these two issues are closely related. For as Wieman notes, only if it is proper to respond to the reality that makes for human good with love, devotion, and commitment is it appropriate to call that reality God. To answer our basic question of whether Wieman is right or wrong in thinking that what makes for the greatest good is the appropriate object of our deepest love, devotion, and commitment we need to look first at what he takes human good to be, and second at what he thinks is particularly important about the creative process that makes for our greatest good. In The Source of Human Good Wieman turns his attention to discussing the nature of good, particularly human good and whatever is productive of it. He begins by distinguishing between intrinsic good and instrumental good. Although many goods are good in most circumstances, they are not always good. Since we can imagine situations in which they are not good, he concludes that 'they do not carry in themselves, intrinsically and essentially, the nature of goodness. Something over and above their bare existence must pertain to them to make them truly good' (SHG, p. 10). From this and other remarks we can conclude that Wieman takes an intrinsic good to be something that is good 'necessarily and always'. Although he tries out a somewhat different theory in The Source of Human Good, his most persistent answer to the question of what is intrinsically good is the answer he learned from R.B. Perry; namely, the satisfaction of interest. So, for

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

413 29

Wieman, the satisfaction of any human interest, drive, or desire is something that is intrinsically good. Of course, it is not always a good thing for an interest to be satisfied, for the satisfaction of certain interests may preclude or obstruct the satisfaction of other interests. Moreover, despite his empiricism, Wieman always believed that in addition to conscious interests or desires, human beings have many unconscious desires. So, even if we could bring about the fulfillment of all our conscious desires, we may fall short of the greatest good possible for us. For the satisfaction of various conscious desires may preclude or obstruct the satisfaction of important unconscious drives, drives that then manifest themselves 'in fits of melancholy, in various forms of anxiety, and in self-defeating behavior'.13 In 1958 Wieman proposed the following account of the greatest good for a human being: 'It is the most complete satisfaction of the individual that is possible when the individual is viewed in the wholeness of his being' (MUC, p. 98). Presumably, the expression 'in the wholeness of his being' is meant to select out those conscious and unconscious desires whose fulfillment provides deep and lasting satisfaction that harmonizes well with the satisfaction of other desires the individual has or might come to have. It is, Wieman thinks, a matter for empirical inquiry to discover what objects of desire do satisfy man 'in the wholeness of his being'. The common notion of an 'instrumental good' is something that is causally productive of whatever is intrinsically good. Some things that are instrumentally good are also intrinsically good. Some things that are instrumentally good are intrinsically neutral or even intrinsically bad. Wieman mentions eating tasteless or nauseating food as an example. Eating tasteless food is, presumably, neither satisfying nor dissatisfying; whereas, eating nauseating food is dissatisfying. But the eating of such food may provide the nourishment necessary for the satisfaction of other desires, thus contributing causally to the realization of intrinsic good. However vague the notion may be, suppose we grant that the satisfaction of a person in the wholeness of her being is the greatest good, at least so far as human beings are concerned. What is it that brings this about? What is it, in Dewey's words, that makes for human good? Wieman's mature answer is that it is a process that occurs in human communities, a process he first called 'the creative event', and later called 'creative interchange'. Before looking at his account of this process, however, we need to note one more point concerning the greatest good at the human level. The greatest good is not itself a realizable state, it is not the end point of a process of change we go through. It's not like getting your Ph.D., something that occurs at a point in time as the result of a process you go through. One never reaches a point where it is true to say that the greatest good has been achieved. The goal is rather to be engaged in a continuing process in

414

WILLIAM L.ROWE ON PHILOSOPHY OF RELIGION 30

WILLIAM L. ROWE

which the individual's deepest interests are being progressively satisfied while new interests are being developed and are themselves about to be in process of being satisfied. So the greatest good for a human being is to be continuously engaged in such a process. Perhaps it will be helpful if we think of the greatest good at the human level as itself an ongoing process: the progressively growing satisfaction of interests in the human individual. Before we see how Wieman characterizes the creative process that progressively yields human good, I want to introduce a word of caution. Unless we are careful, we will end up equating the process that generates the greatest human good with the process that is the greatest human good. For, as I see it, we have here two intimately related processes, not just one process. On the one hand we have the greatest human good, the progressively growing satisfaction of interests in the human individual. On the other hand we have the creative event or creative interchange which is the process that is generating the process that is the progressively growing satisfaction of interests in the human individual. The generating process Wieman calls ereative good, the generated process he calls created good. As I indicated earlier, in The Source of Human Good Wieman presents a different view of what is intrinsically good. He doesn't deny that satisfaction of interest is intrinsically good, but he does introduce the notion of qualitative meaning and suggests that it is also intrinsically good. The reason he does this, I suspect, is because he can more easily characterize the creative event in terms of qualitative meaning than in terms of satisfaction of interest. In any case, he explains that a present event takes on qualitative meaning when it becomes so connected with other past and future events that the felt qualities of those events are experienced in the present event. He gives as an example the hearing of footsteps, a sign that someone is approaching. After hearing those same footsteps for a number of years, the mere hearing of them may arouse feelings associated with that person. As Wieman puts it: * Something of the quality of many events occurring in the past is packed bodily into the present when I hear those footsteps. I even feel by anticipation something of the quality of emerging future events, when foreseen possibilities have been actualized by association with the person now approaching' (SHG, p. 18). Qualitative meaning, then, 'is that connection between events whereby present happenings enable me to feel not only the quality intrinsic to the events now occurring but also the qualities of many other events that are related to them' (SHG, p. 19). You might ask why Wieman thinks qualitative meaning is intrinsically good and how it is related to the satisfaction of interest or desire? I think these are good questions to which he doesn't give convincing answers. He says that qualitative meaning satisfies human want in that it is one ingredient in every instance of satisfaction. This may well be true. But it would

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

415 31

also seem to be true that qualitative meaning is an ingredient in instances of dissatisfaction as well. So, while qualitative meaning may be necessary for satisfaction of human want, it is not sufficient. But our primary interest here is to learn about the creative event, that process that leads to "abundant satisfaction and qualitative meaning'. Having acquainted ourselves with his notion of qualitative meaning, we can now turn to his characterization of creative good, the creative event. Wieman describes the creative event as composed of four sub-events. Each of the sub-events is necessary and together they are sufficient for an event to be a creative event. The four sub-events are: (1) the emerging awareness of qualitative meaning derived from other persons through communication, (2) integrating these new meanings with others previously acquired, (3) expanding the richness of quality in the appreciable world by enlarging its meaning, and (4) deepening the community among those who participate in this total creative event of intercommunication. In the first sub-event one person communicates her feelings and thoughts freely and fully to another who appreciatively grasps the feelings and thoughts being communicated. In the second sub-event these feelings and thoughts that have been appreciatively grasped are integrated into one's own thinking and feeling. This integrating, he explains, 'is largely subconscious, unplanned and uncontrolled by the individual, save only as he may provide conditions favorable to its occurrence' (SHG, p. 59). It is presumably this stage of the process that results in new desires and interests that cannot be predicted in advance. The third sub-event supposedly follows from the first two. One sees one's world differently. Indeed, one's appreciable world is enhanced and transformed. Things that were of no interest may now be seen to be immensely valuable. Other things that were valued before may no longer be valued at all. Finally, the fourth sub-event in the creative event is constituted by a widening and deepening community among those who have participated in the first sub-event. I now respond with deeper understanding to those who have revealed their deeper selves to me. This does not mean that we now constitute one big happy family, or even a mutual admiration society. For, as Wieman remarks, a person 'may disapprove, deny, and repudiate much that has been communicated to him from them, but this is a form of understanding and community' (SHG, p. 64). Well, there you have it - a sketch of the creative event at the level of human beings who are capable of communicating linguistically with one another. According to Wieman, it is this event occurring in our lives that leads to 'the most complete satisfaction of the individual that is possible when the individual is viewed in the wholeness of his being', provided the

416

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 32

WILLIAM L. ROWE

person meets the required conditions. The basic human problem, therefore, is to shape all human conduct and all other conditions so that the creative event can be released to produce maximum good. But here we cannot pursue these matters. Instead, we need to turn to the question we took from Wieman's earlier discussion with Dewey. Is Wieman right in holding that creative good is an appropriate object of our deepest love, devotion, and commitment? Let's concede at the outset that the creative event, as characterized by Wieman, is instrumentally good, that it leads to the satisfaction of human interest. That alone is insufficient to make it an appropriate object of devotion and commitment. After all, the process of digestion is also an instrumental good. So, if we are to conclude that it is appropriate and proper to give love, devotion, and total commitment to the process Wieman has described, we need more to go on than the mere fact that it is causally productive of the ongoing process he calls 'the greatest good'. One possibility we need to consider is Wieman's view that something that is instrumentally good may become, at least on some occasions, intrinsically good. He provides the following example: ... when I chop wood to sustain that other structure of happenings called 'the life of my family in our home', the values of the activity may be purely instrumental if the qualities pertaining to life in my home cannot freely enter conscious awareness as I chop. However, if bonds of meaning are developed between my chopping of the wood and the life of my home, so that the lives of the children and the affection of the wife are vivified in conscious awareness by the very act itself, then the activity ceases to be merely instrumental. Then chopping the wood has taken on those qualities pertaining to the total structure of events called 'the life of my home'. It is an intrinsic good, no matter how fatiguing it may be (SHG, pp. 54-55). The suggestion in this passage is that an activity that is instrumentally good may on occasion actually become intrinsically good. So, although on some occasions chopping wood may be merely instrumentally good, on other occasions that very same activity may also be intrinsically good. Perhaps the same might be true of the creative event. But I think Wieman's discussion here is a bit confused. Although there is nothing wrong with Wieman's example, there is, I think, something wrong with the conclusion he draws from it. For it is inconsistent with his account of intrinsic good. Something is intrinsically good only if it is necessarily and always good. If this is so, then chopping wood can never itself be intrinsically good. I believe the correct conclusion to draw from his example is that the activity of chopping wood may occur as a necessary part of something that is intrinsically good. The something that is intrinsically good is chopping wood while enjoying

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

417 33

the satisfaction of providing for loved ones. This is something that is necessarily and always good. So, unless the creative event is necessarily and always good, it cannot ever be itself intrinsically good. Given Wieman's account, it seems right to count the creative event as a very important instrumental good. For when the proper conditions are present it results in the greatest good humans can achieve. But Wieman balks at equating the creative event with an instrumental good. For the notion of 'an instrument' suggests something that we may control and use to accomplish our own goals and ends. His most persistent theme about the creative event, however, is that while we may be sure it will lead to the most complete satisfaction of the individual that is possible when the individual is viewed in the wholeness of his being, it is not a process we can control or use to realize our values. For it is the nature of this process to transform us, producing in us values that we never intended. As he puts it: The creative event cannot be used to shape the world closer to the heart's desire because it transforms the heart's desire so that one wants something very different from what one desired in the beginning. (SHG, p. 57) So long as we define an instrumental good in terms of its bringing about something that is intrinsically good, we can acknowledge that the creative event is an instrumental good even though we may not be able to use it to bring about the things we presently value. And since the creative event can do its work in our midst without our being conscious of its own features, it would hardly seem to qualify as an intrinsic good. Moreover, any tendency we have to view the creative event as itself intrinsically good is likely due to our failure to adequately distinguish the two processes I mentioned earlier: the generating process and the process that is generated - only the latter, the progressively growing satisfaction of interests in the human individual, is intrinsically good. Perhaps realizing these points, in The Source of Human good Wieman introduces the notion of * Absolute Good' and argues that the creative event is the only absolute good (SHG, pp. 79-81). I'm not going to discuss his account of the creative event as an absolute good other than to say that an absolute good seems to be an important sort of instrumental good. By this I mean that many things are instrumentally good only in certain circumstances and conditions. In other circumstances and conditions they are either instrumentally neutral or instrumentally bad. Wieman's main point about absolute good is that it is good 'under all conditions and circumstances'. I take this to mean that in the world as we know it there are no conditions or circumstances in which the creative event occurs and is not productive of intrinsic good. If this is so, the creative event is an extremely important instrumental good. It is time now to conclude this essay by making some judgments about Wieman's and Dewey's efforts to forge a viable doctrine of God within nat-

418

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 34

WILLIAM L. ROWE

uralism. It should be obvious that any effort will lose much that has been associated with the idea of God in supernaturalism. The most obvious loss is that the naturalist's God is not a person. Persons can be objects of love, devotion, worship, and commitment. Moreover, persons are capable of exercising creative power. And these two features - possessing creative power and commanding devotion and commitment - are intimately connected to the standard use of the religious notion of God. Suppose, then, that we consider how Dewey's God and Wieman's God fare with respect to these two features. To put the point as briefly as possible: Dewey's God can command devotion and commitment, but appears to lack creative power; whereas, Wieman's God possesses creative power, but has some difficulty, I believe, in commanding devotion and commitment. But between the two, I suspect Wieman's God comes closer to accommodating both features than does Dewey's God. Consider first the feature of possessing creative power. Ideal ends are ultimately our creation. Without the exercise of human imagination, no such ideal ends would be created by us. Moreover, the extent to which they have the power to move us to action is only the extent to which we desire their realization. They are incapable of efficient causality. Wieman's God, however, is a force or process that exists independently of human desire and intent. Of course, we can obstruct it, even prevent its occurrence, at least in terms of how it works at the human level through communication with others. But it works the way it does, and cannot be bent to serve our own ends. Its power over us is limited, unlike the power of the supernaturalist's God. But in terms of the feature of creative power, there can be little question that Wieman's God accommodates that feature in significant ways and degrees that Dewey's God does not. In terms of commanding love, devotion, and commitment, however, it is clear that ideal ends are appropriate objects of such attitudes. One can love and be devoted to an ideal almost as easily as one can love and be devoted to a person. But a process or force existing apart from human desire and intent is normally not the sort of thing that it makes sense to love or be devoted to. There are, of course, natural processes that are supportive of human life and other processes that are destructive of it. There are things we can do to support processes of the first sort and things we can do to defeat or delay processes of the second sort. But this is a far cry from loving lifesupporting processes and hating life-threatening processes. The question is whether there is something special about the creative event that makes it appropriate to respond to it with love, devotion, and commitment. As we've seen, Wieman holds that there are two things of special importance about the creative event. First, he argues that it is an absolute good. But the best

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGION WITHIN THE BOUNDS OF NATURALISM

419 35

I've been able to make of this is that it is a particularly important instrumental good. Second, he argues that this creative process can generate created good only if we give ourselves over to it, as opposed to tending to it and trying to use it for our own ends. I think he may be right about this. There may be processes that will produce beneficial changes in us and lead us to more satisfying lives only if we yield to them and allow them to work in us rather than resisting them or trying to manipulate them. But I do not see this as amounting to love or religious devotion. Perhaps, however, the idea of commitment is not inappropriate here. You may conclude that these two attempts to reconstruct the idea of God within naturalism yield gods that are pale, weak, and uninspiring when compared to the robust, powerful gods of supernaturalism. Dewey's God is a human creation, and Wieman's God seems to be either a plurality of natural forces and conditions or a more or less unitary process that has appeared on the scene as an operating force only with the development of beings capable of significant linguistic communication. But if naturalism is true, then if we want to view something in nature or human life as divine, something like what Dewey or Wieman proposes may as good we can do.

Notes 1. The Source of Human Good (Chicago: University of Chicago Press, 1946). Further references to this work will appear in the text as SHG. 2. Naturalism and the Human Spirit, ed. Yervant Krikorian (New York: Columbia University Press, 1944). 3. For an illuminating response to this charge and others see 'Are naturalists materialists?' by John Dewey, Sidney Hook and Ernest Nagel, The Journal of Philosophy, Vol. 42 (1945), 515-530. 4. Although I've suggested that naturalism is to be identified with the conjunction of the three theses, Tl, T2, T3, it should be clear that of the three Tl is fundamental whereas T2 and T3 are derivative. It is by virtue of applying the methods specified by Tl that naturalists propose to justify the claims expressed by T2 and T3. 5. Wieman's commitment to naturalism's fundamental thesis is evident from his earliest writings. For example, in his first book, Religious Experience and Scientific Method (New York: Macmillan, 1926), 23, he says: All knowledge must depend ultimately upon science, for science is nothing else than the refined process of knowing. Scientific method is simply the method of knowing. ... The knowledge of God must be ultimately subjected to scientific method. 6. A Common Faith (New Haven: Yale University Press, 1934). References to this work will appear in the text as CF. 7. The Empirical Theology of Henry Nelson Wieman (Carbondale: Southern Illinois University Press, 1963), p. 33. Further references to this work will appear in the text as ET. 8. Religious Experience and Scientific Method, 9. 9. The Christian Century, Feb. 8,1933,196.

420

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 36

WILLIAM L. ROWE

10. For Wieman's response to Dewey's criticism concerning the unity of Wieman's God see The Christian Century, March 1,1933,299-300. 11. The Christian Century, March 22,1933,395. 12. The Christian Century, April 5,1933,466-467. 13. Man's Ultimate Commitment (Carbondale: Southern Illinois University Press, 1958) 98-99. Further references to this work will appear in the text as MUC.

CHAPTER THIRTY

Religious Pluralism

Abstract. According to religious pluralism, the profound differences among the chief objects of adoration in the great religious traditions are largely due to the different ways in which a single transcendent reality is experienced and conceived in human life. The most prominent developer and defender of religious pluralism in the twentieth century is John Hick. Hick uses the expression 'the Real' to designate the transcendent reality 'authentically experienced' as the different gods and impersonal absolutes worshipped in the major religious traditions. A central claim Hick makes is that, apart from some purely formal characteristics, the Real is ineffable in that the intrinsic properties making up its nature are beyond the scope of any human concepts. I explore this central claim and argue that it implies the dubious, if not incoherent, view that the Real in itself has neither one of many pairs of contradictory properties.

The path to religious pluralism starts with the now commonplace fact that our world contains a number of religious faiths having different ideas as to the nature of the divine reality - for example, whether the divine is personal or impersonal - and different ideas as to whether our destiny, if all goes well, is to live forever in the personal presence of the divine or to lose our personal identity in becoming one with the divine. Clearly not all the different views of the divine can be true. And just as clearly not all the different views of human destiny can be true. If one of these conflicting religious faiths, Islam say, happens to be the truth, then a certain view of the divine and a certain view of human destiny will be true.1 But what then becomes of the faithful members of the other religious traditions, not to mention those who are faithful members of no recognizable religious tradition? Putting aside as too implausible the suggestion that the gods of the other religions are really Allah in disguise, so that devoted Christians, Jews, Hindus etc., are all, however unwittingly, faithful followers of Allah, two main answers are possible. The first is that they (the non-Muslims) are forever shut out of the eternal reward provided for the faithful followers of Islam. This answer, naturally enough, is called exclusivism. The second is that they too may join in the eternal reward enjoyed by faithful Muslims, because they will have a chance in the next life 1 This is an oversimplification. Within each major religious tradition there may be several different conceptions of the divine, as well as different conceptions of human destiny. But typically there will be a historically dominant view.

422

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION I4O

W I L L I A M L. R O W E

to respond favourably to Allah, or Allah may favour them simply because they did their best in this life to follow whatever lights they had to follow. This answer, naturally enough, is called inclusivism^ although there may be some doubt about the eternal destiny of those who go to their graves explicitly rejecting Allah, or those who fail to live well according to whatever lights they were provided with. A further step along the path to religious pluralism occurs when we note, if we do, that each of the major religious traditions seems equally successful in producing, at least now and then, persons of great moral and religious sanctity. Moreover, if some neutral person were to examine the evidence in support of each of these conflicting religious traditions, she might well be unable to determine that one has any better truth conducive reasons in its support than another, judging instead that each is about as epistemically well off as the other.3 And neither of these points is quite what one would expect if one of these religious traditions is basically true in its principal claims, and the others are all false, at least in so far as they disagree with these principal claims. One would normally expect the one true religion to be more productive of moral and religious saints. And one would normally expect the one true religion to have stronger available evidence in its support than do the false religions. If the major religious traditions hold fundamentally conflicting views about the nature of the divine and/or our human destiny, if these different traditions are more or less equally successful in producing human beings who are moral and religious saints, and if the evidence supporting the truth claims of these different traditions is such than no one tradition emerges as distinctly superior, then one of three possibilities would seem to be true. First, all of these religious traditions are false. The ideas of the divine found among them are all illusions, as Freud maintained, creations of the human mind to help us feel safe in a world that far too often appears hostile to our deepest needs. Second, exactly one of these religious traditions happens to be true in its most basic claims, even though to the enquiring outside observer it appears to be no closer to the truth than any other major religious tradition. (Of course, 2 The distinction Pve drawn between exclusivism and inclusivism is due to John Hick. According to Hick, exclusivists and inclusivists agree in holding that a particular one of the world's religions (their own) is the one true religion. But exclusivists hold that apart from the faithful followers of that one true religion there is no salvation, while inclusivists allow that salvation is extended to many who are not followers of the one true religion. It is possible, I think, to believe that one particular religion is basically true without taking a position on whether salvation is or is not extended to many who are not followers of that religion. I suspect that some who call themselves 'exclusivists' hold such a view. For example, I think Alvin Plantinga uses 'exclusivism' to describe a position that is compatible with either exclusivism (Hick's sense) or inclusivism (Hick's sense). See Plantinga's 'A defence of religious exclusivism' in Thomas D. Senor (ed.) The Rationality of Belief and the Plurality of Faith (Ithaca, NY: Cornell University Press, 1995), 191-215. 3 It is unclear how telling this point is. Presumably, the neutral observer has not enjoyed the sorts of experiences that the Muslim or the Christian has, experiences that they may reasonably take as evidence for the existence of Allah, or the heavenly father of Christ.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGIOUS PLURALISM

423 141

the faithful in each of the conflicting religious traditions will be convinced that it is their own religious faith that is the one true faith.) And finally, it is just possible that each of these religious traditions is a valid encounter with a reality that transcends every religious tradition because it transcends all human efforts to conceptualize it or to experience it directly. On this view the different religious traditions include experiences of one and the same transcendent reality. But because this transcendent reality cannot be experienced directly, as it is in itself, and because it transcends all human conceptualization, it can be experienced only as, say, the Heavenly Father of Christ in Christianity, as the God of the Torah in Judaism, as the nonpersonal, infinite Brahman in Hinduism, and so on with the other great religious faiths. The profound differences among the religious traditions are due to the different ways in which the transcendent reality is experienced and conceived in human life. It is this last view that we have come to know as religious pluralism. And its most well known developer and defender among contemporary philosophers of religion, is Professor John Hick.4 So, it is to his view of the matter that I now turn. ii

Hick uses the expression 'the Real' to designate the transcendent reality 'authentically experienced in terms of different sets of human concepts', as the different gods and impersonal absolutes belonging to the major religious traditions.5 And in developing his theory of religious pluralism he relies rather heavily on the Kantian distinction between the Real as it is in itself (the noumenal world) and the Real as it is experienced and conceived by us (the phenomenal world). While this reliance is understandable, it may also be misleading. For it suggests not only what Hick intends, that the Real itself, although beyond human concepts and direct experience, is, nevertheless, experienced through the different divine phenomenal realities met with in different religious faiths, but also, what he may not intend, that just as the phenomenal objects in Kant's philosophy are existing entities (cabbages, stones, etc.), so too the phenomenal objects through which the Real is manifested in various religious traditions (Jahweh, Allah, the Holy Trinity, Shiva, Brahman, the Tao, etc.) are themselves actually existing beings or realities. Thus George Mavrodes was led to view Hick as an exponent of polytheism.6 As Mavrodes remarks in response to Hick's reply to his original piece in which he suggested that Hick may be ' the most important Western philosophical defender of polytheism': 4 Among Hick's many works on this subject see An Interpretation of Religion (London: Macmillan, 1989), 5 and The Rainbow of Faiths (London: SCM Press, 1995). Hick The Rainbow of Faiths, 25. 6 George I. Mavrodes 'Polytheism' in Senor (ed.) The Rationality of Belief and the Plurality of Faith, 261-286.

424

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 142

W I L L I A M L. ROWE

In reading An Interpretation of Religion I got the impression that Hick thought that Allah, the Holy Trinity, Shiva, etc., were the gods worshipped in some various religions. And in some other religions the roughly corresponding objects of adoration were 'impersonal ultimates' e.g., Brahman, the Tao, etc. I also got the impression that all these were distinct from one another. And finally, I had the impression that (on Hick's view) all of these were real beings. That is what led me to the conclusion that Hick was really a serious (descriptive) polytheist.7

In the light of Hick's response to his original piece, Mavrodes now suspects his final impression was mistaken, 'that Hick (despite what he sometimes says) does not think that the gods, etc., of the actual religions - Allah, Shiva, Brahman, the Holy Trinity, and so on - are real at all. '8 I confess that it is bit difficult to determine exactly what Hick's own view is of the ontological status of the personal and impersonal manifestations of the Real in different religious traditions. In part this is due to the fact that in his writings Hick often expresses the view of these deities.that is held by the various religious faiths, quite apart from what his own view may be. It is also very important for him to present the views of the ontological status of these various deities that are compatible with his theory of religious pluralism. Given that there is more than one such view compatible with his theory, we can safely infer that his own view is to be found among them. In the end, however, I think that Hick's own view is probably quite close to the view that Mavrodes has come to attribute to him. In short, I think that Hick's own view comes closest to the view that the gods are projections of the religious imagination, creations of the human mind through which we encounter what is truly ultimate reality. That is, although no such beings actually exist, they are not simply the mental products of inner psychological needs, as Freud and some religious sceptics would say. They are mental products that are appropriate in view of human encounters with what is truly ultimate and beyond all literal description, the Real itself. Here then, is what I take to be Hick's reasoning on this matter. In Christianity, God is held to be a trinity and the sole creator of the world. In Islam, Allah is held to be a purely unitary being and the sole creator of the world. And other monotheistic faiths may hold that their gods are the sole creators of the world. Quite apart, then, from Hick's theory of religious pluralism, it is apparent that not all of these gods can be real beings, for it is impossible that there should exist several different beings, each of which is the sole creator of the world. So, whether Hick's theory of religious pluralism is true or false, it is clear that a polytheism in which all these deities, so described, are real beings is impossible. At best then, a polytheism in which 7 8

George I. Mavrodes 'Response to Hick', Faith and Philosophy, 14 (1997), 289-294. Ibid., 290.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGIOUS PLURALISM

425 143

these beings are reduced to beings whose properties do not entail the nonexistence of other deities can be true. And such a polytheism is compatible with Hick's theory of religious pluralism. For, shorn of their incompatible attributes - being the sole creator of the world, etc. - each deity can be a real being through which the Real is manifested to the faithful in the various monotheistic religious traditions. And similar remarks can be made concerning the impersonal absolutes in other religious traditions. So, given Hick's hypothesis of religious pluralism, Mavrodes's original view that Hick is a polytheist is not altogether incorrect. For Hick explicitly allows that religious pluralism can accommodate such a view. But just as clearly, Hick nowhere suggests that it is his own view. In fact, he notes that * this model will involve extremely awkward issues concerning the relations between the deities and their respective spheres of operation'.9 What is more likely to be his own view is the model on which the gods of the various faiths are 'projections of the religious imagination'. On this model, as I noted above, there are no such real beings or absolutes as Allah, Shiva, Brahman, the Holy Trinity, etc. Are they then merely hallucinations, like Macbeth's dagger? Hick rejects this suggestion because he thinks that these imaginary beings serve as the means by which the Real is experienced in human life. Noting that on this model there are no such real beings, Hick adds: 'But neither on the other hand would they be mere hallucinations, devoid of any objective ground. They would be analogous to what have been called in the literature of parapsychology "veridical hallucinations"'.10 The idea here seems to be that, insofar as religious believers take their experiences to be of objectively existing beings (Jahweh, Allah, Vishnu, the heavenly father of Christ, etc.), their experiences are to be counted as hallucinatory. For the objects they take to be real beings are in fact imaginary, not real. On the other hand, insofar as these non-existing, imaginary entities are vehicles through which the religious believer is appropriately affected by the Real, the experiences are to be counted as veridical. Which of these models does Hick personally think is closest to the truth? As I've indicated, I suspect it is the model on which the gods of the various religions are imaginary beings.11 But Hick's main point in An Interpretation of Religion is to note that his theory of religious pluralism can accommodate either model. And this being so, he concludes: 'It therefore seems wise not to insist upon settling a difficult issue which, in logic, the hypothesis itself leaves open'.12 9

10 Hick An Interpretation of Religion, 275. Ibid., 273. Hick sometimes describes the gods of the various religions as 'experientially real' (An Interpretation of Religion, 242). My gloss on this is that the experiential status of something is settled by how it is experienced. Someone who is hallucinating an object may experience that object as real. But if the object is just a hallucination, its ontological status will be that of an imaginary object. 12 Hick An Interpretation of Religion, 275. 11

426

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 144

W I L L I A M L. ROWE III

Having given a brief account of Hick's theory of religious pluralism, and having described what appears to be his own view of the ontological status of the personal gods and impersonal absolutes worshipped in the world's major religious traditions, it is time to turn our attention to his discussion of the transcendent reality, the Real. To guide us into Hick's discussion of the Real, I will ask two questions and search for an answer to them in Hick's writings. Q,i: Is there a reality designated by Hick's expression 'the Real?' Qa: What are some of the characteristics a reality must have if it is what is designated by Hick's expression 'the Real?' A moment's reflection on these two questions should enable us to see that if we don't already know the answer to Qi, we cannot make any significant progress toward answering it until we learn the answer to Q2. And since I suspect many of us don't know the answer to Qi, I propose that we begin with Qa and look for an answer to it in Hick's writings. Before seeking Hick's answer to Q2, however, we need to note an ambiguity in the expression 'the Real'. For Hick's Kantian approach requires a distinction between the Real as it is in itself and the Real as it is experienced in the personal gods and impersonal absolutes of the great religious faiths. Much can be said about the personal gods and impersonal absolutes (the Real as it is experienced in the great religions). So, let us understand both Qi and Q2 as questions about the Real as it is in itself. What characteristics, then, must be possessed by whatever is designated by Hick's notion of the Real as it is in itself?13 Although allowing that the Real has a nature consisting of one or more properties, Hick holds that the Real is ineffable. He explains his view as follows: By * ineffable' I mean (with a qualification to be mentioned presently) having a nature that is beyond the scope of our networks of human concepts. Thus the Real in itself cannot properly be said to be personal or impersonal, purposive or nonpurposive, good or evil, substance or process, even one or many. However, in denying, for example, that the Real is personal one is not thereby saying that it is impersonal, but rather that this conceptual polarity or dualism does not apply. And the same with the other dualisms'.14

The qualification Hick introduces is due to the fact that in saying that the Real is ineffable, one can hardly go on to say that the Real does not have the characteristic of being ineffable. So, it isn't true that absolutely none of our human concepts apply to the Real. Indeed, Hick allows that some purely 18 14

Unless otherwise noted, I will henceforth us use 'the Real' as short for 'the Real in itself'. Hick The Rainbow of Faiths, 27-28.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGIOUS PLURALISM

427 145

formal concepts do apply to the Real. In An Interpretation of Religion he observes that 'it would not make sense to say of X that none of our concepts apply to it. For it is obviously impossible to refer to something that does not even have the property of "being able to be referred to."'.15 And in The Rainbow of Faiths he allows that some additional description of the Real is required in order to pick it out, and proceeds to offer such a description: ... it would not make sense to speak of an X about which nothing can be said except that it can be referred to. But that's not the case here. This X is postulated as that which there must be if religious experience, in its diversity of forms, is not purely imaginative projection but is also a response to a transcendent reality.16

Here then is another property being ascribed to the Real in itself: being such that if it were not real then religious experience, in its diversity of forms, would be a purely imaginative projection. However, I believe Hick thinks of this property as a formal property of the Real, not a substantial property like * being good', 'being powerful', and the like.17 Of course, if religious experience is not purely imaginative projection, The Real in itself could not have this formal property unless it had the property being real. And being real may well strike us as a substantial property. But there is a tradition associated with Kant in which existence is not taken to be a real predicate. So, perhaps Hick can consistently hold the position he takes in An Interpretation of Religion: that no substantial properties are applicable to the Real. Hick doesn't define what is meant by a formal property, as opposed to a substantial property. But he does give numerous examples of substantial properties, both positive and negative. I assume that a formal property of the Real is some abstract characteristic the Real has that is a condition for our being able either to refer to it or to postulate it as that which is encountered through the personal deities and impersonal absolutes of the major religious traditions. And I assume that a substantial property of the Real would be a property that belongs to its essential nature. And Hick's position, I take it, is that none of our concepts can express any substantial property belonging to the nature of the Real. As he puts it in The Rainbow of Faiths, using' intrinsic qualities' in place of'substantial properties', to say that the Real is ineffable 'means that we cannot properly attribute intrinsic qualities to it,... it means that its nature, infinitely rich in itself, cannot be expressed in our human concepts'.18 If the above is a reasonably correct account of Hick's view, it should be clear that the Real in itself cannot be precisely what Hick so often says it is: a reality that totally transcends the network of human concepts. To say it 15

I6 Hick An Interpretation of Religion, 239. Hick The Rainbow of Faiths, 59-60. This description replaces the Anselmian description 'that than which no greater can be conceived', in IR, a description Hick was led to abandon in light of objections to the effect that it implied that the Real has certain positive substantial properties. See The Rainbow of Faiths, 60, n. 12. 18 Hick The Rainbow of Faiths, 28. 17

428

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 146

W I L L I A M L. R O W E

transcends the network of human concepts implies that the Real does not transcend all human concepts, for 'transcends5 is itself a human concept. It would be better, I think, to say only that the intrinsic qualities constituting the nature of the Real transcend our human concepts. And, indeed, Hick does occasionally put the point in this more limited, cautious way. For example, as we've noted, in The Rainbow of Faiths he explains that by 'ineffable5 he means 'having a nature that is* beyond the scope of our networks of human concepts5.19 This is promising since it allows that formal and perhaps relational characteristics of the Real can be expressed within the network of human concepts. But, some pages later, he seems to take it all back by saying ' It has its own nature, presumably infinite in richness, but that nature is not thinkable in our human terms - and indeed even the concept of a nature, or an essence, belongs to the network of human concepts which the Real totally transcends5.20 Unfortunately, some pages earlier he has characterized the Real as 'an ultimate ineffable reality which is the source and ground of everything5.21 The property being the source and ground of everything surely appears to be a non-formal property, and one that is expressible in our human conceptual systems. Perhaps, however, by distinguishing properties constituting the nature of the Real from formal properties and relational properties that are not constitutive of the nature of the Real, these infelicities can all be parsed in a way that leaves intact what seems to be the most important point Hick needs to make about the Real: that it is ineffable in the sense that the properties constituting its nature are inexpressible by human concepts. For such distinctions may permit us to ascribe to the Real the features it must have if Hick5s theory of religious pluralism is true. If the Real is experienced as personal in religion X and experienced as non-personal in religion Y, we can attribute both relational properties to the Real in itself. But in doing so we won't be expressing any of the intrinsic properties that make up the nature of the Real. My chief difficulty with Hick's idea of the Real is that I cannot see how the Real can avoid having one or the other of two contradictory properties. As we noted earlier, Hick claims that the Real' cannot be said to be one or many, person or thing, substance or process, good or evil, purposive or nonpurposive5.22 Most of the dualisms Hick here mentions involve contraries, like' good or evil5. But' purposive or non-purposive5 suggests contradictories, not contraries. Philip Quinn says that a charitable reading ' demands that we construe as contraries, not contradictories, all the pairs of attributes both of whose members Hick denies are possessed by the noumenal real5.23 While this indeed may be charitable, I don't think it is a correct reading. For Hick, 19

20 21 Ibid., 27. Ibid., 62. Ibid., 27. Hick An Interpretation of Religion, 246. 23 Philip Quinn 'Towards thinner theologies: Hick and Alston on religious diversity', International Journal for Philosophy of Religion, 38 (1995), 145-164. 22

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGIOUS PLURALISM

429 147

I believe, is committed to the view that many contradictory dualisms, as well as contrary dualisms, are inapplicable to the Real, meaning that the Real . possesses neither one of the pair of dualisms. Why does Hick hold this view? I believe he has two reasons. Here is the first: But in denying that the Real is personal one is not saying that it is impersonal, but rather that the personal-impersonal dualism does not apply here. To ask whether the Real is personal or impersonal would be misleading, because it presupposes that it's an entity of the kind that could be personal or impersonal. And the same with the other dualisms.24

Suppose we distinguish contrary properties - being hot or cold; being green or red, being personal or impersonal, etc. - from contradictory properties being hot or non-hot, being green or non-green, being personal or nonpersonal. The water in the pot need be neither hot nor cold, but it must be either hot or non-hot. Hick will agree that the water in the pot must be hot or non-hot, because water is the sort of stuff that could be either - a liquid can be hot and it can be non-hot. How about the musical note, middle C, and the contradictory properties of being sharp or non-sharp? Here Hick can say that middle C has the property of being non-sharp even though it is logically impossible for middle C to be sharp. For middle C is the sort of thing — a musical note - that can be sharp or non-sharp.25 But consider the number two and the contradictory properties of being green or non-green. The number two cannot be green. Nor is the number two the sort of thing - a number - that can be green or non-green. Unlike musical notes that can be sharp, no number can possibly be green. So what must we say about the number two and the question of whether it is green or non-green? Well, it seems evident to me that since the number two cannot possibly be green, it must of necessity be non-green. The conclusion I've just reached, however, conflicts with Hick's argument cited above. For that argument, applied to our present example, implies that the green/non-green dualism does not apply to the number two. According to Hick's argument, to ask whether the number two is green or non-green would be misleading, for the question presupposes that the number two is an entity of the kind that could be green or non-green. And clearly, since no number can be green, the number two is not an entity of the kind that could be green or non-green. My response to this argument is that even though to ask whether the number two is green or non-green may be to presuppose that it's an entity of the kind that could be green or non-green, and would thus be an inappropriate or senseless question if asked by someone who knows that no number can be green, it hardly follows that the proposition that the number two is non-green is false or in some way meaningless. Indeed, the 24 25

Hick The Rainbow of Faiths, 61. That is, some musical notes are sharp while others are non-sharp.

430

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION 148

W I L L I A M L. R O W E

proposition that the number two is non-green is necessarily true. And it is precisely because every number must be non-green that it would make no sense for someone who is aware of that fact to ask whether the number two is green or non-green. In my view, the senselessness of the informed person's asking whether the number two is green or non-green does not rest on the question implying that a number could be green or non-green, but on the person's implying by his question that (he thinks) a number could be either. Given that the person knows that numbers are necessarily non-coloured, it would make no sense for him to ask whether the number two is green or non-green. For since he knows that numbers are necessarily non-coloured, he presumably knows that the number two is necessarily non-green. But, as I've suggested, from the fact that it would not make sense for him to ask whether the number two is green or non-green, it hardly follows that the number two is neither green nor nongreen. Consider the relational property, being identical with this pencil. Clearly, we all agree that this property is one that no human being possesses. We may also agree that no human being could possibly possess that property. What, then, of the property, not being identical with this pencil? Since Socrates does not have the property of being identical with this pencil, does he therefore have the property of not being identical with this pencil? How could he fail to have it? He could fail to have it only by having the property we all agree he cannot have, the property of being identical with this pencil. Of course, knowing all this, it would be inappropriate, perhaps senseless, for me to ask seriously whether Socrates is identical with this pencil or not identical with this pencil. For by asking this question in a serious way I imply that, so far as I know, he could have the property of being identical with this pencil. But the fact that I cannot seriously ask this question doesn't show that Socrates must lack the property of not being identical with this pencil. Consider Moore's example, 'It is raining outside, but I don't believe it'. As Moore noted, it could be true both that it is raining outside and that I don't believe that it is raining outside. But there is, nevertheless, something paradoxical in my sincerely asserting 'It is raining outside but I don't believe it'. For in asserting that it is raining outside, I imply that I believe it is. And when I go on to add that I don't believe it is raining outside, I appear to be rejecting what I have implied in assertively uttering 'It is raining outside'. But from the fact that it would be inappropriate and paradoxical, if not senseless, for me to assertively utter 'It is raining outside, but I don't believe it', it hardly follows that what I assert cannot be true. And I think a similar point holds for 'Is the number two green or non-green?' The fact that it would be inappropriate or senseless for the informed person to seriously ask this question doesn't justify the conclusion that the number two has neither the property of being green nor the property of being non-green.

OTHER WRITINGS IN THE PHILOSOPHY OF RELIGION RELIGIOUS PLURALISM

431 149

In light of these remarks, suppose we ask whether the Real in itself is good or non-good, personal or non-personal. Since Hick clearly states that the Real lacks the property of being good, and lacks the property of being personal, by my lights his view implies that the Real in itself has the property of being non-good, and has the property of being non-personal. Of course, if Hick were to agree that the Real is non-personal, this could create a serious difficulty for the assessment of religions favouring personal deities as opposed to religions favouring non-personal absolutes. For if the Real is non-personal, the religions favouring non-personal absolutes might more closely approximate the Real in itself than do the religions favouring personal deities. Writing of the Real itself, Hick says: ... if we regard the major religious traditions as humanly conditioned responses to such a reality we have a reason to think that these concepts do not apply to it namely, as I pointed out just now that if they did it would have mutually contradictory attributes, such as being personal and being non-personal, being a creator and not being a creator and so on. So if, in view of their fruits in human life, you regard Buddhism, advaitic Hinduism, and Taoism, as well as the theistic faiths, as responses to the ultimate, you must postulate a reality to which these conceptual dualisms do not apply, although it is nevertheless humanly thought and experienced by means of them.26

This passage can only be read as asserting that neither one of these pairs of contradictory attributes is applicable to the Real in itself. And I think there is a good reason why Hick needs to hold that the Real in itself is neither personal nor non-personal. For if the Real in itself should turn out to be personal, as well as the creator of all that is, we run the risk of demoting the Real in itself into one of the personal gods of the theistic religions. And it was, in part, to avoid the problem of holding that one religious faith has the truth, while the others are substantially false, that led Hick to postulate the Real in itself in the first place. So, in order to avoid this possibility, Hick is driven to postulate the Real in itself, declaring that it cannot have either one of the pairs of contradictory properties that get exhibited among the personal gods and non-personal absolutes of the major religious traditions. This is a laudable aim. Unfortunately, it requires the price of postulating a reality that is in itself neither personal nor non-personal, neither finite nor non-finite, neither loving nor non-loving, neither spiritual nor non-spiritual, etc. Our first question, * Is there a reality designated by Hick's expression " the Real?"', can now be dealt with. Hick's answer is that he does not know. He postulates the Real in order to provide a favourable account of the religious diversity in the world, an account he thinks provides the best explanation of the apparent equal ability of these diverse faiths to produce persons of great moral and religious sanctity. But since he thinks there is no way of establishing that religions are not simply illusions, he is not prepared to declare 26

Hick The Rainbow of Faiths, 64 (italics mine).

432

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION I5O

W I L L I A M L. R O W E

that there is a reality designated by 'the Real'. My own answer is that there is no reality designated by Hick's expression 'the Real'. I say this because Hick takes it to be a necessary feature of the Real in itself that it does not have the property of being personal.27 That is, he thinks it is true that the Real in itself is not personal. But if it is. true that the Real in itself is not personal, it is exceedingly difficult to deny, as Hick does, that the Real is non-personal.28 Indeed, I take it to be a necessary truth, if not a truth of logic, that whatever is real is either personal or non-personal. So, by my lights there can be no such thing as Hick's Real in itself.29 27 From the fact that Hick takes it to be a necessary feature of the Real in itself that it does not have the property of being personal, it doesn't strictly follow that the Real doesn't have that property. Hick could be mistaken about this point. But I am here taking Hick's conception of the Real to be definitive of the nature of the Real. 28 One can perhaps delay the denial by taking the 'not' in 'the Real in itself is not personal' as an external, rather than an internal negation. So, 'the Real in itself is not personal' may be taken to express . something like 'it is not the case that the Real in itself is personal'. But, on a Russellian account, the external reading in conjunction with Hick's assumption that' the Real' does refer to a genuine reality will imply the internal reading* on which the Real is asserted to not have the property of being personal. 29 I'm grateful to Michael Bergmann, John Hick, and William Wainwright for insightful comments on an earlier draft of this paper.

William L. Rowe: A Bibliography 1. Books authored 1. Religious Symbols and God: A Philosophical Study ofTillich's Theology, Chicago, IL: The University of Chicago Press, 1968. 2. The Cosmological Argument, Princeton, NJ: Princeton University Press, 1975. Reprinted (with a new foreword) by Fordham University Press, 1998. 3. Philosophy of Religion: An Introduction, first edition: Encino, CA: Dickenson Publishing Company, 1978; second edition: Belmont, CA: Wadsworth Publishing Company, 1993; third edition: Belmont, CA: Wadsworth/Thomson Learning, 2001; fourth edition: Belmont, CA: Thomson Learning, 2007. 4. Thomas Reid on Freedom and Morality, Ithaca, NY: Cornell University Press, 1991. 5. Can God Be Free?, Oxford: Clarendon Press, 2004. 2. Books edited 1. Philosophy of Religion: Selected Readings, edited by William L. Rowe and William J. Wainwright; first edition: New York: Harcourt Brace Jovanovich, 1973; second edition: San Diego, CA: Harcourt Brace Jovanovich, 1989; third edition: Fort Worth, TX: Harcourt Brace College Publishers, 1998. 2. God and the Problem of Evil, edited by William L. Rowe, Maiden, MA: Blackwell Publishers, 2001. 3. Articles, book chapters, encyclopedia and dictionary entries 1. "The Fallacy of Composition," Mind 71 (1962): 87-92. 2. "C. B. Martin's Contradiction in Theology," Australasian Journal of Philosophy 40 (1962): 75-79. 3. "The Meaning of 'God' in Tillich's Theology," The Journal of Religion 42 (1962): 274-86. 4. "Tillich's Theory of Religious Myths," Universitas 2 (1964): 1-13. 5. "Augustine on Foreknowledge and Free Will," The Review of Metaphysics 18 (1964): 356-63. 6. "Tillich's Theory of Signs and Symbols," The Monist 50 (1966): 593-610. 7. "The Cosmological Argument and the Principle of Sufficient Reason," Man and World 1 (1968): 278-92. 8. "Two Criticisms of the Cosmological Argument," The Monist 54 (1970): 441-59. 9. "The Cosmological Argument," Nous 5 (1971): 49-61.

434

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

10. "Neurophysiological Laws and Purposive Principles," The Philosophical Review 80 (1971): 502-508. 11. "The Ontological Argument," in Joel Feinberg (ed.), Reason and Responsibility: Readings in Some Basic Problems of Philosophy, 3rd ed. (Encino and Belmont, CA: Dickenson Publishing Company, 1975), pp. 8-17. 12. "Skepticism and Beliefs about the Future," Philosophical Studies 30 (1976): 105-109. 13. "The Ontological Argument and Question-Begging," International Journal for Philosophy of Religion 7 (1976): 425-32. 14. "Comments on Professor Davis' 'Does the Ontological Argument Beg the Question?'" International Journal for Philosophy of Religion 7 (1976): 443-47. 15. "The Fatalism of 'Diordorus Cronus'" (with Rod Bertolet), Analysis 39 (1979): 137-38. 16. "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16 (1979): 335^1. 17. "On Divine Foreknowledge and Human Freedom: A Reply," Philosophical Studies 37 (1980): 429-30. 18. "Fatalism and Truth," Southern Journal of Philosophy IS (1980): 213-19. 19. "Religious Experience and the Principle of Credulity," International Journal for Philosophy of Religion 13 (1982): 85-92. 20. "Two Criticisms of the Agency Theory," Philosophical Studies 42 (1982): 363-78. 21. "Self-Existence and the Cosmological Argument," Analysis 43 (1983): 61-62. 22. "Rationalistic Theology and Some Principles of Explanation," Faith and Philosophy 1 (1984): 357-69. 23. "Evil and the Theistic Hypothesis: A Response to Wykstra," International Journal for Philosophy of Religion 16 (1984): 95-100. 24. "The Empirical Argument from Evil," in Robert Audi and William J. Wainwright (eds), Rationality, Religious Belief, and Moral Commitment: New Essays in the Philosophy of Religion (Ithaca, NY: Cornell University Press, 1986), pp. 227-47. 25. "Modal Versions of the Ontological Argument," in Louis P. Pojman (ed.), Philosophy of Religion: An Anthology, 1st ed. (Belmont, CA: Wadsworth Publishing Company, 1987), pp. 69-73. 26. "Causality and Free Will in the Controversy between Collins and Clarke," Journal of trie History of Philosophy 25 (1987): 51-67. 27. "Two Concepts of Freedom," (Presidential Address, Central Division) Proceedings and Addresses of the American Philosophical Association 61 (1987): 43-64. 28. "Reid's Conception of Human Freedom," The Monist 70 (1987): 430^1. 29. "Response to Dicker's 'A Refutation of Rowe's Critique of Anselm's Ontological Argument'," Faith and Philosophy 5 (1988): 203-205. 30. "Evil and Theodicy," Philosophical Topics 16 (1988): 119-132. 31. "Causing and Being Responsible for What is Inevitable," American Philosophical Quarterly 26 (1989): 153-59. 32. "Analytical Philosophy and Tillich's Views on Freedom," in Gert Hummel (ed.), God and Being: The Problem of Ontology in the Philosophical Theology of Paul Tillich [Gott und

BIBILIOGRAPHY

435

Sein: das Problem der Ontologie in der philosophischen Theologie Paul Tillichs] (Berlin: Walter De Gruyter Press, 1989), pp. 201-210. 33. Entries for "Samuel Clarke" and "Thomas Reid" in Lawrence C. Becker and Charlotte B. Becker (eds), Encyclopedia of Ethics (New York: Garland Publishing, 1992). 34. "Paradox and Promise: Hick's Solution to the Problem of Evil," in Harold Hewitt, Jr (ed.), Problems in the Philosophy of Religion: Critical Studies of the Work of John Hick (London: Macmillan, 1991), pp. 111-24. 35. "Response to Linda Zabzebski," in Harold Hewitt, Jr (ed.), Problems in the Philosophy of Religion: Critical Studies of the Work of John Hick (London: Macmillan, 1991), pp.13033. 36. "Ruminations about Evil," Philosophical Perspectives, 5, Philosophy of Religion, 1991, edited by James E. Tomberlin (Atascadero, CA: Ridgeview Publishing Company, 1991), pp. 69-88. 37. "Responsibility, Agent-Causation, and Freedom: An Eighteenth-Century View," Ethics 101 (1991): 277-97. 38. "The Rationality of Religious Belief," Contemporary Philosophy 14 (1992): 3-9. 39. "The Problem of Divine Perfection and Freedom," in Eleonore Stump (ed.), Reasoned Faith: Essays in Philosophical Theology in Honor of Norman Kretzmann (Ithaca, NY: Cornell University Press, 1993), pp. 223-33. 40. "John Hick's Contribution to the Philosophy of Religion," in Arvind Sharma (ed.), God, Truth and Reality: Essays in Honour of John Hick (New York: St. Martin's Press, 1993), pp. 18-23. 41. "The Problem of No Best World," Faith and Philosophy 11 (1994): 269-71. 42. "William Alston on the Problem of Evil," in Thomas D. Senor (ed.), The Rationality of Belief and the Plurality of Faith: Essays in Honor of William P. Alston (Ithaca, NY: Cornell University Press, 1995), pp. 71-93. 43. "Religion within the Bounds of Naturalism: Dewey and Wieman," International Journal for Philosophy of Religion 38 (1995): 17-36. 44. Entries for "Agent Causation", "Antinomianism", "Causa Sui", "Clarke, Samuel", "Immanence", "Naturalism, Theological", "Prime Mover", and "Privation" in Robert Audi (ed.), The Cambridge Dictionary of Philosophy (Cambridge: Cambridge University Press, 1995). 45. "The Evidential Argument from Evil: A Second Look," in Daniel Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington, IN: Indiana University Press, 1996), pp. 262-85. 46. "Evil, Problem of," in Donald M. Borchert (editor-in-chief), The Encyclopedia of Philosophy. Supplement (New York: Macmillan Reference USA, 1996), pp. 160-62. 47. "Cosmological Arguments" in Philip L. Quinn and Charles Taliaferro (eds), A Companion to Philosophy of Religion (Oxford: Blackwell, 1997), pp. 331-37. 48. "Circular Explanations, Cosmological Arguments, and The Principle of Sufficient Reason", Midwest Studies in Philosophy 21 (1997): 188-201. 49. "Clarke and Leibniz on Divine Perfection and Freedom", Enlightenment and Dissent (Special Issue on Samuel Clarke), 16 (1997): 60^82.

436

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

50. "Death and Transcendence", in D.Z. Phillips and Timothy Tessin (eds), Religion without Transcendence? (Basingstoke: Macmillan, 1997), pp. 133-148. 51. "God and Evil," The Annual Proceedings of the Center for Philosophic Exchange 28 (199798): 4^15. 52. Entries for "Agnosticism", "Atheism", "Deism", and "Freedom, Divine" in Edward Craig (ed.), Routledge Encyclopedia of Philosophy (London: Routledge, 1998). 53. "Reply to Plantinga," Nous 32 (1998): 545-52. 54. "In Defense of 'The Free Will Defense'," International Journal for Philosophy of Religion 44 (1998): 115-20. 55. "Thomas Reid", entry in Robert L. Arrington (ed.), A Companion to the Philosophers (Maiden, MA: Blackwell, 1999), pp. 469-71. 56. "The Problem of Divine Sovereignty and Human Freedom," Faith and Philosophy 16 (1999): 98-101. 57. "Religious Pluralism," Religious Studies 35 (1999): 139-50. 58. "Evil and God's Freedom in Creation," American Philosophical Quarterly 36 (1999): 10113. 59. "The Metaphysics of Freedom: Reid's Theory of Agent Causation," American Catholic Philosophical Quarterly 74 (2000): 425^6. 60. "Skeptical Theism: A Response to Bergmann," Nous 35 (2001): 297-303. 61. "Grounds for Belief in God Aside, Does Evil Make Atheism More Reasonable than Theism?" in William Rowe (ed.), God and the Problem of Evil (Maiden, MA: Blackwell, 2001), pp. 124.-37. 62. "Reply to Howard-Snyder and Bergmann," in William Rowe (ed.), God and the Problem of Evil (Maiden, MA: Blackwell, 2001), pp. 155-58. 63. "Can God Be Free?" Faith and Philosophy 19 (2002): 405-24. 64. "Alternate Possibilities and Reid's Theory of Agent-Causation," in David Widerker and Michael McKenna (eds),_Mora/ Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities (Aldershot: Ashgate, 2003), pp. 219-50. 65. "Divine Freedom," in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, Summer 2003 edition, URL = . 66. "Reflections on the Craig-Flew Debate," in Stan W. Wallace (ed.), Does God Exist? The Craig-Flew Debate (Aldershot: Ashgate, 2003), pp. 65-73. 67. "Thomas Reid's Theory of Freedom and Responsibility," in Terence Cuneo and Rene van Woudenberg (eds), The Cambridge Companion to Thomas Reid (New York: Cambridge University Press, 2004), pp. 222^2. 68. "Cosmological Arguments," in William E. Mann (ed.), The Blackwell Guide to the Philosophy of Religion (Oxford: Blackwell, 2005), pp. 103-16. 69. "Divine Power, Goodness, and Knowledge," in William J. Wainwright (ed.), The Oxford Handbook of Philosophy of Religion (Oxford: Oxford University Press, 2005), pp. 15-34. 70. "Response to Hasker," Religious Studies 41 (2005): 463-66. 71. "Response to Almeida," Philosophical Papers 35 (2006): 27-28.

BIBILIOGRAPHY

437

72. "Friendly Atheism, Skeptical Theism, and the Problem of Evil," International Journal for Philosophy of Religion 59 (2006): 79-92. 73. "Free Will, Moral Responsibility and the Problem of 'Oomph'," The Journal of Ethics 10 (2006): 295-313. 74. "Evil, The Problem of [Addendum]", in Donald M. Borchert (editor-in-chief), Encyclopedia of Philosophy, 2nd ed. (Detroit, MI: Macmillan Reference USA, 2006), vol. 3, pp. 477-78. 75. "Does Panentheism Reduce to Pantheism? A Response to Craig," International Journal for Philosophy of Religion 61 (2007): 65-67. 76. "Paul Tillich", entry in Graham Oppy and Nick Trakakis (eds), History of Western Philosophy of Religion, vol. 5 (Acumen Publishing, forthcoming). 4. Review articles 1. "God and Other Minds," Nous 3 (1969): 259-84 (review of Alvin Plantinga, God and Other Minds). 2. "A Critical Notice of Tillich: An Essay on the Role of Ontology in his Philosophical Theology by Alastair M. Macleod," The Canadian Journal of Philosophy 5 (1975): 615-26. 5. Abstracts 1. "Plantinga on Possible Worlds and Evil," The Journal of Philosophy 70 (1973): 554-55 (abstract of a paper read at a symposium of the Eastern Division Meeting of the American Philosophical Association in December 1973). 2. "The Problem of Evil and Some Varieties of Atheism," The Journal of Philosophy 54 (1977): 731-32 (abstract of an invited paper read at the Eastern Division Meeting of the American Philosophical Association in December 1977). 3. "Rationalistic Theology and Some Principles of Explanation," Nous 17 (1983): 74. 6. Book reviews 1. "The Blue and Brown Books, by Ludwig Wittgenstein," The Chicago Theological Seminary Register 50 (I960): 12-13. 2. "Paul Tillich and the Christian Message, by George H. Tavard," The Chicago Theological Seminary Register 53 (1963): 23-25. 3. "Systematic Theology, Vol. Ill: Life and the Spirit, History and the Kingdom of God, by Paul Tillich," The Philosophical Review 65 (1966): 260-62. 4. "The Logic of Religion, by Jospeh M. Bochenski," The Philosophical Review 76 (1967): 53336.

438

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

5. "Philosophy and Religion, by Axel Hagerstrom," The Philosophical Review 77 (1968): 24850. 6. "Evil and the God of Love, by John Hick," Journal of Philosophy 66 (1969): 271-76. 7. "The Fabric of Paul Tillich's Theology, by David H. Kelsey," The Philosophical Review 78 (1969): 552-54. 8. "God and Timelessness, by Nelson Pike," The Philosophical Review 81 (1972): 372-76. 9. "God and the Universe of Faiths, by John Hick," The Philosophical Review 85 (1976): 13336. 10. "Wittgenstein and Religious Belief, by W. Donald Hudson," International Studies in Philosophy 10 (1978): 247-48. 11. "Divine Commands and Moral Requirements, by Phillip L. Quinn," Philosophical Review 89 (1980): 637-39. 12. "The Existence of God, by Richard Swinburne," Philosophical Books 22 (1981): 174^77. 13. "Does God Have a Nature?, by Alvin Plantinga," The Philosophical Review 92 (1983): 305306. 14. "Theism, by Clement Dore," Faith and Philosophy 3 (1986): 202-206. 15. "Practical Ethics, by Thomas Reid, edited by Knud Haakonssen," Ethics 102 (1991): 187-88. 16. "Mystic Union: An Essay in the Phenomenology of Mysticism, by Nelson Pike," The Philosophical Quarterly 45 (1995), 375-77. 17. "On the Nature and Existence of God, by Richard M. Gale," Journal of the American Academy of Religion 63 (1995): 592-95. 18. "The Metaphysics of Free Will, by John Martin Fischer," Ethics 107 (1996): 141^3. 19. "Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul K. Moser," Australasian Journal of Philosophy 81 (2003): 608-610.

Index (William L. Rowe is referred to as WR throughout the index, except for his own main entry)

Adams, Marilyn 143 Adams, Robert 267-8, 286 on divine freedom 259-60 Alston, William P. xiv, 164 'Bambi' example, critique 141, 142 free will theodicy 148-9, 179-82 WR's critique 149-57, 180 on problem of evil 139-61 evidential argument 139-40 'Sue' example, critique 142, 143, 180-82 theodicies, scrutiny 179-82 on Tillich 9-10, 12-13, 49-50, 57 on values 159-60 Anselm, St belief in God 215 concept of God 354-5, 363 on existence 354 on greatness 355 ontological argument 355 works Monologium 353, 354, 355 Proslogium II/III xvi, 353, 354, 355, 363, 364 Aquinas, St Thomas concept of God 24 cosmological argument 323, 337-8 on creation 286, 287-8 on existence 337-8 Arius 212 Athanasius 212 atheism and existence of evil 61-4, 119, 193-206, 208, 216 friendly 61, 66, 67 indifferent 61 unfriendly 61 see also theism atheist definition 61 theist's view of 66-7 Augustine, St on foreknowledge and free will 371-8

free will theodicy 204-5, 284 on inability to do evil 273 on turning evil to good 75 'Bambi'(suffering fawn) see under evil Bayes's Theorem, application 167-71 belief basic 388-9, 391, 394 criterion for 392 theist's 393 tree example 393, 394-5 see also God, belief in; religious belief Bergmann, Michael, defence of sceptical theism 241-7 WR's response 241-7, 249-52 Berkeley, George 54 Big Bang theory 339 Bonaventure, St 337, 389 Broad, C.D. 379, 383-4 credulity principle 381 Calvin, John 389 Clarke, Samuel xv, 255, 283, 309, 323 on divine freedom 272-3, 273-4, 275-6 on existence of dependent beings 328, 330 Collins, Anthony 273 CORNEA (Condition of Reasonable Epistemic Access) 171, 223, 224 cosmological arguments 323-35, 337-49 Aquinas on 323, 337-8 criticisms, refutation of 307-21 definition 337 Edwards on 317-18 essence of 323-4 Hepburn on 309-10 history of 323, 337 Islamic 337 types 337 WR on xv-xvi, 303-21, 323-35, 337-50 see also God, existence of Craig, William 338 creation, Aquinas on 286, 287-8

440

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

see also God, freedom in creation credulity principle xvi Broad 381, 384 and religious experience 379-85 Swinburne 379, 384 versions 385 dependent beings, existence of 325-30 Clarke on 328, 330 Dewey, John xvii, 401 A Commmon Faith 402, 406, 408 concept of God 404-8, 418 on the religious 403-4 religious humanism 407-8 on supernaturalism 403 Wieman, debate 410, 411-12 divine freedom Clarke on 272-3, 273-4, 275-6 and perfection 255-65 WR on xv, 255-65, 267-79 Robert Adams on 259-60 to do evil 256-7 to do the right thing 257-8 see also God, freedom in creation Edwards, Paul on cosmological argument 317-18 Tillich, criticism of 47-8,49, 54-5, 56 evil Alston on 139-61 and atheism vs theism 193-206 'Bambi' (suffering fawn) example 63, 77-8, 91, 122, 140, 152, 160, 163, 165, 176, 178, 195, 207-8, 233 Alston's critique 141, 142, 144, 145-6, 147, 153-4, 155-9 Delmas Lewis's critique 83-7 pointlessness 207, 208, 210, 224 Wykstra's critique 78-81, 210-11, 224-6 and belief in God 105, 106 and concept of God 105 and divine freedom 256-7, 269-77 and existence of God 69-89, 91-103, 194-6, 203-5, 233-8, 241-7 evidential argument 69-89, 139-40, 163-83, 223-8, 229-32, 233-40, 241-7, 249-52 existence of, and atheism 61-4, 119, 193-205, 208, 216

Hick's theodicy 98, 100-3, 112-14, 145, 216, 229 inability to do 273 intrinsic 105 moral 91 example 92-8 natural 91, 205 necessity for 72-4 pointlessness of 120-21, 210 WR's critique 121-6 problem of Alston on 139-61 Hick on 105-116 and theodicy 98-103 WR on xiv-xv, 119-36, 163-83 Rowe on xiv-xv 'Sue' (murdered girl) example 92, 122, 140, 149-50, 160, 163, 165, 176, 178, 195, 196, 198, 233 Alston's critique 142, 143, 144, 145-6, 150-51, 155-9 free will theodicy 180-82 Wykstra's parent analogy 175-7 suffering, as instance of 61-2, 70 turned into good 75-6 exclusivism 421 inclusivism, distinction 422fn2 existence Anselm on 354 Aquinas on 337-8 contingent things 354 of dependent beings 325-9 Clarke on 328, 330 impossible things 354 of necessary being 304-5, 324 necessary vs possible things 353-4 Expression Thesis 294-5 Fitzpatrick, F.J. 89 foreknowledge, and free will 371-8 Free Will Defence, Plantinga's xv, 124, 125, 187-91 free will theodicy Alston's 148-9 Augustine's 204-5, 284 essence of 204 freedom, Locke on 272

God as being-itself 16-17

INDEX belief in Anselm's 215 as basic belief 389-90, 391, 392, 393, 397 and evil 105, 106 rationality of 387-8, 389-90 Rowe on xvi-xvii concepts of 281, 364 Anselm's 354-5, 363 Aquinas 24 Dewey's 404-8, 418 and evil 105 Wieman's 409-10, 412, 418 doctrine of, Tillich 3-34 existence of cosmological argument 323-4, 337, 353 evidential argument 139-40, 163-86 and evil 69-89, 91-103, 194-6, 203-5, 233-8, 241-7 ontological argument 353-62 a posteriori arguments 353 a priori arguments 353 teleological argument 353 experience of 381-2 freedom in creation 258-65, 281-2, 284, 285-9, 293-9 Adams on 259-60 and evil 269-77 Kretzmann on 287-8 and theism 272 Wainright on 262 WR on 255-65, 267-79, 281-300 as ultimate concern 3-14, 16 see also divine freedom good, intrinsic 106 greatness, Anselm on 355 Hasker, William xv, 130, 131, 132, 134-6 Hepburn, Ronald, on cosmological argument 309-10 Hick, John Mavrodes on 423-4 aspolytheist 423, 425 problem of evil 105-16 on The Real xvii, 424, 425, 426-32 critique 428-31 on religious pluralism xvii, 423-5 theodicy 98, 99-103, 106, 112-14, 145, 216, 229

441

WR's critique 114-16 Howard-Snyder, Daniel xiii, 187 Howard-Snyder, Daniel & Frances, Jove story 289-92 Howard-Snyder, Daniel & Mike Bergmann, WR's response to 249-52 human good as process 413-14 Wieman on 412-17 Hume, David 65, 307, 317 inclusivism 422 exclusivism, distinction 422fn2 Kretzmann, Norman 287-8 Leibniz, G.W. 74 best of all worlds 267, 268-9, 272, 282, 284, 285, 293, 323 Lewis, Delmas 'The Problem with the Problem of Evil' 83 WR's 'Bambi', critique 83-7 Locke, John 54, 55 on freedom 272 Mavrodes, George, on Hick 423-4 Moore, G.E. 65, 108, 125, 326, 375 Morris, Thomas 293, 296, 297-8 Expression Thesis 294-5 naturalism xvii, 401, 419 basic thesis 402 see also Dewey; supernaturalism; Wieman necessary being, existence of 324 nonbeing, Tillich on 11-12, 13 O'Leary-Hawthorne, John 187 ontological argument 387 Anselm 355-6 Gaunilo's critique 357 Kant's critique 357-8 WR's critique 359-62 for God's existence 353-62 Plantinga's modal versions 363-7 WR's critique xvi, 365-7 Organic Unities, Principle of 284 Otto, Rudolf 4, 382 Paul, apostle 81,213,227,380,381

442

WILLIAM L. ROWE ON PHILOSOPHY OF RELIGION

Plantinga, Alvin 180, 389, 392, 393, 394, 395 on evil necessity for 72-4 pointlessness of 120-1 WR's critique 121-6 Free Will Defence xv, 124, 125, 187-91 ontological argument, modal versions 363-7 WR's critique 365-7 sceptical theism 216 WR's response to 233-40 polytheism 423 and religious pluralism 425 Quinn, Philip 260-61, 428 The Real definition 423 Hick on xvii, 424, 425, 426-32 ineff ability 426 Reid, Thomas 271 religion definition 403 philosophy of xiii the religious, Dewey on 403-4 religious belief rationality of 387-99 types 387 see also God, belief in religious experience credulity principle 379-85 definition 379-80 as delusion 382-3 Russell on 384 Saul of Tarsus 380, 381 religious humanism, Dewey 407-8 religious pluralism definition 423 Hick on xvii, 423-5 and polytheism 425 religious symbols 35-58 see also under Tillich Rowe, William L. Alston's theodicy, critique 149-57 on Augustine on divine foreknowledge and human will 371-8 on belief in God xvi-xvii bibliography 433-8 career xiii-xiv

on the cosmological argument xv-xvi, 303-21 on divine freedom xv, 255-65, 267-79, 281-300 Hick's theodicy, critique 114-16 on the ontological argument xvi, 353-62, 363-7 on problem of evil xiv-xv, 61-252, 269-72 on religious experience 379-86 on religious naturalism 401-20 on religious pluralism 421-32 on Tillich's theology xiv, 3-34, 35-58 Russell, Bertrand 308, 312, 314 on religious experience 383 Russell, Bruce 92, 140 Schleiermacher, Friedrich 382 sceptical theism see under theism Smith, Quentin 338 Stace, W.T. 54 states of affairs contingent 282, 283 intrinsic value 284-5 model of 283 possible 282, 283 'Sue' (murdered girl) see under evil suffering gratuitous 140 as instance of evil 61-2, 70 Sufficient Reason, Principle of 304, 309, 315-16, 330-34, 338 supernaturalism, Dewey on 403 Swinburne, Richard credulity principle 379, 384 theodicy 216 symbols see religious symbols theism defence of 64-6 definition 194 empirical, Wieman's 408-19 expanded 81-2, 194, 212, 223, 227 and God's freedom in creation 272 restricted 81, 82, 194, 212, 223 sceptical xv, 209-10, 216, 219, 241 WR's response to Bergmann 241-7, 249-52 standard 129, 212, 223, 227 see also atheism theist

INDEX atheist's view of 66-7 basic belief 267, 393 definition 61 theodicies Alston's scrutiny of 179-82 Hick's, and evil 98, 99-103, 106, 112-14, 145, 216, 229 Leibniz'99 purpose 98-9 Swinburne's 216 see also free will theodicy Tillich, Paul xiii, xiv Alston on 9-10, 12-13, 49-50, 57 atheism 61 on being-itself 13-15 doctrine of God 3-34 as being-itself 3, 7-8, 9, 10, 13, 16, 19, 20-21, 22-5, 27-34, 36, 37, 38-9, 40, 41-2, 51 development 25-7 discourse about 42-4,47 as nonsymbolic 20-21, 41, 47-8, 51 as symbolic 16, 18, 21-2, 36-7, 40, 44-6, 49, 51 as ultimate concern 3-14, 16, 26 Edwards, criticism by 47-8, 49, 54-5, 56 on nonbeing 11-12, 13 religious symbols 35-58 immanent level 35 main problem 50-51 primary 35, 36

443

secondary 35-6 semantic principle 51-4 transcendent level 35 WR's defence 51-8 theism 61 Urban's criticisms of 17-20 Tooley, Michael 160 transworld depravity 187, 188 sanctity 187, 188 Trinity 212 Urban, William M., criticism of Tillich 17-20 values, Alston on 159-60 Wainright, William, on divine freedom 262 Wieman, Henry Nelson xiii, xiv, xvii concept of God 409-10, 412, 418 Dewey, debate 410, 411-12 empirical theism 408-19 on human good 412-17 Wykstra, Stephen 126, 127, 128, 164, 172-6, 209, 213-14, 217 'Bambi' example, critique 78-81, 210-11, 224-6 sceptical theism 209-10 'Sue' example, parent analogy 175-7 WR's response to 223-8 Zagzebski, Linda, WR's response to 229-33